what is data analysis in education

5 Ways Your Institution Can Leverage Your Data Analytics

• Data Analytics 101

Think about all the different data sets your school has. From student GPAs to class section counts, nearly everything can be put into a data set, creating a culture of big data. How are you using all of this information to your school’s advantage?

When used strategically, big data and data analytics in education has the power to transform an institution. As a professor of higher education at North Carolina State University and a previous director of institutional research, Stephen Porter knows this first-hand. “Institutions are rushing with their analytics and not thinking strategically about what data they have and how it can help their predictions,” Porter said.

What is Big Data?

The term “big data” refers to the immense amount of information that’s created and collected on a daily basis. Big data has become so colossal that traditional data management tools, like Excel spreadsheets and manual record-keeping, are no longer able to keep up.

Because the volume of data has significantly increased with the digital revolution, the need for a robust data analytics program is an absolute must if you want to put your information to use in a strategic way.

In particular, educators can use data analytics to record and analyze the following data sets:

• Student Data : Demographics like age, ethnicity and gender; whether they are full-time or part-time; if they take classes online, on campus or a mix of the two.
• Course Data : Enrollment headcounts, grades and completion rates by program or section.
• Instructor Data : Demographics like age, ethnicity and gender; salary information; productivity levels.
• Facility Data : Classroom utilization and resource allocation, like how many hours a week each room is being used.

Benefits of Using Data Analytics Tools

• Create a Data-Driven Decision-Making Culture – With data analytics, there’s no need to rely on blind faith when making decisions. You’ll have numbers and statistics to back up your decisions, which can lead to more successful outcomes.
• Access Data Easier – Data analytics tools depend on the same type of technological infrastructure to capture, store and organize information, so it’s easy to find what you need.
• Find Information More Quickly – Since your data lives in one place, there’s no need to search through dozens of files and folders to find one report, making the process much quicker.

With these in mind, we’ve compiled a list of some of the most effective ways educational institutions can leverage their data analytics.

Help Your Students Succeed

Enrolling a new student is far more expensive than retaining a current one. That’s why one of the most advantageous ways to use data analytics in education involves identifying challenges students may have or are currently having in their academic career.

Predicting Success

When determining which students to accept to your institution, looking at certain academic analytics can tell you which candidates are the most likely to succeed and which may be more likely to drop out or fail their classes. This can help you make a judgment call before they even walk onto your campus.

Say a student was accepted to your institution and wants to study engineering. By using data you already have about this student — like their SAT scores, high school GPA and individual class grades — you can assess whether or not they would be likely to succeed in the engineering program. Did they struggle with math? If so, the engineering track might require additional math support for this particular student. In this instance, the student can meet with their advisor to review other options, like exploring a different program or beginning with remedial math courses.

Help Students Who May Be Struggling

Your data can also help you intervene before a problem has the chance to become a serious issue for you and your students. If a student fails their classes, placing them on academic probation at the end of a semester doesn’t help anyone. But by sharing certain learning analytics between professors and advisors throughout a term, steps can be taken to prevent the student from failing in the first place.

Porter has seen institutions leverage this technique in practical ways. “With the right data and software program, institutions can set up an alert system that notifies a student’s advisor if they are failing their classes,” he said. “The advisor can then intervene and try to find a solution, which is a much more proactive approach compared to just letting a student fall through the cracks.”

Do they need tutoring? Are they struggling to balance their personal life with school? Or are they just in the wrong program? Their advisor will be able to assess the situation and figure out a solution that will benefit both the student and your institution.

Share a Public Fact Book

Deciding on where we pursue our education is one of the biggest decisions we make in our lives, and students spend months — if not years — researching their options. Because of this, data analytics in education is useful when it comes to recruiting and educating applicants.

Assembling data sets into a public fact book that’s accessible on your website is a great way to share information that your prospects need to make a decision. This can include data like:

• Class sizes
• Student to teacher ratios
• Student outcomes

The example below shows the course success rates for French classes. In 2015, the success rate was 61% but it jumped to 83% in 2016. This data would give a student who wants to study French a good idea of how successful the program is, which may make them more likely to attend your institution.

See how customizable reporting software could help you leverage your institution’s data. >>

Evaluate how physical space is being used.

From classrooms to energy consumption, physical space is costing you money. But fortunately, your data can help you make efficient use of every square inch of your campus.

Tracking Resources

Where are your students at specific times of the day and days of the week? If your data tells you that they aren’t very excited to take English courses on Friday afternoons, you can potentially reduce the energy consumption for that part of campus by changing the days or times these courses are offered.

This also helps you plan better. Since you know that no one enjoys a Friday afternoon English class, you can plan these classes to better match your students’ preferences, which could also help increase enrollment.

Section Fill Rates

If ENG 101 offers 11 sections for the fall semester and only four are full, it might be time to figure out why. Is it the time of day? The day of the week? The professor?

This can also help you in the opposite scenario, like if CHEM 202 only offers three sections and they all have a long waitlist. That may be a clear sign that you need to add more sections and/or hire additional professors.

These are all questions you can dive deeper into, which can open up more possibilities that can help you save additional resources.

Track Enrollment Trends

Your data sets will tell you everything you need to know about students who are applying, enrolling and graduating from your institution, which is essential when it comes to planning and recruiting.

For example, the data below shows the enrollments for white females between the ages of 18 and 20 for the fall semesters of 2015 and 2016.

As you can see, there was a significant dip — by 50%! — in enrollments for this demographic in 2016. What could have happened to cause this? Noticing the problem is the first step. Once you’re aware, you can have your institutional research department explore this further and come to a conclusion. Then you can fix it!

Location, Location, Location!

Where do your applicants live? If you can easily see that prospective students from certain cities or towns are applying, being accepted and graduating from your institution, you can better tailor your marketing efforts to students in the same areas. You may also have the flexibility to increase offerings at strategically located satellite campuses.

Improve Communication Between Departments

It’s not surprising that there are silos in campuses all over the country. Programs can have disparate data coming from a variety of sources, which can make it hard to share information between professors, let alone other departments.

However, using data analytics in education — combined with a structured reporting software — can help you build a more collaborative culture that can seriously benefit your institution. For example, sharing analytics about enrollment trends has the potential to save much-needed resources. If the natural science program is losing enrollment while biological science is increasing enrollment, why not combine these programs? They are similar but with different appeals, which can simply be addressed by the way the program is marketed.

Porter believes that institutions “need to break down silos so there is more data at the fingertips of those in charge of predicting and planning.” And while some departments may be wary about sharing their data, it’s important to understand that a lot of it is submitted to the federal government and is more transparent now than ever before. Plus, the more information you have, the easier it becomes to more effectively predict and plan.

How to Avoid Potential Challenges When Getting Started

Leveraging your data to achieve these benefits begins with a solid plan of action. Make sure to do the following before implementing any data analytics solution:

• Create Goals – Your data analytics program will open up a lot of doors for you, so it could become overwhelming if you don’t enter into the new process without clear-cut goals in mind.
• Set Expectations – You’ll need your employees to be on board with your data analytics platform and strategies, which is why you should have open communication and set expectations for your team before implementing a solution.
• Make it Accessible – For full transparency, make sure your employees can access data where and when they need it. That means setting the appropriate permissions and organizing data properly. For example, teachers will need access to different data sets than the administration office.

These are only a small fraction of the ways data analytics in education can empower you to improve the future of your institution. To see more, sign up for the free version of Precision Campus and start putting your data to work!

Our Founder and President Eric Spear has been right where you are: in charge of an unfathomable amount of data and responsible for creating countless end user reports. With the dawn of cloud computing, Eric recognized the opportunity to combine his expertise with the latest technology to support the modern-day campus’ needs and solve many of their previously unsolvable challenges.

Need help finding data analytics software?

7 Tips to Help You Find the Right Higher Education Data Analytics Software

4 Smart Tips to Help You Find the Best Higher Education Data Analytics Software

5 Tips to Help You Succeed with Higher Education Data Software

Toll Free 603.828.2521

Sign up for our Newsletter

This site belongs to UNESCO's International Institute for Educational Planning

IIEP Learning Portal

Search form

• issue briefs
• Monitor learning

Using data to improve the quality of education

There is a worldwide concern that learning outcomes have not kept pace with the expansion of education. The extent of the learning deficit is largely unknown because many countries have few systematic data on who is learning and who is not. Learning assessments provide data on the status of learning, which can be used to monitor the quality of systems and student learning outcomes. Regular monitoring can reveal changes over time in response to interventions to improve student outcomes, providing feedback and additional data for decision-making.

Learning data, in conjunction with other dimensions of quality such as context, teaching and learning environment, and learner characteristics can reveal the factors that most affect learning outcomes. By revealing gaps in student achievement and service provision, data can be used to identify those groups that are being underserved and are underperforming. Once identified, such inequities can be addressed.

Data can be used to hold the system accountable for the use of resources by showing whether increased public investment in education has resulted in measurable gains in student achievement. Although direct accountability for results rests mainly with the school, the enabling policy and practice environment is the responsibility of decision-makers at all administrative levels.

Which actor needs which type of data?

Data-driven decisions to improve learning are taken at each level of the system. The specificity of the data decreases from school to national level and the time lag between data collection and application increases. Decisions concerning individual students, classes, and schools are made locally, where raw data are produced. System-wide decisions based on aggregated data are made nationally.

Classroom teachers

Classroom teachers manage the teaching and learning process. They monitor students’ learning by informal means, such as quizzes and games, and formative tests. Teachers use the data to assess a student’s performance, strengths, weaknesses, and progress. Additional information on an individual student’s background allows the teacher to diagnose possible causes of poor performance and apply remedies. The data can also be used for self-evaluation to identify where teachers could improve their pedagogy or classroom management.

Head teachers

Head teachers assess the school’s overall performance. They examine student achievement and attainment, staff performance, and use of school resources. Head teachers set and monitor school practices, programmes, and policies. They need raw achievement data, information on teachers’ classroom practices and contribution to student outcomes, and on their own performance as rated by supervisors.

Parents and communities

Parents and communities require information on students’ achievement, including their strengths and weaknesses, and any behavioural issues. They are concerned about public examination results, since performance determines their children’s progress to further education or employment. Parents and school staff can discuss and agree an agenda for action to support student needs. Parents can support school improvement through parent-teacher associations and school boards.

District and provincial level actors

District level actors have responsibility for oversight of the management and quality of schools in the district. They collect and aggregate school data on student attendance and achievement, teacher attrition and absenteeism, and resources. They play an important role in the identification of the resource needs of schools, in monitoring standards and recommending improvement measures.

Provincial level administrators, coordinators, and supervisors make decisions based on evidence of an issue serious enough, or an opportunity good enough, to warrant commitment of time and provincial resources. Their focus is on how to plan and use interventions to provide large groups of schools with the resources and expertise to set up and evaluate their education programmes and, guided by evaluation results, to adopt procedures to improve effectiveness.

National level officials

National level officials make broad policy decisions on links between government directives and the plans and resources needed to comply with those directives. They need substantial system-wide information on current student outcomes and associated factors, together with data on long-term trends. These are collected and collated to provide the basis for decisions on the whole or on a major part of the education system. Data sources include EMIS, national examination results, and learning assessments.

What information can the data provide and how can it be used?

Learning data, augmented with background data, provide information on how well students are learning, what factors are associated with achievement, and which groups perform poorly. This information can be used for system analysis, improved resource allocation, agenda setting or during the policy-cycle.

Education system analysis

Education systems may be analyzed in terms of:

• What students are learning;
• Whether what they learn responds to parents’, community, and country needs and aspirations (relevance);
• How well resources are used to produce results (internal efficiency);
• What the main factors influencing learning are; and
• Which aspects of the system require improvement.

If the data show some groups’ learning outcomes are low due to their location, ethnicity, religion, or disability, measures can be taken to provide additional resources, such as teachers or books, aimed at improving their achievement.

Improved resource allocation

The data may reveal issues with the provision and use of resources. School infrastructure, availability of instructional materials, and the use of instructional time influence learning outcomes. Improved instructional materials with information on their use may contribute to better achievement.

Agenda setting and policy-making

According to Clarke (2017), there are differences between countries at different income levels in the focus of their policy and design. Generally, high-income countries with established assessment programmes use data for sector-wide reforms or a programme of interventions aimed at improving learning outcomes. Low-income countries that are beginning to use the programmes tend to identify a few separate issues, such as resource allocation or teacher qualifications, as responsible for poor achievement. Resulting policies include a few discrete interventions.

Data analysis can identify areas that require improvement, from which agenda for action can be designed. For example, Meckes and Carrasco found that in Chile, publication of the correlation between students’ socio-economic status and their achievement prompted demands for policies to address equity issues (Raudonyte, 2019).

Seychelles’ use of SACMEQ findings in 2000 provides an example of using assessment results for policy formulation. SACMEQ data indicated large differences in learning outcomes among pupils in the same school, attributable to a long-established practice of streaming by ability from Grade 1. By Grade 6 the learning achievement between girls and boys had widened to such an extent that there were more girls in the elite class and more boys in the inferior class. Effective communication channels, an enabling political context, and effective dialogue among actors contributed to the decision to adopt a de-streaming policy (Leste 2005 quoted in Raudonyte, 2019).

The regular collection of learning and other related data to monitor policy implementation can inform on the status of planned activities, reveal implementation challenges, pinpoint early indications of impact, and suggest modifications to adjust shortcomings. For example, the Learn to Read initiative in Madhya Pradesh was monitored on a monthly basis through standardized tests to detect shortcomings and adjust implementation (Tobin et al., 2015).

National assessments can be used to gauge the impact of policy on learning outcomes and to provide feedback to address shortcomings. In theory, there should be a seamless progression from testing through agenda setting, policy formulation, implementation, and monitoring and evaluation based on more testing. In practice, such a feedback mechanism is often less well organized. This may be due, among other things, to lack of experience with using assessments, weak technical capacity, poor coordination between assessment and decision-making bodies, and funding shortfalls.

Challenges to data use

For data to be used effectively they must be actionable, available to all who are in a position to act and presented in an appropriate form for each group of stakeholders. Barriers to data use include the following:

Data availability

Inadequate funding of an assessment programme can mean the programme cannot be completed. Delays in analysis can prevent data from being released in a timely manner. Results may be withheld if they are below expectations. Findings may be dismissed if they do not respond to the needs of the system, or are not actionable or linked to viable policy options.

Access problems

Data access problems include: a failure to communicate results to both the public and those who are in a position to act; results retained within a ministry of education to restrict their use by other stakeholders and prevent the media and public from lobbying for action; the content and format of the reports may not be suited to some or all target groups, who need a variety of data and presentation modes.

Quality issues

Issues with the design, relevance, and credibility of the assessment programme can lead to data being withheld or ignored. Real or perceived deficiencies in assessment instrumentation, sampling and analysis can raise validity and relevance issues. Occasional or ill-designed assessments mean that skills and content are not comparable over time. Caution is needed when developing policy messages based on assessment results without an analysis of supplementary data.

Limited capacity and skills to assess and use the data

Ministries of education may lack experience with national assessments, have poorly established decision-making procedures and low technical capacity. Technical personnel may lack expertise in assessment design, in-depth data analysis, and interpretation. This may result in recommendations being superficial and uninformative. Policy-makers may not understand the implications of the assessment or may not focus on the analysis due to time constraints. Data collection, analysis, availability, and use may be adversely affected by funding constraints.

Political climate

Conflict and political unrest may impact assessment implementation. Political sensitivities due to low levels of achievement can prevent data use. There may be a lack of political will to act on a recommendation.

Minimizing the challenges

Credibility and acceptability issues can be addressed by involving all relevant stakeholders in the design and implementation of an assessment. The assessment team should have the technical competence to design, administer the assessment and analyze results. Ongoing technical training of existing and potential staff is necessary to ensure quality and to allow for attrition.

Building local capacity or establishing a regional coordinating body are possibilities. Both options require substantial investment in capacity building that could be costly and time-consuming.

Judicious use of media channels at all stages of the assessment including dissemination of results, and regular stakeholder discussions will ensure the public are kept informed. Distribution will be facilitated if there is a budget for dissemination, a dissemination plan and if the reports prepared are tailored to different users’ needs.

Existing structures, policy-making and decision-making processes within ministries can also be a barrier to data use. In order to adapt to a data-driven decision-making culture, ministries of education may need to restructure and redefine the roles and responsibilities within the organization. Links among staff and with relevant outside institutions need to be established and sustained.

Australia. Department of Foreign Affairs and Trade. 2018. ‘ Education learning and development module: Learning assessment ’ . Canberra: DFAT.

Best, M.; Knight, P.; Lietz, P.; Lockwood, C.; Nugroho, D.; Tobin, M. 2013. The impact of national and international assessment programmes on education policy, particularly policies regarding resource allocation and teaching and learning practices in developing countries. Final report. London: EPPI-Centre, Social Science Research Unit, Institute of Education, University of London.

Birdsall, N.; Bruns, B.; Madan, J. 2016. Learning data for better policy: A global agenda. Washington, DC: Center for Global Development.

Clarke, P. 2017. ‘ Making use of assessments for creating stronger education systems and improving teaching and learning’ . Paper commissioned for the 2017/18 Global Education Monitoring Report, Accountability in education: Meeting our commitments. Paris: UNESCO.

Custer, S.; King, E. M.; Atinc, T. M.; Read, L.; Sethi, T. 2018. Towards data driven education systems: Insights into using information to measure results and manage change . Washington, DC: Center for Universal Education at Brookings/AidData.

De Chaisemartin, T.; Schwanter, U. 2017. Ensuring learning data matters . IIEP-UNESCO Learning Portal.

Mählck, L.; Ross, K. N. 1990. Planning the quality of education: The collection and use of data for informed decision-making . Paris: IIEP-UNESCO.

Postlethwaite, T. N., Kellaghan, T. 2008. National assessments of educational achievement . Paris: IIEP-UNESCO.

Raudonyte, I.2019. Use of learning assessment data in education policy-making . Paris: IIEP-UNESCO.

Ross, K. N. 1997. ‘Research and policy: a complex mix’. In: IIEP Newsletter , 15 (1), pp. 1-–4.

Saito, M. 2015. The use of learning assessments in policy and planning. IIEP-UNESCO Learning Portal.

Tobin, M.; Lietz, P.; Nugroho, D.; Vivekanandan, R.; Nyamkhuu, T. 2015. Using large-scale assessments of students’ learning to inform education policy. Insights from the Asia-Pacific region . Melbourne/ Bangkok: ACER/ UNESCO.

UNESCO; IIEP Pôle de Dakar; World Bank; UNICEF. 2014. Education sector analysis methodological guidelines. Vol. 1: Sector-wide analysis, with emphasis on primary and secondary education . Dakar: UNESCO. IIEP Pôle de Dakar.

UNESCO Office Bangkok and Regional Bureau for Education in Asia and the Pacific. 2013. The use of student assessment for policy and learning improvement . Bangkok: UNESCO Office Bangkok.

UNESCO Office Bangkok and Regional Bureau for Education in Asia and the Pacific. 2017. Analyzing and utilizing data for better learning outcomes . Paris/ Bangkok: UNESCO/ UNESCO Office Bangkok.

UNESCO Office Bangkok and Regional Bureau for Education in Asia and the Pacific. 2017. Large-scale assessment data and learning outcomes: Linking assessments to evidence-based policy making and improved learning . Bangkok: UNESCO Office Bangkok.

UNESCO-UIS. 2018. SDG 4 data digest: Data to nurture learning . Montreal: UIS.

Willms, J. D. 2018. Learning divides: Using data to inform educational policy . Information Paper No. 54. Montreal: UIS.

Related information

• The use of learning assessment data
• Quality of education

What Data Analysis Is and the Skills Needed to Succeed

Use the tools and techniques of data analysis to make sense of mountains of information..

U.S. News & World Report Education takes an unbiased approach to our recommendations. When you use our links to buy products, we may earn a commission but that in no way affects our editorial independence.

Popular Data Analysis Courses

Provider : Udacity

Cost : $399 per month or $1,356 for four months

Skill Level : Intermediate , Advanced

75% off with code: USNEWS2021

Provider : Coursera

Cost : $39 per month after free trial

Skill Level : Beginner

From counting steps with a smartwatch to visiting this website, nearly everything we do generates data. But just collecting statistics, measurements and other numbers and storing the information is not enough. How we harness data is the key to success in our digital world.

Getty Images

What Is Data Analysis and Why Is it Necessary?

How many steps you took today doesn’t mean anything unless you know information like how many steps you took yesterday, how many steps you take on average and how many steps you should be taking.

When you gather information, organize it and draw conclusions and insights, then you can make better decisions, improve operations, fine-tune technology and so on. Data analysis includes evaluating and recapping information, and it can help us understand patterns and trends.

Types of Data Analysis

There are four main types of data analysis: descriptive, diagnostic, predictive and prescriptive. These data analysis methods build on each other like tiers of a wedding cake.

Descriptive Data Analysis

Descriptive statistics tell you what is in the data you’ve gathered. Building blocks include how many data points you have, average and median measurements, the amount of variation within your data, and the certainty those things provide about your results.

Diagnostic Data Analysis

Diagnostic data analysis – also called causal analysis – examines the relationships among data to uncover possible causes and effects. To accomplish this, you might look for known relationships to explain observations or use data to identify unknown relationships.

Predictive Data Analysis

Building on diagnostic data analysis is predictive analysis, where you use those relationships to generate predictions about future results. These “models” can range from equations in a spreadsheet to applications of artificial intelligence requiring vast computing resources.

Predictive modeling is the heart of analysis, says Nick Street, professor of business analytics and associate dean for research and Ph.D. programs at the University of Iowa’s Tippie College of Business.

“My poll needs to be correct about the people who are going to vote, and my self-driving car has to be correct about whether that’s a stop sign or not,” Street says.

Prescriptive Data Analysis

Often, the goal of data analysis is to help make sound decisions. While all types of data analysis can help you accomplish this, prescriptive data analysis provides a deeper understanding of costs, benefits and risks. Basically, prescriptive data analysis helps us answer the question, “What should I do?”

The most common kind of prescriptive analysis is optimization, or figuring out "the best results under the given circumstances," according to a post at Data Science Central. So, given a set of constraints, which inputs provide the most benefit for the lowest cost and least amount of risk. For example, a particular step in surgery might reduce the risk of infection but increase the risk of other complications.

In Street’s work, data can inform a decision by predicting how likely a patient is to get an infection without the step in surgery that is supposed to reduce infection risk. That way, a doctor could determine whether the extra step is actually beneficial, or if the step could be removed from the surgical process.

Of course, while a data analyst can provide the prescriptive analysis, a doctor would need to interpret the probability and make a decision based on the data.

“I’m not qualified to make that decision,” Street says of a data analyst’s role. “I can just tell you that for this person it’s (63%).”

Looking for more course options?

Data Analysis Tools, Techniques and Methods

Data analysis involves a spectrum of tools and methodologies with overlapping goals, strengths and capabilities. Here is how each working part contributes to effective data analysis.

The Data Analysis Phases

There are different ways of looking at the phases of data analysis. Here is a typical framework.

1. Data Requirements

You need to know the questions you want to answer and determine what data you require in order to find the answer.

2. Data Collection

This involves identifying data that might answer your questions, determining what steps are required to gather the data, and understanding what strengths and weaknesses each type of data might present. Not all data is strong or relevant for answering your question.

Charlie McHenry, a partner at consulting firm Green Econometrics, says figuring out which data matters to answer a question might seem difficult, but the information you need is often hiding in plain sight.

For example, consider the data gathered from business systems, surveys and information downloaded from social media platforms. You might also consider purchasing commercial data or using public datasets.

“Every enterprise has a fire hose of collectable data,” McHenry says.

3. Data Cleansing

This is the most delicate stage of data analysis, and it often takes the most time to accomplish. All data comes in “dirty,” containing errors, omissions and biases. While data doesn’t lie, accurate analysis requires identifying and accounting for imperfections.

For example, lists of people often contain multiple entries with different spellings. The same person might appear with the names Anne, Annie and Ann. At least one of those is misspelled, and treating her as three separate people is always incorrect.

4. Data Analysis

The meatiest phase is applying descriptive, diagnostic, predictive and prescriptive analysis to the data. At first, the results may be baffling or contradictory, but always keep digging.

Just be vigilant and look for these common errors:

• False positives that seem important but are actually coincidental.
• False negatives, which are important relationships that are hidden by dirty data or statistical noise.
• Lurking variables, where an apparent relationship is caused by something the data didn’t capture.

5. Data Interpretation

This stage is where a data analyst must practice careful judgment and has the most chance to be wrong. It’s up to an analyst to determine which models, statistics and relationships are actually important.

Then the data analyst must understand and explain what the models do and do not mean. For instance, political scientists and journalists often build models to predict a presidential election by using polls. In 2008 and 2012, those models correctly predicted the results. In 2016, those models showed lower levels of certainty, and the candidate they said was more likely to win did not. By ignoring the change in certainty, many people were shocked by the election results, falling prey to confirmation bias because they only saw data that supported their beliefs about who would win.

6 . Data Visualization

Staring at equations and columns of numbers is not appealing to many people. That’s why a data analyst has to make the numbers “friendly” by transforming data into visuals like charts and graphs. Modern data visualization takes this a step further and includes digital graphics and dashboards of interrelated charts that people can explore online.

Data Analysis Tools

While there are countless tools for each phase of data analysis, the most popular tools break down in the following way:

Data Collection

• SurveyMonkey: Do you need to collect data from your users or customers? There are many tools for online surveys, but SurveyMonkey is popular with analysts for its ease of use, features and capabilities. You can apply it to survey all users, only a random portion or a sample of the public.
• Data.world: There is a lot of data already out there, much more than any person can find just by searching the web. While data.world’s primary emphasis is allowing companies to host and analyze their own data in the cloud, its community portal has a rich set of datasets you can use. Other go-to data collections include: FRED for economic data, ESRI ArcGIS Online for geographic data and the federal government’s Data.gov .
• Google Analytics: Google produces a tool for tracking users online. If you have a website, you can use this free tool to measure virtually any aspect of user behavior. Competitors include Adobe Marketing Cloud, Open Web Analytics and Plausible Analytics.

Data Storage

• Microsoft Excel : The Swiss Army knife of data analysis, current versions of the Microsoft Excel spreadsheet can store up to 1 million rows of data. It also has basic tools for manipulating and visualizing data. Excel is available in desktop, mobile and online versions. Competitors include Google Sheets, Apple’s Numbers and Apache OpenOffice.
• PostgreSQL: One of the most popular of the traditional database systems, PostgreSQL can store and query gigabytes of information split into “tables” for each kind of data. It has the SQL language built in (see below), can be used locally or in the cloud, and can be integrated with virtually any programming language. Competitors include Microsoft SQL Server, Microsoft Access and MySQL.
• MongoDB: This is a popular “nonrelational” database. MongoDB combines data so that all the information related to a given entity, such as customers, is stored in a single collection of nested data. Competitors include Apache CouchDB, Amazon DynamoDB and Apache HBase.

Data Manipulation/Programming

Of course, gathering and storing data aren’t enough. Data analysis involves tools to clean data, then transform it, summarize it and develop models from it.

• SQL: The go-to choice when your data gets too big or complex for Excel, SQL is a system for writing “queries” of a database to extract and summarize data matching a particular set of conditions. It is built into relational database programs and requires one to work. Each database system has its own version of SQL with varying levels of capability.
• R : R is the favored programming language of statisticians. It is free and has a large ecosystem of community-developed packages for specific analytical tasks. It especially excels in data manipulation, data visualization and calculations, while being less used for advanced techniques requiring heavy computation.
• Python : Python is the second-most-popular programming language in the world. It is used for everything from building websites to operating the International Space Station. In data analysis, Python excels at advanced techniques like web scraping (automatically gathering data from online sources), machine learning and natural language processing.

Data Visualization

• Tableau : Analysts swear by this desktop program’s compatibility with nearly any data source, ability to generate complex graphics, and capability of publishing interactive dashboards that allow users to explore the data for themselves.
• Google Data Studio : Similar in some ways to Tableau, this is a web-based tool that focuses on ease of use over complex capabilities. It’s strongly integrated with other Google products, and many say it produces the best-looking results out of the box.
• Microsoft Power BI : No list of data visualization tools would be complete without Microsoft Power BI. It’s tightly linked with Microsoft’s desktop, database and cloud offerings, and focuses on allowing users to create their own dashboards and visualizations.

Data Warehousing

Left flowing, the “fire hose" of data McHenry describes quickly overwhelms most databases. Where can you store a clearinghouse of information? Here are some options:

• Oracle Database : Known as “Big Red,” Oracle is famed for its ability to scale vast quantities of data. Oracle Database allows users to store and analyze big data using familiar database formats and tools like SQL.
• Amazon Redshift : Amazon Redshift is pitched as a more affordable alternative to Oracle Database. As part of Amazon Web Services, it integrates well with their other services, but it can only be used as part of the AWS cloud offerings.
• Domo: Domo combines the capabilities of a data warehouse like Oracle or Amazon Redshift with a functionality similar to Microsoft Power BI. It is used by organizations that want to allow many employees to gain access to a data warehouse.

Example of Data Analysis at Work

Putting together all the pieces of the data analysis puzzle might seem complex, but the time and resources required are worth the gains, says Pentti Tofte, vice president and head of analytics at the property insurer FM Global.

FM’s goal is not just to set insurance rates, but also to help customers reduce them, Tofte says. His inspectors visit more than 100,000 properties every year and record more than 700 pieces of data. Combining that information with data related to risks like fires and hurricanes, FM can then provide recommendations to the companies it insures.

“We believe most loss is preventable,” Tofte says. “We use data to tell them what losses to expect where and which vulnerabilities to prioritize.”

How Does Data Analysis Relate to Other Data and Business Functions?

Data analysis exists as a continuum of techniques, three of the most common being data analytics, data science and data mining.

Data Analysis vs. Data Analytics

Some people use these terms interchangeably. Data analysis also is often considered to be a subset of data analytics. Generally, data analytics covers a forward-looking outlook, or predicting future actions or results.

Data Analysis vs. Data Science

Data science takes analysis a step further by applying techniques from computer science to generate complex models that take into account large numbers of variables with complex (and sometimes poorly understood) interrelationships.

Data Analysis vs. Data Mining

Data mining goes even deeper by automating the process of discovery. Software is developed to find relationships and build models from extremely large datasets. Data mining is extremely powerful, but the resulting models require extensive evaluation to ensure they are valid.

How to Sharpen Your Data Analysis Skills

So you want to learn more about data analysis, but where to start? There is no right answer for everyone. And with such a large topic, don’t expect shortcuts. Here are a few places to get started.

If you never took a statistics class, it’s time to read The Cartoon Guide to Statistics . While it’s no replacement for a semester-long class, it’s more than enough to get you started.

Speaking of classes, there are some very good options for free online. Coursera , Udacity and Khan Academy offer relevant classes for free, although some features may require a paid upgrade. As you get more advanced, you can access a library of great tutorials at KDNuggets .

To get started right now, check out YouTube, where you will find a nearly never-ending collection of videos on data analysis. I highly recommend tuning in to The Ohio State University professor and Nobel Fellow Bear Braumoeller’s online lectures that address data literacy and visualization.

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 09 February 2017

What data and analytics can and do say about effective learning

• Jason M. Lodge   ORCID: orcid.org/0000-0001-6330-6160 1 &
• Linda Corrin 2  

npj Science of Learning volume  2 , Article number:  5 ( 2017 ) Cite this article

8831 Accesses

27 Citations

24 Altmetric

Metrics details

• Communication
• Human behaviour
• Interdisciplinary studies

The collection and analysis of data about learning is a trend that is growing exponentially in all levels of education. Data science is poised to have a substantial influence on the understanding of learning in online and blended learning environments. The mass of data already being collected about student learning provides a source of greater insights into student learning that have not previously been available, and therefore is liable to have a substantial impact on and be impacted by the science of learning in the years ahead. 1

However, despite the potential evident in the application of data science to education, several recent articles, e.g., 2 , 3 have pointed out that student behavioural data collected en masse do not holistically capture student learning. Rogers 4 contends that this positivist view of analytics in education is symptomatic of issues in the social sciences more broadly. While there is undeniable merit in bringing a critical perspective to the use of data and analytics, we suggest that the power and intent of data science for understanding learning is now becoming apparent. The intersection of the science of learning with data and analytics enables more sophisticated ways of making meaning to support student learning.

Learning analytics and the science of learning

The concept of learning analytics emerged less than a decade ago to describe the analysis of student data to inform the improvement of learning and learning environments. 5 Learning analytics involves the integration and analysis of data from multiple sources to inform action. It is a rapidly growing field that has been built upon a foundation informed by not only by data science but also psychology, business analytics, and the science of learning (see also 6 ). Studies of learning analytics have been conducted in areas such as the support of student learning through the provision of automated feedback 7 and curriculum design. 8

The strength of learning analytics as a growing field of research and practice is that it leverages the increasing body of data about student behaviour and engagement generated as more technology is used in teaching and learning. The resulting explosion of data has led to many possibilities such as better monitoring of student progress, identification of students “at risk”, new insights into students’ patterns of behaviour, and real-time intervention in digital environments. The use of these data also raises concerns such as ethical use of data, the quality of models underpinning learning analytics systems, and appropriate interpretation of data. 9 There are several reasons why arguments against the use of these data through learning analytics to understand effective learning are symptomatic of a field still to reach its full potential. We will briefly address these in turn.

Inferring learning from behaviour

Common criticisms of learning analytics suggest that behavioural data alone cannot be used to determine the quality of learning. 3 But what is forgotten in these discussions is that the use of behavioural data to understand student learning is far from a novel approach. Researchers working within the science of learning have been using similar inferential methods for decades, particularly psychological scientists and cognitive neuroscientists. Through carefully designed experimental studies, researchers can make inferences about the learning process on the basis of these data. Learning analytics as a methodology can learn from this experience.

What experimental studies provide are models of how learning works that can then be used as a way of understanding and predicting the learning process in real-life settings. 10 For example, laboratory studies provide suggestions as to the type of behaviours evident when a student gets confused when undertaking a learning activity (e.g., ref. 11 ). When these same markers become evident when students learn in a real online environment, we can say with some confidence that the student could be confused.

Having identified potential confusion, appropriate educational interventions can be made. This could be an automated feedback message within the online learning system, or some form of communication from/with the teacher. Behavioural data can also be used to track students’ approaches to study. For example, the frequency and sequence with which they engage with learning activities can be tracked. 12 While this may not directly measure student learning, it can have a positive influence on the student’s learning environment by helping to identify strategies that could improve how they plan and regulate their study.

With some care about the inferences being made on the basis of data, the science of learning and learning analytics can not only learn from each other but could form a fruitful collaboration. Psychological science in particular can provide options for how best to infer learning from behavioural data. Learning analytics provides new tools for the science of learning to assess these options to understand learning in real-life digital environments.

Through collecting, integrating, and analysing data, learning analytics provides opportunities for further examination of how observations from the laboratory translate to the classroom. Log files and audit trails from online learning systems accumulate behavioural data about students as they learn in digital environments that can then be compared and contrasted with behavioural data collected in experimental settings. Learning analytics, and data science more broadly, can therefore help to bridge education, psychology, and neuroscience through a common focus on behaviour.

Data and design

In the real-life educational environments, creating meaning from data requires making reference to how learning activities are designed. 13 Similar to the way in which the design of an experiment allows for inferences to be made about learning in the lab, learning design allows for inferences to be made about learning in the classroom. The conditions in both cases give meaning to the data. Examining data about student behaviour with reference to a particular learning design helps teachers to see if students engaged with activities in the way they expected. If not, this might mean the design may need to be reviewed and improved. Again this points to the strength of learning analytics to connect the laboratory and the classroom by bridging student behaviour with learning design.

In conversations about big data there is often an assumption that the data and analysis will automatically provide an “answer” to student learning. The illusion that the collection of big data sets will ultimately lead to conclusive facts about learning is a challenge that the field of learning analytics must address. What is clear at this stage in the development of learning analytics is that the teacher remains central to the process of linking analyses with appropriate educational actions. 14 As the designer of the learning activities, the teacher is best placed to be able to determine if student patterns of behaviour match with the pedagogical reasons for why that activity should lead to student learning.

Learning analytics is not just about big data

It is easy to criticise the field of learning analytics as being overly focussed on isolated behavioural markers. It is true that, if this were all that this field represents, there would be limited value in it. However, learning analytics now encompasses a growing range of methods for understanding learning. Its strength is that, when used strategically, it builds on the outcomes of other disciplines, especially research in education and psychology. There is also potential for computational neuroscience to assist in the construction and refining of analytical models that will make better predictions about student learning as they progress. 15 While learning analytics may not provide the ultimate answer to improving learning, there is potential for the field to help bridge some gaps between education, psychology, and neuroscience by providing deeper insight into student behaviour as they learn in real educational settings.

For learning analytics and behavioural data to be useful for understanding student learning, it is important to determine what we want to know, what is already known, and how this relates to design. Only when these factors are determined, we can identify what data are needed. Identifying the right data is crucial to getting learning analytics right and to realise its potential for bridging the gap between the laboratory and the classroom. Some of these data are easy to access, some may not be available. Some data that are available are not useful. Big data is important, but so too is small data about individuals and particular learning tasks.

The science of learning therefore has a critical role to play in informing how learning analytics evolves over time. Laboratory studies help to verify patterns seen in real-life environments by exposing them to controlled conditions. The science of learning can inform the development of learning analytics through the provision of theories and methodologies that will help to move both fields forward. Learning analytics can help to bridge the gap between neuroscience, psychology, and education by providing a way of observing student behaviour as they learn outside the laboratory. The combination of learning analytics and the science of learning therefore has the potential to provide more powerful ways to monitor and support students as they learn.

Siemens, G. & Baker, R. J. S. Learning analytics and educational data mining: towards communication and collaboration. Proceedings of the 2nd International Conference on Learning Analytics and Knowledge (ACM, 2012).

Beer, C., Tickner, R., & Jones, D. Three paths for learning analytics and beyond: moving from rhetoric to reality. Proceedings of the 31st Annual Conference of the Australasian Society for Computers in Learning in Tertiary Education (Macquarie University, 2014).

Lodge, J. M. & Lewis, M. J. In Future Challenges, Sustainable Futures. Proceedings ascilite Wellington 2012 (eds. Brown, M., Hartnett, M., & Stewart, T.) (ascilite, 2012).

Rogers, T. Critical realism and learning analytics research: epistemological implications of an ontological foundation. Proceedings of the Fifth International Conference on Learning Analytics and Knowledge (ACM, 2015).

Siemens, G. & Long, P. Penetrating the fog: analytics in learning and education. EDUCAUSE Review , 46 , 30 (2011).

Gašević, D., Dawson, S. & Siemens, G. Let’s not forget: learning analytics are about learning. TechTrends 59 , 64–71 (2015).

Google Scholar  

Sonnenberg, C. & Bannert, M. Discovering the effects of metacognitive prompts on the sequential structure of srl-processes using process mining techniques. J. Learn. Analytics 2 , 72–100 (2015).

Article   Google Scholar  

Rienties, B. & Toetenel, L. The impact of 151 learning designs on student satisfaction and performance: social learning (analytics) matters. Proceedings of the Sixth International Conference on Learning Analytics & Knowledge (ACM, 2016).

Corrin, L. & de Barba, P. in Rhetoric and Reality: Critical Perspectives on Educational Technology . (eds. Hegarty, B. McDonald, J. & Loke, S-K) (ascilite, 2014).

Kennedy, G. E. & Judd, T. S. Expectations and reality: evaluating patterns of learning behaviour using audit trails. Comp. Edu. 49 , 840–855 (2007).

D’Mello, S., Lehman, B., Pekrun, R. & Graesser, A. Confusion can be beneficial for learning. Learn. Instr. 29 , 153–170 (2014).

Coffrin, C., Corrin, L., De Barba, P., & Kennedy, G. In Proceedings of the 4th International Conference on Learning Analytics and Knowledge (eds. Pistilli, M., Willis, J., Koch, D., Arnold, K., Teasley, S., & Pardo, A.) (ACM, 2014).

Lockyer, L., Heathcote, E. & Dawson, S. Informing pedagogical action: aligning learning analytics with learning design. Am. Beh. Sci. 57 , 1439–1459 (2013).

Bakharia, A. et al . In Proceedings of the 6th International Conference on Learning Analytics and Knowledge (eds. Reiners, T., et al. ) (ACM, 2016).

Anderson, J. R., Pyke, A. A. & Fincham, J. M. Hidden stages of cognition revealed in patterns of brain activation. Psy. Sci. 27 , 1215–1226 (2016).

Download references

Acknowledgements

The Australian Research Council provided funding for the work underpinning this commentary as part of the Special Research Initiative for the Science of Learning Research Centre (Project No. SR120300015).

Competing interests

The authors declare that they have no competing interests.

Author information

Authors and affiliations.

Melbourne Centre for the Study of Higher Education & ARC-SRI Science of Learning Research Centre, University of Melbourne, Melbourne, Australia

Jason M. Lodge

Melbourne Centre for the Study of Higher Education, University of Melbourne, Melbourne, Australia

Linda Corrin

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Jason M. Lodge .

Rights and permissions

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/

Reprints and permissions

About this article

Cite this article.

Lodge, J.M., Corrin, L. What data and analytics can and do say about effective learning. npj Science Learn 2 , 5 (2017). https://doi.org/10.1038/s41539-017-0006-5

Download citation

Received : 06 September 2016

Revised : 27 October 2016

Accepted : 05 December 2016

Published : 09 February 2017

DOI : https://doi.org/10.1038/s41539-017-0006-5

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Predictive video analytics in online courses: a systematic literature review.

• Ozan Raşit Yürüm
• Tuğba Taşkaya-Temizel
• Soner Yıldırım

Technology, Knowledge and Learning (2023)

Research trends in measurement and intervention tools for self-regulated learning for e-learning environments—systematic review (2008–2018)

• Elizaphan Maina
• Robert Oboko

Research and Practice in Technology Enhanced Learning (2020)

Using Technology to Structure and Scaffold Real World Experiential Learning in Distance Education

• Nikki James
• Andrea Humez
• Philipp Laufenberg

TechTrends (2020)

Utilising learning analytics to support study success in higher education: a systematic review

• Dirk Ifenthaler
• Jane Yin-Kim Yau

Educational Technology Research and Development (2020)

Data-driven unsupervised clustering of online learner behaviour

• Robert L. Peach
• Sophia N. Yaliraki
• Mauricio Barahona

npj Science of Learning (2019)

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Educational Data Analytics for Teachers and School Leaders

• Open Access
• © 2023

You have full access to this open access Book

• Sofia Mougiakou 0 ,
• Dimitra Vinatsella 1 ,
• Demetrios Sampson   ORCID: https://orcid.org/0000-0002-0174-0137 2 ,
• Zacharoula Papamitsiou 3 ,
• Michail Giannakos   ORCID: https://orcid.org/0000-0002-8016-6208 4 ,
• Dirk Ifenthaler   ORCID: https://orcid.org/0000-0002-2446-6548 5

University of Piraeus, Piraeus, Greece

You can also search for this author in PubMed   Google Scholar

Department of Digital Systems, University of Piraeus, Piraeus, Greece

Department of technology management, sintef digital, trondheim, norway.

Dept of Computer Science, Norwegian Univ of Science & Technology, Trondheim, Norway

Learning, Design and Technology, University of Mannheim, Mannheim, Germany

• Fully explains the complex ethics and privacy issues inherent in using educational data analytics
• Provides guidance for teaching practical data analytics to educators and administrators
• Discusses the use of educational data analytics for both online and blended learning

Part of the book series: Advances in Analytics for Learning and Teaching (AALT)

14k Accesses

9 Citations

3 Altmetric

Buy print copy

• Compact, lightweight edition
• Dispatched in 3 to 5 business days
• Free shipping worldwide - see info
• Durable hardcover edition

Tax calculation will be finalised at checkout

Table of contents (4 chapters)

Front matter, online and blended teaching and learning supported by educational data.

• Sofia Mougiakou, Dimitra Vinatsella, Demetrios Sampson, Zacharoula Papamitsiou, Michail Giannakos, Dirk Ifenthaler

Adding Value and Ethical Principles to Educational Data

• Learning Analytics
• Teaching Analytics

Back Matter

• Educational Data Analytics
• Instructional design
• Evidence based education
• Teachers' competence

About this book

Educational Data Analytics (EDA) have been attributed with significant benefits for enhancing on-demand personalized educational support of individual learners as well as reflective course (re)design for achieving more authentic teaching, learning and assessment experiences integrated into real work-oriented tasks.

This open access textbook is a tutorial for developing, practicing and self-assessing core competences on educational data analytics for digital teaching and learning. It combines theoretical knowledge on core issues related to collecting, analyzing, interpreting and using educational data, including ethics and privacy concerns. The textbook provides questions and teaching materials/ learning activities as quiz tests of multiple types of questions, added after each section, related to the topic studied or the video(s) referenced. These activities reproduce real-life contexts by using a suitable use case scenario (storytelling), encouraging learners to link theory with practice; self-assessed assignments enabling learners to apply their attained knowledge and acquired competences on EDL. 

By studying this book, you will know where to locate useful educational data in different sources and understand their limitations; know the basics for managing educational data to make them useful; understand relevant methods; and be able to use relevant tools; know the basics for organising, analysing, interpreting and presenting learner-generated data within their learning context, understand relevant learning analytics methods and be able to use relevant learning analytics tools; know the basics for analysing and interpreting educational data to facilitate educational decision making, including course and curricula design, understand relevant teaching analytics methods and be able to use relevant teaching analytics tools; understand issues related with educational data ethics and privacy.

This book is intended for school leadersand teachers engaged in blended (using the flipped classroom model) and online (during COVID-19 crisis and beyond) teaching and learning; e-learning professionals (such as, instructional designers and e-tutors) of online and blended courses; instructional technologists; researchers as well as undergraduate and postgraduate university students studying education, educational technology and relevant fields.

Authors and Affiliations

Sofia Mougiakou, Dimitra Vinatsella

Demetrios Sampson

Zacharoula Papamitsiou

Dept of Computer Science, Norwegian Univ of Science & Technology, Trondheim, Norway

Michail Giannakos

Dirk Ifenthaler

About the authors

Dimitra Vinatsella holds a B.Sc. in “Informatics & Telecommunications” (2003) and a M.Sc. in ”Communication Systems and Networks” (2007) from National and Kapodistrian University of Athens. She started to work as a Product Development Manager in Vodafone Greece responsible to act as an overall project manager for the planning, development, roll out and post launch monitoring of retail

commercial Value Added Services and Products. In 2007, she joined the Greek Ministry of Education, Research and Religious Affairs, as Computer Science Teacher and since 2012 she has been working in a cross-functional team in the Center of Informatics and New Technologies of the Directorate of Secondary Education of Piraeus. She has a long experience in e-Learning utilizing innovative technologies and learning management systems and she has participated successfully in several scientific research European programmes, including the Learn2Analyze project. Currently, she is a Ph.D. Candidate at the Department of Digital Systems, University of Piraeus, Greece.

Bibliographic Information

Book Title : Educational Data Analytics for Teachers and School Leaders

Authors : Sofia Mougiakou, Dimitra Vinatsella, Demetrios Sampson, Zacharoula Papamitsiou, Michail Giannakos, Dirk Ifenthaler

Series Title : Advances in Analytics for Learning and Teaching

DOI : https://doi.org/10.1007/978-3-031-15266-5

Publisher : Springer Cham

eBook Packages : Education , Education (R0)

Copyright Information : The Editor(s) (if applicable) and The Author(s) 2023

Hardcover ISBN : 978-3-031-15265-8 Published: 29 October 2022

Softcover ISBN : 978-3-031-15268-9 Published: 29 October 2022

eBook ISBN : 978-3-031-15266-5 Published: 28 October 2022

Series ISSN : 2662-2122

Series E-ISSN : 2662-2130

Edition Number : 1

Number of Pages : XII, 238

Number of Illustrations : 12 b/w illustrations, 82 illustrations in colour

Topics : Education, general , Statistics, general

• Publish with us

Policies and ethics

• Find a journal
• Track your research

Quick Links

• Leadership Donors

Using Analytics to Find Value in Education Data

• Data, Privacy, Security, and Regulation
• Future of Work

The Learning Network

The following insights are derived from a recent Assembly talk featuring Hywel Benbow , regarding applying data analytics to enhance the performance of educational institutions. We presented a few question to Hywel during this talk, and what follows is a summary of his responses.

Could you speak to some overarching and generalizable ways educational organizations could leverage data to drive performance, growth, and innovation?  

The transformative power of analytics is undeniable, it plays a pivotal role in revolutionizing education on multiple fronts. Firstly, it streamlines administrative tasks from admissions to astute resource allocation, boosting organizational efficiency. It also brings a new dimension to staff recruitment and retention strategies by tailoring hiring processes to specific needs such as hiring teachers adept at teaching multiple subjects during periods of growth and expansion, optimizing the recruitment process and supporting strategies for future campaigns.

On the educational side, data analytics transforms the learning experience. Analytics facilities personalized learning by recognizing individual students’ needs, preferences, abilities, and learning pace. Adaptive learning systems further fine-tune the educational journey. These platforms adjust in real time ensuring the learning journey of every student is optimized. Beyond real time adjustments, analytics can proactively identify students who might be at risk, facilitating timely interventions, and predicting future student performance, ensuring educators remain prepared. 

The impact doesn’t stop with analytics alone. Emerging technologies like AI, machine learning, augmented and virtual reality are seen as innovative and immersive tools, enhancing the learning experience and removing some of the barriers, particularly for students with physical or geographical limitations. Collectively, analytics and emerging technologies are reshaping education, making it more efficient, personalized, and responsive to the needs of students and institutions. 

Can you share your thoughts on how education institutions can build the capabilities to be data and insight-driven?  

Education institutions are poised at an interesting juncture where data-informed insights can significantly enhance both operational efficiency and the quality of education. To transition into being data and insight-driven, several integral steps need to be taken.

Firstly, there needs to be strong leadership support to establish this culture—leadership teams must embrace and champion the use of data to inform decisions and enhance operational efficiency, essentially setting the tone for the entire institution. Creating general awareness among faculty, leadership, and board members is also crucial. An informed and enthusiastic faculty can be the linchpin in effectively transitioning to a data-informed approach. 

Awareness alone isn’t enough, promoting and embedding data fluency and literacy is another pivotal element. It’s not about transforming educators into analytics but ensuring they are equipped with the wider skillset to use data effectively. Collaboration between data analysts and domain experts is encouraged to elevate the quality of questions asked and the analytical outcomes.

Infrastructure is the backbone of any data initiative. As such, ensuring that the organization’s data and digital ecosystem is aligned with the organizations overarching strategy becomes indispensable.

Finally, there is deep responsibility in handling data. Data governance, which includes maintaining data quality, proper data collection, and data protection, is underscored as a foundational element for effective data-informed initiatives. Without robust governance, even the most insightful data can become a liability.

Could you talk about a few examples in which you used data analytics to help stakeholders in educational organizations make informed decisions, focusing on decisions that are commercial or business-facing?   

While data analytics is a necessity to succeed commercially, two key areas where data has a distinct impact across K-12 schools, universities, and colleges are admissions and customer experience.

Data analytics plays a crucial role in dissecting the admissions processes, pinpointing where potential students fall out of the admissions funnel through to gaining a deeper understanding of the end-to-end journey students and parents undergo, from enrolling to graduating or leaving. 

This is closely tied to the concept of customer experience in education, especially in K-12 schools, where students and parents constitute two distinct customer groups. Recognizing the needs and desires of both groups is pivotal, the essence of education is inherently emotional thereby magnifying the importance of catering to the aspirations and concerns of both these groups. Happy students and satisfied parents are critical for student retention. Happy parents also become promoters, providing brand loyalty and positive comments leading to increased admissions and revenue, which is paramount from a commercial perspective.

But there is another layer to this relationship. Parental and student feedback, often rich with insights about strengths and challenges is incredibly valuable. When channeled constructively this feedback can identity issues, themes, trends and help transform performance or highlight where there is a perception challenge. On a positive front, feedback can be turned into compelling narratives, which when used for marketing and communications, can attract prospective new students.

How have you used data to build capacity and improve performance of individuals and teams in the educational organizations you led? How can data analytics help educators optimize their impact in the classroom, help students improve performance and learning outcomes, and help administrators streamline and optimize services?  

Data analytics excels at improving efficiency and productivity, an example being automating repetitive tasks, freeing professionals to focus on more impactful endeavors. This liberation inadvertently paves the way for rich, spontaneous collaborations. When analytics experts liaise with professionals from diverse departments, the combined outcomes can be quite profound.

For educators, data analytics isn’t just a tool, it offers valuable support, enabling data-informed decision-making. It guides educators with insights, these alongside their contextual knowledge enhance teaching and student performance. The real power of data analytics comes when different datasets are utilized together, creating comprehensive and timely views of students encompassing a fuller perspective on their needs. 

Technology plays a large part but should be supportive of the wider goals rather than the driver. Innovations like Chat GPT exemplify how technology can be a catalyst for creativity and exploration and helping to streamline tasks. The integration of technology in education equips students and teachers with assistive tools for enhancing the way they work, this, when used well improves productivity and outcomes. With the job market continually evolving towards tech-centric roles, it’s imperative for students to be fluent in the developing and using technology. Being able to incorporate different tools, thought their own learning and through seeing how their school utilize them is a positive development.

What are some risks to be cognizant of when applying data analytics in the context of educational organizations? What precautions should be taken to make sure sensitive student data is kept secure?  

Firstly, data security and privacy take center stage. Education settings deal with a considerable amount of sensitive data, therefore safeguarding these datasets is paramount. This involves secure storage and tailored access rights for individuals like administrators, teachers, and board members. Regular reviews and adjustments are also essential to ensure the right access is in place as roles are and projects evolve. 

Misinterpreting data is a significant concern, particularly in unfamiliar contexts. Without understanding the context, erroneous conclusions can be drawn, leading to ill-advised decisions. Collaborative efforts between education experts and data analysts, involving open communication, relentless questioning, and mutual guidance, effectively mitigate this risk.

Upholding ethical standards is also key to preserve trust within the educational community. It’s essential that collected data serves its intended purpose, transparency with students, parents, and staff on how data is used is also essential. When integrating external datasets with internal ones, latent biases may seep in, skewing analyses and outcomes. Such biases, as evidenced in certain outputs of machine learning tools including ChatGPT, can lead to flawed conclusions and actions. Balancing data security, interpretation, and ethical considerations remain central in the educational landscape. 

How should educational organizations deal with potential biases in educational data, that impact decision making, especially in relation to students, assessment, and performance evaluation.  

There is an inherent responsibility to ensure that biases embedded within educational data don’t compromise the decision-making process, especially concerning assessments and performance evaluations. It’s important to have a clear understanding of the data’s origin and composition, especially when using data in the context of predictive modeling for student performance and education analytics, as doing so can help recognize and address potential biases.

Recognizing the data’s nuances and inherent biases is the first step to effectively counteract them. And this vigilance amplifies when transitioning data insights across varied cultural and demographic contexts. For instance, utilizing data derived from Western educational settings within a predictive model for Middle Eastern ones, teeming with a plethora of nationalities and diverse educational backgrounds, can lead to glaring inaccuracies. Such transitions don’t just risk misinterpretation; they can inadvertently entrench biases.

Many large data sets come from Western countries with mature data collection systems, making them more suitable for use in similar contexts. While these can be a valuable source in insights for policy development in the same geographies, they may not resonate with the same authenticity in countries with an emerging education landscape.

By critically evaluating data sources, being sensitive to contextual nuances, and continuously checking for biases, the education sector can ensure that data analytics supports improved outcomes.

What is the current state of the application of data analytics to augment human and business performances in the educational landscape, in the context of the Middle East (UAE)? How does this state compare to how data analytics is being used in the educational landscape in the US and the UK?   

Have you faced any resistance or hesitancy to implementing data analytics? And how did you possibly deal with that?  

In the Middle East, data analytics is a rapidly emerging field that has witnessed significant growth in the past few years. This development is a response to the region’s strong emphasis on personalized learning, driven by the diverse cultural and linguistic backgrounds of students and the increasing demand from parents for customized education to enhance student outcomes. As a result, advanced analytics tools are gaining traction in education, not only for personalized learning but also for enhancing efficiency through the analysis of student and staff performance. The integration of AI technology, such as Chat GPT, is also growing in Middle Eastern educational settings, much the same as the rest of the world. 

Comparing the Middle East to more mature regions like Europe and the U.S., the Middle East is rapidly narrowing the gap, with its best schools performing at the same level as their peers in established regions. However, the Middle East is still in the process of building its data infrastructure, while mature regions have well-established systems in place helping to shape policy and strategy. In the coming years as more historical data is available the Middle East will be on a similar footing.

In industries where data analytics is less mature or has been less of an everyday feature, resistance and apprehension are common. Overcoming this resistance entails shifting the focus from uncovering problems to collaboratively identifying and solving them. An internal marketing approach can showcase the benefits of data analytics, emphasizing its goal to enhance efficiency and address specific concerns. Collaboration and a problem-solving orientation are pivotal in demonstrating the value of data analytics. It’s very much a “how can data analytics help” approach that removes resistance.

Looking forward, the Middle East is uniquely positioned. Unhindered by some of the legacy constraints that occasionally tether older regions, it possesses the agility to leapfrog into the future. As the regions continues its upward trajectory, emerging countries may increasingly look to the Middle East for insights into adapting and innovating educational practices, establishing the region as a hub for educational innovation. 

What do you see the future of data analytics in the educational landscape to be in the aforementioned regions? In what specific ways do you foresee data analytics disrupting the educational landscape, for good and for bad, particularly in relation to the advent of Artificial Intelligence and Large Language Models such as ChatGPT?  

The future of education analytics holds great promise, driven by increasingly accessible and affordable technologies. The Middle East, in particular, is poised for substantial advancements over the next 5 to 10 years. This progress is attributed to the accumulation of more and better data and, substantial government involvement in elevating educational standards, facilitated by the region’s openness to change, unlike more established systems in the US and the UK. An unexpected proponent of this evolution was the COVID-19 pandemic. It laid bare the essence of robust infrastructure capable of underpinning remote and digital education. The Middle East, and in particular schools in the UAE demonstrated greater preparedness in this regard.

Predictive analytics is expected to play an expanded role in education, particularly in assessing students’ future performance. As schools amass more data, it will allow predictive analytics to be increasingly leveraged to anticipate student outcomes and enhance personalized learning. 

While implementing such technology can be a challenge for state-level schools or those with budget constraints, the decreasing cost of technology over time is expected to increase accessibility. Assistive technologies like Chat GPT and Google Bard are likely to become more common in education, complementing traditional teaching and learning tools. It’s pertinent to remember these are assistive technologies and are there to support elements of teaching rather than replace it. As technology continues to evolve, it is poised to have a substantial impact on education in the years ahead, from teaching, learning and administrative perspectives.

What are a few important challenges the education sector has to grapple with related to data infrastructure and governance?  

In terms of governance and infrastructure, building a culture where the use of data within an organization becomes common place is vital. Governance, often seen as a strict term, should be associated with security, protection, and ethics. Elevating data literacy and fluency levels across the organization can help mitigate governance challenges, as it supports everyone in understanding the importance of data protection, data quality and data ethics. 

Another daunting hurdle is the fragmentation of data, addressing the challenge of data silos is a challenge. Organizations need to move away from having data scattered across different unconnected systems and toward building a cohesive ecosystem that seamlessly manages data. The real power of data is unlocked when users are able to utilize different datasets together to uncover deeper insights. Ensuring the digital and data ecosystem is aligned allows this to happen.

The availability of cloud systems has become more cost-effective, encouraging a shift from legacy practices to more cohesive systems. Scalability is also a key challenge, as the volume of data collected in the education sector is expected to increase significantly. The infrastructure required by to succeed must be both expansive and agile, able to evolve with surges and respond to unforeseen shifts, much like the pandemic-induced digital pivot.

The mounting volume also signals an impending demand for adept data analysts. This surge, however, is not exclusive to the education sector, culminating in a fierce battle for talent acquisition. Efforts to integrate digital analytics into curriculums, such as the UK’s digital curriculum, aim to address this challenge and ensure a future workforce with ingrained data analysis skills. 

The bedrock of any successful endeavor is a robust foundation, and this axiom holds true, unequivocally, when it comes to data and analytics.

As a closing remark, do you have any advice for students and parents as to why they should care about data analytics?  

Understanding the whole picture of student learning outcomes is essential, given that various factors can influence different student experiences. For instance, a high-ability student whose potential might remain unfulfilled, due to unrecognized capabilities or challenges in their wellbeing. Harnessing diverse data, can equip educators with a better perspective of the student, allowing them to understand which areas need additional focus or attention. Collecting the right data and using it in the right ways minimized the number of students who may miss reaching their potential. 

From a parent’s perspective, having teachers understand a child’s needs, strengths, weaknesses, preferred pace of learning, and interaction style is crucial for effective education. While every parent values the safety and happiness of their children, they are also discerning about their child’s preparedness for subsequent academic phases, regardless of age, curriculum or geographical nuances. Integrating data analytics proves invaluable in catering to these aspirations. The goal of personalized learning is to ensure that each learner reaches their full potential. This means recognizing that not every student excels in the same way or at the same pace. Data analytics can play a key role in helping educators, counselors, and teachers set students up for success and guide them through their educational journey.

Engage With Us

Join Our Community

Ready to dive deeper with the Digital Data Design Institute at Harvard? Subscribe to our newsletter, contribute to the conversation and begin to invent the future for yourself, your business and society as a whole.

How higher-education institutions can transform themselves using advanced analytics

Leaders in most higher-education institutions generally understand that using advanced analytics can significantly transform the way they work by enabling new ways to engage current and prospective students, increase student enrollment, improve student retention and completion rates , and even boost faculty productivity and research. However, many leaders of colleges and universities remain unsure of how to incorporate analytics into their operations and achieve intended outcomes and improvements. What really works? Is it a commitment to new talent, technologies, or operating models? Or all of the above?

To answer these questions, we interviewed more than a dozen senior leaders at colleges and universities known for their transformations through analytics. We also conducted in-depth, on-campus visits at the University of Maryland University College (UMUC), a public institution serving primarily working adults through distance learning, and Northeastern University, a private nonprofit institution in Boston, to understand how their transformations went. 1 Our research base included presidents, vice presidents of enrollment management, chief data officers, provosts, and chief financial officers. In September 2017, we conducted on-campus visits to meet with leaders at several levels at both the University of Maryland University College (UMUC) and Northeastern University. We thank these leaders for generously agreeing to have their observations and experiences included in this article. We combined insights from these interviews and site visits with those gleaned from our work with more than 100 higher-education engagements across North America over the past five years, and we tapped McKinsey’s wide-ranging expertise in analytics-enabled transformations in both the public and private sectors.

Our conversations and engagements revealed several potential pitfalls that organizations may face when building their analytics capabilities—as well as several practical steps education leaders can take to avoid these traps.

Understanding the challenges

Advanced analytics use cases.

Northeastern used advanced analytics to help grow its U.S. News & World Report ranking among national universities from 115 in 2006 to 40 in 2017.

UMUC used advanced analytics to achieve a 20 percent increase in new student enrollment while spending 20 percent less on marketing.

Transformation through advanced analytics can be difficult for any organization; in higher education, the challenges are compounded by sector-specific factors related to governance and talent. Leaders in higher education cannot simply pay lip service to the power of analytics; they must first address some or all of the most common obstacles.

Being overly focused on external compliance . Many higher-education institutions’ data analytics teams focus most of their efforts on generating reports to satisfy operational, regulatory, or statutory compliance. The primary goal of these teams is to churn out university statistics that accrediting bodies and other third parties can use to assess each institution’s performance. Any requests outside the bounds of these activities are considered emergencies rather than standard, necessary assignments. Analytics teams in this scenario have very limited time to support strategic, data-driven decision making.

Isolating the analytics program in an existing department . In our experience, analytics teams in higher-education institutions usually report to the head of an existing function or department—typically the institutional research team or the enrollment-management group. As a result, the analytics function becomes associated with the agenda of that department rather than a central resource for all, with little to no contact with executive leadership. Under this common scenario, the impact of analytics remains limited, and analytics insights are not embedded into day-to-day decision making of the institution as a whole.

Failing to establish a culture of data sharing and hygiene . In many higher-education institutions, there is little incentive (and much reluctance) to share data. As a result, most higher-education institutions lack good data hygiene —that is, established rules for who can access various forms of data, as well as formal policies for how they can share those data across departments. For example, analytics groups in various university functions may use their own data sets to determine retention rates for different student segments—and when they get together, they often disagree on which set of numbers is right.

Compounding this challenge, many higher-education institutions struggle to link the myriad legacy data systems teams use in different functions or working groups. Even with the help of a software platform vendor, the lead time to install, train, and win buy-in for these technical changes can take time, perhaps two to three years, before institutions see tangible outcomes from their analytics programs. In the meantime, institutions struggle to instill a culture and processes built around the possibilities of data-driven decision making.

Lacking the appropriate talent. Budgets and other constraints can make it difficult for higher-education institutions to meet market rates for analytics talent. Colleges and universities could potentially benefit from sourcing analytics talent among their graduate students and faculty, but it can be a struggle to attract and retain them. Furthermore, to successfully pursue transformation through analytics, higher-education institutions need leaders who are fluent in not only management but also data analytics and can solve problems in both areas.

Would you like to learn more about the Social Sector Practice ?

Deploying best practices.

These challenges can seem overwhelming, but transformation through analytics is possible when senior leaders in higher-education institutions endeavour to change both operations and mind-sets.

Leaders point to five action steps to foster success:

Articulate an analytics mandate that goes beyond compliance . Senior leaders in higher education must signal that analytics is a strategic priority. Indeed, to realize the potential of analytics, the function cannot be considered solely as a cost center for compliance. Instead, this team must be seen as a source of innovation and an economic engine for the institution. As such, leaders must articulate the team’s broader mandate. According to the leaders we interviewed, the transformation narrative must focus on how analytics can help the institution facilitate the student journey from applicant to alumnus while providing unparalleled learning, research, and teaching opportunities, as well as foster a strong, financially sustainable institution.

Establish a central analytics team with direct reporting lines to executive leaders . To mitigate the downsides of analytics teams couched in existing departments or decentralized across several functions, higher-education leaders must explicitly allocate the requisite financial and human resources to establish a central department or function to oversee and manage the use of analytics across the institution. This team can be charged with managing a central, integrated platform for collecting, analyzing, and modeling data sets and producing insights quickly.

For example, UMUC has a designated “data czar” to help define standards for how information is captured, managed, shared, and stored online. When conflicts arise, the data czar weighs in and helps de-escalate problems. Having a central point of contact has improved the consistency and quality of the university’s data: there is now a central source of truth, and all analysts have access to the data. Most important, the university now has a data evangelist who can help cultivate an insights-driven culture at the institution.

In another example, leaders at Northeastern created an analytics center of excellence structured as a “virtual” entity. The center is its own entity and is governed by a series of rotating chairs to ensure the analytics team is aware of and paying equal attention to priorities from across the university.

In addition to enjoying autonomous status outside a subfunction or single department, the analytics team should report to the most-senior leaders in the institution—in some cases, the provost. When given a more substantial opportunity to influence decisions, analytics leaders gain a greater understanding of the issues facing the university and how they affect the institution’s overall strategy. Leaders can more easily identify the data sets that might provide relevant insights to university officials—not just in one area, but across the entire organization—and they can get a jump-start on identifying possible solutions.

Analysts at Northeastern, for instance, were able to quantify the impact of service-learning programs on student retention, graduation, and other factors, thereby providing support for key decisions about these programs.

Win analytics buy-in from the front line and create a culture of data-driven decision making . To overcome the cultural resistance to data sharing, the analytics team must take the lead on engendering meaningful communications about analytics across the institution. To this end, it helps to have members of the centralized analytics function interact formally and frequently with different departments across the university. A hub-and-spoke model can be particularly effective: analysts sit alongside staffers in the operating units to facilitate sharing and directly aid their decision making. These analysts can serve as translators, helping working groups understand how to apply analytics to tackle specific problems, while also taking advantage of data sets provided by other departments. The university leaders we spoke with noted that their analysts may rotate into different functional areas to learn more about the university’s departments and to ensure that the department leaders have a link back to the analytics function.

How to improve student educational outcomes: New insights from data analytics

Of course, having standardized, unified systems for processing all university data can help enable robust analysis. However, universities seeking to create a culture of data-driven decision making need not wait two years until a new data platform is up and running. Instead, analysts can define use cases—that is, places where data already exist and where analysis can be conducted relatively quickly to yield meaningful insights. Teams can then share success stories and evangelize the impact of shared data analytics, thereby prompting others to take up their own analytics-driven initiatives.

The analysts from UMUC’s decision-support unit sometimes push relevant data and analyses to the relevant departments to kick-start reflection and action, rather than waiting for the departments to request the information. However, the central unit avoids producing canned reports; analysts tend to be successful only when they engage departments in an honest and objective exploration of the data without preexisting biases.

Strengthen in-house analytical capabilities . The skills gap is an obvious impediment to colleges’ and universities’ attempts to transform operations through advanced analytics—thus, it is perfectly acceptable to contract out work in the short term. However, while supplementing a skills gap with external expertise may help accelerate transformations, it can never fully replace the need for in-house capacity; the effort to push change across the institution must be owned and led internally.

To do so, institutions will need to change their approaches to talent acquisition and development . They may need to look outside usual sources to find professionals who understand core analytics technologies (cloud computing, data science, machine learning, and statistics, for instance) as well as design thinking and operations. Institutions may also need to appeal to new hires with competitive financial compensation and by emphasizing the opportunity to work autonomously on intellectually challenging projects that will make an impact on generations of students and contribute to an overarching mission.

Do not let great be the enemy of good . It takes time to launch a successful analytics program. At the outset, institutions may lack certain types of data, and not every assessment will yield insightful results—but that is no reason to pull back on experimentation. Colleges and universities can instead deploy a test-and-learn approach: identify areas with clear problems and good data, conduct analyses, launch necessary changes, collect feedback, and iterate as needed. These cases can help demonstrate the impact of analytics to other parts of the organization and generate greater interest and buy-in.

Subscribe to the Shortlist

McKinsey’s new weekly newsletter, featuring must-read content on a range of topics, every Friday

Realizing impact from analytics

It is easy to forget that analytics is a beginning, not an end . Analytics is a critical enabler to help colleges and universities solve tough problems—but leaders in higher-education institutions must devote just as much energy to acting on the insights from the data as they do on enabling analysis of the data. Implementation requires significant changes in culture, policy, and processes. When outcomes improve because a university successfully implemented change—even in a limited environment—the rest of the institution takes notice. This can strengthen the institutional will to push further and start tackling other areas of the organization that need improvement.

Some higher-education institutions have already overcome these implementation challenges and are realizing significant impact from their use of analytics. Northeastern University, for example, is using a predictive model to determine which applicants are most likely to be the best fit for the school if admitted. Its analytics team relies on a range of data to make forecasts, including students’ high school backgrounds, previous postsecondary enrollments, campus visit activity, and email response rates. According to the analytics team, an examination of the open rate for emails was particularly insightful as it was more predictive of whether students actually enrolled at Northeastern than what the students said or whether they visited campus.

Meanwhile, the university also looked at National Student Clearinghouse data, which tracks where applicants land at the end of the enrollment process, and learned that the institutions it had considered core competitors were not. Instead, competition was coming from sources it had not even considered. It also learned that half of its enrollees were coming from schools that the institution’s admissions office did not visit. The team’s overall analysis prompted Northeastern to introduce a number of changes to appeal to those individuals most likely to enroll once admitted, including offering combined majors. The leadership team also shifted some spending from little-used programs to bolster programs and features that were more likely to attract targeted students. Due in part to these changes, Northeastern improved its U.S. News & World Report ranking among national universities from 115 in 2006 to 40 in 2017.

In another example, in 2013 UMUC was trying to pinpoint the source of a decline in enrollment. It was investing significant dollars in advertising and was generating a healthy number of leads—however, conversion rates were low. Data analysts at the institution assessed the university’s returns on investment for various marketing efforts and discovered a bottleneck—UMUC’s call centers were overused and underresourced. The university invested in new call-center capabilities and within a year realized a 20 percent increase in new student enrollment while spending 20 percent less on advertising.

The benefits we discussed barely scratch the surface; the next wave of advanced analytics will, among other things, enable bespoke, personalized student experiences, with teaching catered to students’ individual learning styles and competency levels. To realize the great promise of analytics in the years to come, senior leaders must focus on more than just making incremental improvements in business processes or transactions. Our conversations with leaders in higher education point to the need for colleges and universities to establish a strong analytics function as well as a culture of data-driven decision making and a focus on delivering measurable outcomes. In doing so, institutions can create significant value for students—and sustainable operations for themselves.

Marc Krawitz is an associate partner in McKinsey’s New Jersey office. Jonathan Law is a partner in the New York office and leads the Higher-Education Practice. Sacha Litman is an associate partner in the Washington, DC, office and leads public and social sector analytics.

The authors would like to thank business and technology leaders at the University of Maryland University College and Northeastern University for their contributions to this article.

Explore a career with us

Related articles.

Shaking up the leadership model in higher education

Three more reasons why US education is ready for investment

Boosting productivity in US higher education

• Review Article
• Open access
• Published: 22 June 2020

Teaching analytics, value and tools for teacher data literacy: a systematic and tripartite approach

• Ifeanyi Glory Ndukwe 1 &
• Ben Kei Daniel 1  

International Journal of Educational Technology in Higher Education volume  17 , Article number:  22 ( 2020 ) Cite this article

27k Accesses

40 Citations

14 Altmetric

Metrics details

Teaching Analytics (TA) is a new theoretical approach, which combines teaching expertise, visual analytics and design-based research to support teacher’s diagnostic pedagogical ability to use data and evidence to improve the quality of teaching. TA is now gaining prominence because it offers enormous opportunities to the teachers. It also identifies optimal ways in which teaching performance can be enhanced. Further, TA provides a platform for teachers to use data to reflect on teaching outcome. The outcome of TA can be used to engage teachers in a meaningful dialogue to improve the quality of teaching. Arguably, teachers need to develop their teacher data literacy and data inquiry skills to learn about teaching challenges. These skills are dependent on understanding the connection between TA, LA and Learning Design (LD). Additionally, they need to understand how choices in particular pedagogues and the LD can enhance their teaching experience. In other words, teachers need to equip themselves with the knowledge necessary to understand the complexity of teaching and the learning environment. Providing teachers access to analytics associated with their teaching practice and learning outcome can improve the quality of teaching practice. This research aims to explore current TA related discussions in the literature, to provide a generic conception of the meaning and value of TA. The review was intended to inform the establishment of a framework describing the various aspects of TA and to develop a model that can enable us to gain more insights into how TA can help teachers improve teaching practices and learning outcome. The Tripartite model was adopted to carry out a comprehensive, systematic and critical analysis of the literature of TA. To understand the current state-of-the-art relating to TA, and the implications to the future, we reviewed published articles from the year 2012 to 2019. The results of this review have led to the development of a conceptual framework for TA and established the boundaries between TA and LA. From the analysis the literature, we proposed a Teaching Outcome Model (TOM) as a theoretical lens to guide teachers and researchers to engage with data relating to teaching activities, to improve the quality of teaching.

Introduction

Educational institutions today are operating in an information era, where machines automatically generate data rather than manually; hence, the emergence of big data in education ( Daniel 2015 ). The phenomenon of analytics seeks to acquire insightful information from data that ordinarily would not be visible by the ordinary eyes, except with the application of state-of-the-art models and methods to reveal hidden patterns and relationships in data. Analytics plays a vital role in reforming the educational sector to catch up with the fast pace at which data is generated, and the extent to which such data can be used to transform our institutions effectively. For example, with the extensive use of online and blended learning platforms, the application of analytics will enable educators at all levels to gain new insights into how people learn and how teachers can teach better. However, the current discourses on the use of analytics in Higher Education (HE) are focused on the enormous opportunities analytics offer to various stakeholders; including learners, teachers, researchers and administrators.

In the last decade, extensive literature has proposed two weaves of analytics to support learning and improve educational outcomes, operations and processes. The first form of Business Intelligence introduced in the educational industry is Academic Analytics (AA). AA describes data collected on the performance of academic programmes to inform policy. Then, Learning Analytics (LA), emerged as the second weave of analytics, and it is one of the fastest-growing areas of research within the broader use of analytics in the context of education. LA is defined as the "measurement, collection, analysis and reporting of data about the learner and their learning contexts for understanding and optimising learning and the environments in which it occurs" ( Elias 2011 ). LA was introduced to attend to teaching performance and learning outcome ( Anderson 2003 ; Macfadyen and Dawson 2012 ). Typical research areas in LA, include student retention, predicting students at-risk, personalised learning which in turn are highly student-driven ( Beer et al. 2009 ; Leitner et al. 2017 ; Pascual-Miguel et al. 2011 ; Ramos and Yudko 2008 ). For instance, Griffiths ( Griffiths 2017 ), employed LA to monitor students’ engagements and behavioural patterns on a computer-supported collaborative learning environment to predict at-risk students. Similarly, Rienties et al. ( Rienties et al. 2016 ) looked at LA approaches in their capacity to enhance the learner’s retention, engagement and satisfaction. However, in the last decade, LA research has focused mostly on the learner and data collections, based on digital data traces from Learning Management Systems (LMS) ( Ferguson 2012 ), not the physical classroom.

Teaching Analytics (TA) is a new theoretical approach that combines teaching expertise, visual analytics and design-based research, to support the teacher with diagnostic and analytic pedagogical ability to improve the quality of teaching. Though it is a new phenomenon, TA is now gaining prominence because it offers enormous opportunities to the teachers.

Research on TA pays special attention to teacher professional practice, offering data literacy and visual analytics tools and methods ( Sergis et al. 2017 ). Hence, TA is the collection and use of data related to teaching and learning activities and environments to inform teaching practice and to attain specific learning outcomes. Some authors have combined the LA, and TA approaches into Teaching and Learning Analytics (TLA) ( Sergis and Sampson 2017 ; Sergis and Sampson 2016 ). All these demonstrate the rising interest in collecting evidence from educational settings for awareness, reflection, or decision making, among other purposes. However, the most frequent data that have been collected and analysed about TA focus on the students (e.g., different discussion and learning activities and some sensor data such as eye-tracking, position or physical actions) ( Sergis and Sampson 2017 ), rather than monitoring teacher activities. Providing teachers access to analytics of their teaching, and how they can effectively use such analytics to improve their teaching process is a critical endeavour. Also, other human-mediated data gathering in the form of student feedback, self and peer observations or teacher diaries can be employed to enrich TA further. For instance, visual representations such as dashboards can be used to present teaching data to help teachers reflect and make appropriate decisions to inform the quality of teaching. In other words, TA can be regarded as a reconceptualisation of LA for teachers to improve teaching performance and learning outcome. The concept of TA is central to the growing data-rich technology-enhanced learning and teaching environment ( Flavin 2017 ; Saye and Brush 2007 ). Further, it provides teachers with the opportunity to engage in data-informed pedagogical improvement.

While LA is undeniably an essential area of research in educational technology and the learning sciences, automatically extracted data from an educational platform mainly provide an overview of student activities, and participation. Nevertheless, it hardly indicates the role of the teacher in these activities, or may not otherwise be relevant to teachers’ individual needs (for Teaching Professional Development (TPD) or improvement of their classroom practice). Many teachers generally lack adequate data literacy skills ( Sun et al. 2016 ). Teacher data literacy skill and teacher inquiry skill using data are the foundational concepts underpinning TA ( Kaser and Halbert 2014 ). The development of these two skills is dependent on understanding the connection between TA, LA and Learning Design (LD). In other words, teachers need to equip themselves with knowledge through interaction with sophisticated data structures and analytics. Hence, TA is critical to improving teachers’ low efficacy towards educational data.

Additionally, technology has expanded the horizon of analytics to various forms of educational settings. As such, the educational research landscape needs efficient tools for collecting data and analyzing data, which in turn requires explicit guidance on how to use the findings to inform teaching and learning ( McKenney and Mor 2015 ). Increasing the possibilities for teachers to engage with data to assess what works for the students and courses they teach is instrumental to quality ( Van Harmelen and Workman 2012 ). TA provides optimal ways of performing the analysis of data obtained from teaching activities and the environment in which instruction occurs. Hence, more research is required to explore how teachers can engage with data associated with teaching to encourage teacher reflection, improve the quality of teaching, and provide useful insights into ways teachers could be supported to interact with teaching data effectively. However, it is also essential to be aware that there are critical challenges associated with data collection. Moreover, designing the information flow that facilitates evidence-based decision-making requires addressing issues such as the potential risk of bias; ethical and privacy concerns; inadequate knowledge of how to engage with analytics effectively.

To ensure that instructional design and learning support is evidence-based, it is essential to empower teachers with the necessary knowledge of analytics and data literacy. The lack of such knowledge can lead to poor interpretation of analytics, which in turn can lead to ill-informed decisions that can significantly affect students; creating more inequalities in access to learning opportunities and support regimes. Teacher data literacy refers to a teachers’ ability to effectively engage with data and analytics to make better pedagogical decisions.

The primary outcome of TA is to guide educational researchers to develop better strategies to support the development of teachers’ data literacy skills and knowledge. However, for teachers to embrace data-driven approaches to learning design, there is a need to implement bottom-up approaches that include teachers as main stakeholders of a data literacy project, rather than end-users of data.

The purpose of this research is to explore the current discusses in the literature relating to TA. A vital goal of the review was to extend our understanding of conceptions and value of TA. Secondly, we want to contextualise the notion of TA and develop various concepts around TA to establish a framework that describes multiple aspects of TA. Thirdly, to examine different data collections/sources, machine learning algorithms, visualisations and actions associated with TA. The intended outcome is to develop a model that would provide a guide for the teacher to improve teaching practice and ultimately enhance learning outcomes.

The research employed a systematic and critical analysis of articles published from the year 2012 to 2019. A total of 58 publications were initially identified and compiled from the Scopus database. After analysing the search results, 31 papers were selected for review. This review examined research relating to the utilisation of analytics associated with teaching and teacher activities and provided conceptual clarity on TA. We found that the literature relating to conception, and optimisation of TA is sporadic and scare, as such the notion of TA is theoretically underdeveloped.

Methods and procedures

This research used the Tripartite model ( Daniel and Harland 2017 ), illustrated in Fig.  1 , to guide the systematic literature review. The Tripartite model draws from systematic review approaches such as the Cochrane, widely used in the analyses of rigorous studies, to provide the best evidence. Moreover, the Tripartite model offers a comprehensive view and presentation of the reports. The model composes of three fundamental components; descriptive (providing a summary of the literature), synthesis (logically categorising the research based on related ideas, connections and rationales), and critique (criticising the novel, providing evidence to support, discard or offer new ideas about the literature). Each of these phases is detailed fully in the following sections.

Tripartite Model. The Tripartite Model: A Systematic Literature Review Process ( Daniel and Harland 2017 )

To provide clarity; the review first focused on describing how TA is conceptualised and utilised. Followed by the synthesis of the literature on the various tools used to harvest, analyse and present teaching-related data to the teachers. Then the critique of the research which led to the development of a conceptual framework describing various aspects of TA. Finally, this paper proposes a Teaching Outcome Model (TOM). TOM is intended to offer teachers help on how to engage and reflect on teaching data.

TOM is a TA life cycle which starts with the data collection stage; where the focus is on teaching data. Then the data analysis stage; the application of different Machine Learning (ML) techniques to the data to discover hidden patterns. Subsequently, the data visualisation stage, where data presentation is carried out in the form of a Teaching Analytics Dashboard (TAD) for the teacher. This phase is where the insight generation, critical thinking and teacher reflection are carried out. Finally, the action phase, this is where actions are implemented by teachers to improve teaching practice. Some of these actions include improving the LD, changing teaching method, providing appropriate feedback and assessment or even carrying out more research. This research aims to inform the future work in the advancement of TA research field.

Framing research area for review

As stated in the introduction, understanding current research on TA can be used to provide teachers with strategies that can help them utilise various forms of data to optimise teaching performance and outcome. Framing the review was guided by some questions and proposed answers to address those questions (see Table  1 )

Inclusion and exclusion criteria

The current review started with searching through the Scopus database using the SciVal visualisation and analytical tool. The rationale for choosing the Scopus database is that it contains the largest abstract and citation database of peer-reviewed research literature with diverse titles from publishers worldwide. Hence, it is only conceivable to search for and find a meaningful balance of the published content in the area of TA. Also, the review included peer-reviewed journals and conference proceedings. We excluded other documents and source types, such as book series, books, editorials, trade publications on the understanding that such sources might lack research on TA. Also, this review excluded articles published in other languages other than English.

Search strategy

This review used several keywords and combinations to search on terms related to TA. For instance: ’Teaching Analytics’ AND ’Learning Analytics’ OR ’Teacher Inquiry’ OR ’Data Literacy’ OR ’Learning Design’ OR ’Computer-Supported Collaborative Learning’ OR ’Open Learner Model’ OR ’Visualisation’ OR ’Learning Management System’ OR ’Intelligent Tutoring System’ OR ’Student Evaluation on Teaching’ OR ’Student Ratings’.

This review searched articles published between 2012 to 2019. The initial stage of the literature search yielded 58 papers. After the subsequent screening of previous works and removing duplicates and titles that did not relate to the area of research, 47 articles remained. As such, a total of 36 studies continued for full-text review. Figure  2 , shows the process of finalising the previous studies of this review.

Inclusion Exclusion Criteria Flowchart. The selection of previous studies

Compiling the abstracts and the full articles

The review ensured that the articles identified for review were both empirical and conceptual papers. The relevance of each article was affirmed by requiring that chosen papers contained various vital phrases all through the paper, as well as, title, abstract, keywords and, afterwards, the entire essay. In essence, were reviewed giving particular cognisance and specific consideration to those section(s) that expressly related to the field of TA. In doing as such, to extract essential points of view on definitions, data sources, tools and technologies associated with analytics for the teachers. Also, this review disregarded papers that did not, in any way, relate to analytics in the context of the teachers. Finally, 31 articles sufficed for this review.

Systematic review: descriptive

Several studies have demonstrated that TA is an important area of inquiry ( Flanders 1970 ; Gorham 1988 ; Pennings et al. 2014 ; Schempp et al. 2004 ), that enables researchers to explore analytics associated with teaching process systematically. Such analytics focus on data related to the teachers, students, subjects taught and teaching outcomes. The ultimate goal of TA is to improve professional teaching practice ( Huang 2001 ; Sergis et al. 2017 ). However, there is no consensus on what constitutes TA. Several studies suggest that TA is an approach used to analyse teaching activities ( Barmaki and Hughes 2015 ; Gauthier 2013 ; KU et al. 2018 ; Saar et al. 2017 ), including how teachers deliver lectures to students, tools usage pattern, or dialogue. While various other studies recognise TA as the ability to applying analytical methods to improve teacher awareness of student activities for appropriate intervention ( Ginon et al. 2016 ; Michos and Hernández Leo 2016 ; Pantazos et al. 2013 ; Taniguchi et al. 2017 ; Vatrapu et al. 2013 ). A hand full of others indicate TA as analytics that combines both teachers and students activities ( Chounta et al. 2016 ; Pantazos and Vatrapu 2016 ; Prieto et al. 2016 ; Suehiro et al. 2017 ). Hence, it is particularly problematic and challenging to carry out a systematic study in the area of analytics for the teachers to improve teaching practice, since there is no shared understanding of what constitutes analytics and how best to approach TA.

Researchers have used various tools to automatically harvest important episodes of interactive teacher and student behaviour during teaching, for teacher reflection. For instance, KU et al. ( 2018 ), utilised instruments such as; Interactive Whiteboard (IWB), Document Camera (DC), and Interactive Response System (IRS) to collect classroom instructional data during instruction. Similarly, Vatrapu et al. ( 2013 ) employed eye-tracking tools to capture eye-gaze data on various visual representations. Thomas ( 2018 ) also extracted multimodal features from both the speaker and the students’ audio-video data, using digital devices such as cameras and high-definition cameras. Data collected from some of these tools not only provide academics with real-time data but also attract more details about teaching and learning than the teacher may realise. However, the cost of using such digital tools for large-scale verification is high, and cheaper alternatives are sort after. For instance, Suehiro et al. ( 2017 ) proposed a novel approach of using e-books to extract teaching activity logs in a face-to-face class efficiently.

Vatrapu ( 2012 ) considers TA as a subset of LA dedicated to supporting teachers to understand the learning and teaching process. However, this definition does not recognise that both the learning and teaching processes are intertwined. Also, most of the research in LA collects data about the student learning or behaviour, to provide feedback to the teacher ( Vatrapu et al. 2013 ; Ginon et al. 2016 ; Goggins et al. 2016 ; Shen et al. 2018 ; Suehiro et al. 2017 ), see, for example, the iKlassroom conceptual proposal by Vatrapu et al. ( 2013 ), which highlights a map of the classroom to help contextualise real-time data about the learners in a lecture. Although, a few research draw attention to the analysis of teacher-gathering and teaching practice artefacts, such as lesson plans. Xu and Recker ( 2012 ) examined teachers tool usage patterns. Similarly, Gauthier ( 2013 ) extracted the analysis of the reasoning behind the expert teacher and used such data to improve the quality of teaching.

Multimodal analytics is an emergent trend used to complement available digital trace with data captured from the physical world ( Prieto et al. 2017 ). Isolated examples include the smart school multimodal dataset conceptual future proposal by Prieto et al. ( 2017 ), which features a plan of implementing a smart classroom to help contextualise real-time data about both the teachers and learners in a lecture. Another example, Prieto et al. ( 2016 ), explored the automatic extraction of orchestration graphs from a multimodal dataset gathered from only one teacher, classroom space, and a single instructional design. Results showed that ML techniques could achieve reasonable accuracy towards automated characterisation in teaching activities. Furthermore, Prieto et al. ( 2018 ) applied more advanced ML techniques to an extended version of the previous dataset to explore the different relationships that exist between datasets captured by multiple sources.

Previous studies have shown that teachers want to address common issues such as improving their TPD and making students learn effectively ( Charleer et al. 2013 ; Dana and Yendol-Hoppey 2019 ; Pennings et al. 2014 ). Reflection on teaching practice plays an essential role in helping teachers address these issues during the process of TPD ( Saric and Steh 2017 ; Verbert et al. 2013 ). More specifically, reflecting on personal teaching practice provides opportunities for teachers to re-examine what they have performed in their classes ( Loughran 2002 ; Mansfield 2019 ; Osterman and Kottkamp 1993 ). Which, in turn, helps them gain an in-depth understanding of their teaching practice, and thus improve their TPD. For instance, Gauthier ( 2013 ), used a visual teach-aloud method to help teaching practitioners reflect and gain insight into their teaching practices. Similarly, Saar et al. ( 2017 ) talked about a self-reflection as a way to improve teaching practice. Lecturers can record and observe their classroom activities, analyse their teaching and make informed decisions about any necessary changes in their teaching method.

The network analysis approach is another promising field of teacher inquiry, especially if combined with systematic, effective qualitative research methods ( Goggins et al. 2016 ). However, researchers and teacher who wish to utilise social network analysis must be specific about what inquiry they want to achieve. Such queries must then be checked and validated against a particular ontology for analytics ( Goggins 2012 ). Goggins et al. ( 2016 ), for example, aimed at developing an awareness of the types of analytics that could help teachers in Massive Open Online Courses (MOOCs) participate and collaborate with student groups, through making more informed decisions about which groups need help, and which do not. Network theory offers a particularly useful framework for understanding how individuals and groups respond to each other as they evolve. Study of the Social Network (SNA) is the approach used by researchers to direct analytical studies informed by network theory. SNA has many specific forms, each told by graph theory, probability theory, and algebraic modelling to various degrees. There are gaps in our understanding of the link between analytics and pedagogy. For example, which unique approaches to incorporating research methods for qualitative and network analysis would produce useful information for teachers in MOOCs? A host of previous work suggests a reasonable path to scaling analytics for MOOCs will involve providing helpful TA perspectives ( Goggins 2012 ; Goggins et al. 2016 ; Vatrapu et al. 2012 ).

Teacher facilitation is considered a challenging and critical aspect of active learning ( Fischer et al. 2014 ). Both educational researchers and practitioners have paid particular attention to this process, using different data gathering and visualisation methods, such as classroom observation, student feedback, audio and video recordings, or teacher self-reflection. TA enables teachers to perform analytics through visual representations to enhance teachers’ experience ( Vatrapu et al. 2011 ). As in a pedagogical environment, professionals have to monitor several data such as questions, mood, ratings, or progress. Hence, dashboards have become an essential factor in improving and conducting successful teaching. Dashboards are visualisation tools enable teachers to monitor and observe teaching practice to enhance teacher self-reflection ( Yigitbasioglu and Velcu 2012 ). While a TAD is a category of dashboard meant for teachers and holds a unique role and value [62]. First, TAD could allow teachers to access students learning in an almost real-time and scalable manner ( Mor et al. 2015 ), consequently, enabling teachers to improve their self-knowledge by monitoring and observing students activities. TAD assists the teachers in obtaining an overview of the whole classroom as well as drill down into details about individual and groups of students to identify student competencies, strengths and weaknesses. For instance, Pantazos and Vatrapu ( 2016 ) described TAD for repertory grid data to enable teachers to conduct systematic visual analytics of classroom learning data for formative assessment purposes. Second, TAD also allows for tracking on teacher self-activities ( van Leeuwen et al. 2019 ), as well as students feedback about their teaching practice. For example,Barmaki and Hughes ( 2015 ) explored a TAD that provides automated real-time feedback based on speakers posture, to support teachers practice classroom management and content delivery skills. It is a pedagogical point that dashboards can motivate teachers to reflect on teaching activities, help them improve teaching practice and learning outcome ( 2016 ). The literature has extensively described extensively, different teaching dashboards. For instance, Dix and Leavesley ( 2015 ), broadly discussed the idea of TAD and how they can represent visual tools for academics to interface with learning analytics and other academic life. Some of these academic lives may include schedules such as when preparing for class or updating materials, or meeting times such as meeting appointments with individual or collective group of students. Similarly, Vatrapu et al. ( 2013 ) explored TAD using visual analytics techniques to allow teachers to conduct a joint analysis of students personal constructs and ratings of domain concepts from the repertory grids for formative assessment application.

Systematic review: synthesis

In this second part of the review process, we extracted selected ideas from previous studies. Then group them based on data sources, analytical methods used, types of visualisations performed and actions.

Data sources and tools

Several studies have used custom software and online applications such as employing LMS and MOOCs to collect online classroom activities ( Goggins et al. 2016 ; KU et al. 2018 ; Libbrecht et al. 2013 ; Müller et al. 2016 ; Shen et al. 2018 ; Suehiro et al. 2017 ; Vatrapu et al. 2013 ; Xu and Recker 2012 ). Others have used modern devices including eye-tracker, portable electroencephalogram (EEG), gyroscope, accelerometer and smartphones ( Prieto et al. 2016 ; Prieto et al. 2018 ; Saar et al. 2017 ; Saar et al. 2018 ; Vatrapu et al. 2013 ), and conventional instruments such as video and voice recorders ( Barmaki and Hughes 2015 ; Gauthier 2013 ; Thomas 2018 ), to record classroom activities. However, some authors have pointed out several issues with modern devices such as expensive equipment, high human resource and ethical concerns ( KU et al. 2018 ; Prieto et al. 2017 ; Prieto et al. 2016 ; Suehiro et al. 2017 ).

In particular, one study by Chounta et al. ( 2016 ) recorded classroom activities using humans to code tutor-student dialogue manually. However, they acknowledged that manual coding of lecture activities is complicated and cumbersome. Some authors also subscribe to this school of thought and have attempted to address this issue by applying Artificial Intelligence (AI) techniques to automate and scale the coding process to ensure quality in all platforms ( Prieto et al. 2018 ; Saar et al. 2017 ; Thomas 2018 ). Others have proposed re-designing TA process to automate the process of data collection as well as making the teacher autonomous in collecting data about their teaching ( Saar et al. 2018 ; Shen et al. 2018 ). Including using technology that is easy to set up, effortless to use, does not require much preparation and at the same time, not interrupting the flow of the class. In this way, they would not require researcher assistance or outside human observers. Table  2 , summarises the various data sources as well as tools that are used to harvest teaching data with regards to TA.

The collection of evidence from both online and real classroom practice is significant both for educational research and TPD. LA deals mostly with data captured from online and blended learning platforms (e.g., log data, social network and text data). Hence, LA provides teachers with data to monitor and observe students online class activities (e.g., discussion boards, assignment submission, email communications, wiki activities and progress). However, LA neglects to capture physical occurrences of the classroom and do not always address individual teachers’ needs. TA requires more adaptable forms of classroom data collection (e.g., through video- recordings, sensor recording or by human observers) which are tedious, human capital intensive and costly. Other methods have been explored to balance the trade-off between data collected online, and data gathered from physical classroom settings by implementing alternative designs approach ( Saar et al. 2018 ; Suehiro et al. 2017 ).

Analysis methods

Multimodal analytics is the emergent trend that will complement readily available digital traces, with data captured from the physical world. Several articles in the literature have used multimodal approaches to analyse teaching processes in the physical world ( Prieto et al. 2016 ; Prieto et al. 2017 ; Prieto et al. 2018 ; Saar et al. 2017 ; Thomas 2018 ). In university settings, unobtrusive computer vision approaches to assess student attention from their facial features, and other behavioural signs have been applied ( Thomas 2018 ). Most of the studies that have ventured into multimodal analytics applied ML algorithms to their captured datasets to build models of the phenomena under investigation ( Prieto et al. 2016 ; Prieto et al. 2018 ). Apart from research areas that involve multimodal analytics, other areas of TA research have also applied in ML techniques such as teachers tool usage patterns ( Xu and Recker 2012 ), online e-books ( Suehiro et al. 2017 ), students written-notes ( Taniguchi et al. 2017 ). Table  3 outlines some of the ML techniques applied from previous literature in TA.

Visualisation methods

TA allows teachers to apply visual analytics and visualisation techniques to improve TPD. The most commonly used visualisation techniques in TA are statistical graphs such as line charts, bar charts, box plots, or scatter plots. Other visualisation techniques include SNA, spatial, timeline, static and real-time visualisations. An essential visualisation factor for TA is the number of users represented in a visualisation technique. Serving single or individual users allows the analyst to inspect the viewing behaviour of one participant. Visualising multiple or group users at the same time can allow one to find strategies of groups. However, these representations might suffer from visual clutter if too much data displays at the same time. Here, optimisation strategies, such as averaging or bundling of lines might be used, to achieve better results. Table  4 represents the visualisation techniques mostly used in TA.

Systematic review: critique

Student evaluation on teaching (set) data.

Although the literature has extensively reported various data sources used for TA, this study also draws attention to student feedback on teaching, as another form of data that originates from the classroom. The analytics of student feedback on teaching could support teacher reflection on teaching practice and add value to TA. Student feedback on teaching is also known as student ratings, or SET is a form of textual data. It can be described as a combination of both quantitative and qualitative data that express students opinions about particular areas of teaching performance. It has existed since the 1920s ( Marsh 1987 ; Remmers and Brandenburg 1927 ), and used as a form of teacher feedback. In addition to serving as a source of input for academic improvement ( Linse 2017 ), many universities also rely profoundly on SET for hiring, promoting and firing instructors ( Boring et al. 2016 ; Harland and Wald 2018 ).

Technological advancement has enabled institutions of Higher Education (HE) to administer course evaluations online, forgoing the traditional paper-and-pencil ( Adams and Umbach 2012 ). There has been much research around online teaching evaluations. Asare and Daniel ( 2017 ) investigated the factors influencing the rate at which students respond to online SET. While there is a verity of opinions as to the validity of SET as a measure of teaching performance, many teaching academics and administrators perceive that SET is still the primary measure that fills this gap ( Ducheva et al. 2013 ; Marlin Jr and Niss 1980 ). After all, who experiences teaching more directly than students? These evaluations generally consist of questions addressing the instructor’s teaching, the content and activities of the paper, and the students’ own learning experience, including assessment. However, it appears these schemes gather evaluation data and pass on the raw data to the instructors and administrators, stopping short of deriving value from the data to facilitate improvements in the instruction and the learning experiences. This measure is especially critical as some teachers might have the appropriate data literacy skills to interpret and use such data.

Further, there are countless debates over the validity of SET data ( Benton and Cashin 2014 ; MacNell et al. 2015 ). These debates have highlighted some shortcomings of student ratings of teaching in light of the quality of instruction rated ( Boring 2015 ; Braga et al. 2014 ). For Edström, what matters is how the individual teacher perceives an evaluation. It could be sufficient to undermine TPD, especially if the teachers think they are the subjects of audit ( Edström 2008 ). However, SET is today an integral part of the universities evaluation process ( Ducheva et al. 2013 ). Research has also shown that there is substantial room for utilising student ratings for improving teaching practice, including, improving the quality of instruction, learning outcomes, and teaching and learning experience ( Linse 2017 ; Subramanya 2014 ). This research aligns to the side of the argument that supports using SET for instructional improvements, to the enhancement of teaching experience.

Systematically, analytics of SET could provide valuable insights, which can lead to improving teaching performance. For instance, visualising SET can provide some way, a teacher can benchmark his performance over a while. Also, SET could provide evidence to claim for some level of data fusion in TA, as argued in the conceptualisation subsection of TA.

Transformational TA

The growing research into big data in education has led to renewed interests in the use of various forms of analytics ( Borgman et al. 2008 ; Butson and Daniel 2017 ; Choudhury et al. 2002 ). Analytics seeks to acquire insightful information from hidden patterns and relationships in data that ordinarily would not be visible by the natural eyes, except with the application of state-of-the-art models and methods. Big data analytics in HE provides lenses on students, teachers, administrators, programs, curriculum, procedures, and budgets ( Daniel 2015 ). Figure  3 illustrates the types of analytics that applies to TA to transform HE.

Types of analytics in higher education ( Daniel 2019 )

Descriptive Analytics Descriptive analytics aims to interpret historical data to understand better organisational changes that have occurred. They are used to answer the "What happened?" information regarding a regulatory process such as what are the failure rates in a particular program ( Olson and Lauhoff 2019 ). It applies simple statistical techniques such as mean, median, mode, standard deviation, variance, and frequency to model past behaviour ( Assunção et al. 2015 ; ur Rehman et al. 2016 ). Barmaki and Hughes ( 2015 ) carried out some descriptive analytics to know the mean view time, mean emotional activation, and area of interest analysis on the data generated from 27 stimulus images to investigate the notational, informational and emotional aspect of TA. Similarly, Michos and Hernández-Leo ( 2016 ) demonstrated how descriptive analytics could support teachers’ reflection and re-design their learning scenarios.

Diagnostic Analytics Diagnostic analytics is higher-level analytics that further diagnoses descriptive analytics ( Olson and Lauhoff 2019 ). They are used to answer the "Why it happened?". For example, a teacher may need to carry out diagnostic analytics to know why there is a high failure rate in a particular programme or why students rated a course so low for a specific year compared to the previous year. Diagnostic analytics uses some data mining techniques such as; data discovery, drill-down and correlations to further explore trends, patterns and behaviours ( Banerjee et al. 2013 ). Previous research has applied the repertory grid technique as a pedagogical method to support the teachers perform knowledge diagnostics of students about a specific topic of study ( Pantazos and Vatrapu 2016 ; Vatrapu et al. 2013 ).

Relational Analytics Relational analytics is the measure of relationships that exists between two or more variables. Correlation analysis is a typical example of relational analytics that measures the linear relationship between two variables ( Rayward-Smith 2007 ). For instance, Thomas ( 2018 ) applied correlation analysis to select the best features from the speaker and audience measurements. Some researchers have also referred to other forms of relational analytics, such as co-occurrence analysis to reveal students hidden abstract impressions from students written notes ( Taniguchi et al. 2017 ). Others have used relational analytics to differentiate critical formative assessment futures of an individual student to assist teachers in the understanding of the primary components that affect student performance ( Pantazos et al. 2013 ; Michos and Hernández Leo 2016 ). A few others have applied it to distinguish elements or term used to express similarities or differences as they relate to their contexts ( Vatrapu et al. 2013 ). Insights generated from this kind of analysis can be considered to help improve teaching in future lectures and also compare different teaching styles. Sequential pattern mining is also another type of relational analytics used to determine the relationship that exists between subsequent events ( Romero and Ventura 2010 ). It can be applied in multimodal analytics to cite the relationship between the physical aspect of the learning and teaching process such as the relationship between ambient factors and learning; or the investigation of robust multimodal indicators of learning, to help in teacher decision-making ( Prieto et al. 2017 ).

Predictive Analytics Predictive analytics aims to predict future outcomes based on historical and current data ( Gandomi and Haider 2015 ). Just as the name infers, predictive analytics attempts to predict future occurrences, patterns and trends under varying conditions ( Joseph and Johnson 2013 ). It makes use of different techniques such as regression analysis, forecasting, pattern matching, predictive modelling and multi-variant statistics ( Gandomi and Haider 2015 ; Waller and Fawcett 2013 ). In prediction, the goal is to predict students and teachers activities to generate information that can support decision-making by the teacher ( Chatti et al. 2013 ). Predictive analytics is used to answer the "What will happen". For instance, what are the interventions and preventive measures a teacher can take to minimise the failure rate? Herodotou et al. ( Herodotou et al. 2019 ) provided evidence on how predictive analytics can be used by teachers to support active learning. An extensive body of literature suggests that predictive analytics can help teachers improve teaching practice ( Barmaki and Hughes 2015 ; Prieto et al. 2016 ; Prieto et al. 2018 ; Suehiro et al. 2017 ) and also to identify group of students that might need extra support to reach desired learning outcomes ( Goggins et al. 2016 ; Thomas 2018 ).

Prescriptive Analytics Prescriptive analytics provides recommendations or can automate actions in a feedback loop that might modify, optimise or pre-empt outcomes ( Williamson 2016 ). It is used to answer the "How will it best happen?". For instance, how will teachers make the right interventions for students that have been perceived to be at risk to minimise the student dropout rate or what kinds of resources are needed to support students who might need them to succeed? It determines the optimal action that enhances the business processes by providing the cause-effect relationship and applying techniques such as; graph analysis, recommendation engine, heuristics, neural networks, machine learning and Markov process ( Bihani and Patil 2014 ; ur Rehman et al. 2016 ). For example, applying curriculum Knowledge graph and learning Path recommendation to support teaching and learners learning process ( Shen et al. 2018 ).

Actionable Analytics Actionable analytics refers to analytics that prompt action ( Gudivada et al. 2016 ; Gudivada et al. 2018 ; Winkler and Söllner 2018 ). Norris et al. ( 2008 ) used the term action analytics to describe "the emergence of a new generation of tools, solutions, and behaviours that are giving rise to more powerful and effective utilities through which colleges and universities can measure performance and provoke pervasive actions to improve it". The educational sector can leverage some of these innovative, new and cutting edge technologies and techniques such as Natural Language Processing (NLP) ( Sergis and Sampson 2016 ; Taniguchi et al. 2017 ), big data analytics ( Goggins et al. 2016 ) and deep learning ( Prieto et al. 2018 ) to support teacher in both the teaching and learning processes.

Institutional Transformation Data in themselves are not useful; they only become valuable if they can be used to generate insight. In other words, analytics can be applied to institutional data to optimise productivity and performance of the institutional operations, thereby providing value that can transform the institutional practices. In education, there are various purposes of analytics, ranging from those that provide institutions with an overview or deep-down microscopic view of individual students, faculty, curriculum, programs, operations and budgets, to those capable of predicting future trends. Unveiling the value of TA empowers the teachers to identify issues and transform difficulties into opportunities. These opportunities can be employed to optimises the institutional processes, enhance learner experiences and improve teaching performance. TA and LA both play a vital role in effectively reforming and transforming the educational sector to catch up with the fast pace at which data generates. For example, with the extensive use of online and blended learning platforms, the application of analytics will enable institutional stakeholders at all levels to gain new insights into educational data. Today, the HE sector is at crossroads, where there is a need for synergies in learning research and data analytics to transform the way teaching and learning are fundamentally carried out.

The link between TA, LA and LD

Primarily, TA aims to link the centrepiece of LA and remodel them to address teaching challenges. More specifically, TA argues that connecting and analysing insights generated from LA methods and tools with those generated from in-class methods and tools, through TA tools could support teacher reflection and improve TPD based on evidence. Hence, this concept is presented further in the next subsection.

Conceptual framework of TA

Based on the different perceptions of TA described in previous reviews, this study proposes a conceptual framework for TA to model the complex interaction existing around TA. Three nodes (LA, TA and LD) are interconnected to each other forming a triadic network with the teacher at the centre, performing value-added interactions to make informed based decisions. Each part of this interconnection forms a triangle, totalling three triangles (A, B and C) (see Fig.  4 ).

Conceptualisation of TA. Triadic TA Conceptual Framework

The proposed framework is not bound to any particular implementation of learning or design technology. Instead, the point is to describe the elements of analytics and data sources that are key for each domain to guide the use of analytical methods, tools and technology to support the multiple dimensions of learning design successfully.

This triad illustrates the interaction occurring between the teacher, the LA and the LD, to inform TPD. Hernández-Leo et al. ( 2019 ) argued that LD could contribute to structuring and orchestrating the design intent with learners digital trace patterns, advancing the knowledge and interpretation of LA. LA tailored to fit the design intent could be considered by teachers as contributing to the enhancement of the LD in subsequent design interactions. For example, LA could be an information tool to inform the tutors or designers of pedagogical decision making ( Persico and Pozzi 2015 ). Hence, a teacher may want to utilise LA to make just-in-time pedagogical decisions, such as grouping students based on their performance.

Similarly, a teacher may want to investigate if the estimated time taken for students to carry out learning tasks is reasonable or whether adjustments need to be made to the course design ( Hernández-Leo et al. 2019 ; Pozzi and Persico 2013 ). This domain can also provide teachers with analytics regarding the challenges and difficulties students face in the problem-solving phase while performing a task. In return, they give the teacher information in the form of TAD summarising the various challenges students encountered with that activity. They may also provide solutions on how to address them. For example, an early alert system that instantiates a dashboard for instructors using some metrics calculations such as login counts and page views ( Thille and Zimmaro 2017 ). The data sources in the LA node can improve teachers’ awareness, which could also lead to the improvement of LD and help to distinguish design elements that could modify future designs. Data collection in this domain is mostly automatic through virtual learning environments (e.g., LMS, MOOCs). Other forms of data collection may include social media platforms (e.g., Facebook, Tweeter), wearable sensors (e.g., eye-trackers, EEG), software tools that support and collect data related to specific student activities and attendance ( Bakharia et al. 2016 ; Bos and Brand-Gruwel 2016 ).

This triangle represents the relationship between the teacher, the LD and TA. While experiencing LD, TA endeavours to handle continues teachers’ engagement, progression, achievement and learners satisfaction ( Bakharia et al. 2016 ; Sergis and Sampson 2017 ). For example, exploring the impact of video shot on instructor performance and student learning. Using MOOC AB testing, teachers could experiment whether a difference in video production setting would have any impact on the instructors acting performance, or whether any changes in format and instructors performance will result in detectable differences in student viewing behaviour ( Chen et al. 2016 ).

Further, data sources in TA could assist teacher reflection on the impacts of their LD. Data collection could also be automatic by the use of wearable sensors on the teachers while performing teaching activities, also known as in-class analytics. Several institutions now record video contents of their face-to-face classes. Some others even go a step further by collecting their physiological data. These datasets, as mentioned earlier, have a way of exemplifying and illustrating things that ordinarily, a book of pedagogy cannot convey, in providing systematic feedback for the teachers. It involves capturing data during a traditional in-class, face-to-face teacher-centric instruction or teacher-student interaction (where students learn by directly or indirectly interacting with instructors in a lab or lecture hall) and analysing data to identify areas of possible improvements. The kind of data usually captured in this setting are audio, video, body movement, brain activity, cortex activity, to mention just a few. For example, a teacher can perform diagnostic analysis on class recorded videos to expose what is intrinsic during his lecture. This kind of diagnostic analysis could help teachers understand more about their teaching and discover areas of further improvement. SET is another form of data about the teachers; they are collected via the institutional application platforms ( Hernández-Leo et al. 2019 ) and can be visualised to improve teaching performance..

Analytics that happens in the LD involves the visualisation of teaching design to facilitate teacher reflection on the lesson plan, visualisation of the extent to which the lesson plan aligns with the educational objectives, and finally, validation of the lesson plan to highlight potential inconsistencies in the teaching design. For example, a teacher can visualise the number of assessment activities of the lesson plan or the various types of educational resources used in the lesson plan, to know if they are still valid or obsolete. Similarly, a teacher could analyse the time allocated for each lesson activity, to find out if the time allocated for each activity is good enough, or visualise the level of inconsistencies of time misappropriations and imbalances between the overall lesson plan and the individual lesson activities.

This area presents the communication between the teacher, the LA and the TA. Chinchu Thomas ( 2018 ) explored the correlation between student ratings on teaching and student physiological data. Similarly, Schmidlin ( 2015 ) established how to analyse and cross-reference data without decrypting the data sources. Hence, we argue that SET could be linked with LA such as student digital traces from LMS ( Stier et al. 2019 ) and other forms of data (such as attendance data), without compromising privacy. This claim for data fusion could support the teachers to make informed-decisions in new ways. For example, analytics performed on linked datasets could quickly reveal those student opinions that may not count at the end of the semester courses.

Visualisations that could quickly realise students with low participation rates and link it to their opinions, without revealing any identity. Additionally, teachers may be interested in comparing the view of students with low participation rate with those of high participation rate. This kind of information may lead teachers towards making explicit judgements with evidence. A tutor may choose to disregard the opinions of those students that participated less than 20 per cent in-class activities and assignments, as well as had a low attendance rate. Hence, narrowing concentration more on the opinions of students that participated in improving teaching practice.

However, considering ethical concerns, data fusion at the individual level still requires explicit and informed consent from the students whose data are collected ( Menchen-Trevino 2016 ). Other issues such as privacy concerns, data fusion can be problematic as this usually requires that the teachers know student identities. However, from a programmatic perspective, extra measures can be put in place to address this concern. Algorithms can be interfaced to mask student identities to some other unique identities to make them anonymous but linked ( Schmidlin et al. 2015 ) to provide a richer set of data for the teacher to make informed decisions.

Teachers can get a better picture towards improving the context in which learning happens, only if they can be informed about both how they teach and how students learn. Hence, this framework aims to continually provide teachers with interesting information from intelligent feedback based on data generated from users and learning context to improve their learning design and teaching outcome continuously.

Teaching Outcome Model (TOM)

Design-based research advances instructional design work, theory, and implementation as iterative, participatory, and located rather than processes "owned and operated" by designers of instructions ( Wang and Hannafin 2005 ). TOM is an iterative process that follows a design-based research approach to guide teachers, researchers, faculty and administrators on how to utilise data to improve the quality of teaching and learning outcome. This model enables teachers to investigate and evaluate their work using data. Consequently, improving the teacher use of data to inform teaching practice. To build more awareness with regards to teaching data, TOM models TA through iterative cycles of data collection, data analysis, data visualisation and action stages which are interdependent of each other (see Fig.  5 ). Design-based research, as a pragmatic methodology, can guide TOM while generating insights that can support teacher reflections on teaching and student learning. Conversely, TOM ensures that design-based research methodologies can be operational and systemised. Following the various stages outlined in the model, teachers can regularly identify, match and adjust teaching practice, and learning design to all the learners need.

Teaching Outcome Model. TA Life cycle

In the data collection stage, a constant stream of data accumulates from the digital traces relating to teaching daily activities and engagements, including structured and unstructured data, visual and non-visual data, historical and real-time data. It is also important to note that the rate at which diverse data accumulates in our educational system will keep growing. According to Voithofer and Golan ( 2018 ), there are several ways to mine teaching and learning data without professional knowledge that is beyond the necessary teacher training experience in data literacy, administering learning design and class orchestration. Subscribing to this school of thought, adopting Big data infrastructure in our institutions will guarantee easy access to data by the various stakeholders, this will also mitigate the bottleneck of disparate data points existing in our educational sector. Therefore, enabling educators to focus more attention on instruction, setting up interactive class activities, and participating more on discussions that will create more data for evidence-based decision making. Also, the misuse of data is a broad primary concern ( Roberts et al. 2017 ). One critical matter is identifying the types of data that can be collected, analysed and visualized; to ensure that the right people have access to the data for the right purpose. As such, implementing data governance policies around institutional data such as; ’open definition of purpose, scope and boundaries, even if that is broad and in some respects, open-ended’ is critical ( Kay et al. 2012, p 6 ). This sort of measure will introduce clarity and address issues around who controls what data as well as security and privacy issues around data.

Analysis stage

This step involves the different ways of working with data to ensure data quality. Professionals such as data scientists, programmers, engineers and researchers need to work together with the teachers at this level. They can apply data mining techniques, statistical methods, complex algorithms, and AI techniques (such as NLP, AI, ML, deep learning) to adequately transform data into the useful analytical process. Analytics in the education space presents in diverse forms including, descriptive, diagnostic, predictive and prescriptive. These different forms of analytics can be utilised to offer a high-level view or fine-grained view of individual learners, teacher, faculty and their various activities, engagements and behaviours. Unravelling the value of data analytics empowers teachers and researchers to identify problems and transform challenges into opportunities that can be utilised to support teacher reflection and enrich teacher data-literacy experiences. For example, teachers can apply NLP on text data to gather topics from discussion posts, contributions participants have made within collaborative projects and their sentiments.

Furthermore, ML techniques could be combined with TA to enhance teaching outcome. For instance, chatbots could support the teacher by acting as a teacher assistant in large classes. An essential consideration in analytics, however, is that data can be easily de-identified ( Roberts et al. 2017 ; Cumbley and Church 2013 ), especially when data sets increase in size and scope and are combined to generate big data. To resolve these concerns, a particular university introduced a two-stage method of data de-identification coupled with data governance to restrict data access ( De Freitas et al. 2015 ).

Visualisation stage

This stage ensures data presentation in useful and meaningful ways to teachers. Empowering teachers with interactive visual interfaces and dashboards that facilitate teacher cognition and promote reflection about pre-processed and fine-grained teaching and learning activities. Through TAD, can project real-time and historical information from different data sources that might not be necessarily interoperable, and results summarised ( Moore 2018 ). However, visualisation is "what you see is what you get"; meaning that information presentation method may affect its interpretation, and consequently, may influence decision-making. Hence, it is necessary to address issues around visualisations in diverse forms such as; visual analytics and exploratory data analysis to create room for visual interactivity, exploratory visualisation to discover trends, patterns, relationships and behaviours. For example, a teacher can use a TAD to monitor student engagement. When the student engagement is poor, it may prompt the teacher to take necessary actions such as; changing teaching material and making it more interactive. Additionally, there are also questions around privacy, such as who has access to visualisations relevant to an instructor, such as other faculty members participating in the course, directly or indirectly, administrators, researchers, potential employees of other institutions.

Action stage

At this stage, informed-decision leads to action and actions unavoidably reshape our environment; subsequently, regenerate new data. Additionally, there is a to create tools that will be useful to the teacher to understand and make meaning of data quickly. Actions taken by teachers can be used to improve the course design and assessment (value-added formative assessment). In any case, predictive analytics prompts an epistemological question; how should we ensure effective action by the teacher based on flawed predictions such that the system does not collapse?

Discussion and conclusion

This article presents the result of a systematic literature review aimed at describing the conception, and synthesis of the current research on the notion of TA, to provide insight into how TA can be used to improve the quality of teaching. The first part of the article described what is meant by TA to consolidate the divergent discourse on TA. The review showed that TA applies to analytics on teaching activities as well as methods of improving teachers’ awareness on students’ activities, including supporting the teachers to understand student learning behaviours to provide adequate feedback to teachers. In essence, the primary goal of TA is to improve teaching performance. The literature also revealed the several tools and methods are available for extracting digital traces associated with teaching in addition to traditional student evaluation tools. However, one of the main challenges recognised was the cost associated with some devices used to capture in-class activities, and ML techniques have been proposed to minimise this challenge.

The literature has also recognised teacher inquiry as a promising area of research in TA and came to a consensus that methods, like multimodal analytics and SNA, could help promote teacher inquiry and teacher reflection. Visualisations and visual analytics techniques are very significant in TA and also encourage teacher inquiry. The use of visualisation dashboards and TAD are essential tools that the modern-day teachers require to carry out a continuous and efficient reflection on teaching practice.

The emphasis of the synthesis of TA was clearly on data collection, analysis and visualisation, as illustrated in Fig.  6 . In the literature, the various kinds of data collected and used to improve teaching practice, include:

Digital trace data; "records of activity (trace data) undertaken through an online information system (thus, digital)" [119]. They incorporate various activities generated from custom applications and learning environments that leave digital footprints.

Image data are photographic or trace objects that represent the underlying pixel data of an area of an image element.

Physiological data are body measurement based on body-mounted sensors ( Lazar et al. 2017 ), used to extract data from teachers while performing classroom teaching activities.

Audio-video stream data or recorded lecturer data with captured physical teaching activities and students learning activities. Hence, attainable with mounted cameras, computer or mobile cameras connected to applications like Zoom and Skype, eye tracks with recording capabilities and digital cameras connected to learning environments such as Eco365.

Social data are data with online social activities, including utilising the repertory grid technique to collect students’ assessment data from social media sites.

Text data, including quantitative and qualitative data, data generated from text documents such as discussion forums, students essay or articles, emails and chat messages.

Dimensions of TA. Illustration of TA based on the literature

Analysis in this context refers to the application of Educational Data Mining (EDM) and deep learning techniques mostly used to process data. EDM approaches is a complicated process that requires an interweaving of various specialised knowledge and ML algorithms, especially to improve teaching and learning ( Chen 2019 ). NLP and classification are the two main EDM techniques applied in TA. However, the review also recognised the use of other methods such as clustering and deep learning techniques, to support teachers.

As commonly said, a picture is worth more than a thousand words; visualisation can effectively communicate and reveal structures, patterns and trends in variables and their interconnections. Research in TA has applied several visualisation techniques including Network, Timeline, Spatial, Table and Statistical Graphs. For instance, SNA is a form of visual analytics that is used to support teachers to determine how different groups interact and engage with course resources. Identifying differences in interaction patterns for different groups of students may result in different learning outcomes, such as, how access patterns of successful groups of students differ from that of unsuccessful students. Applying visualisation techniques can support teachers in areas such as advising underperforming students about effective ways to approach study. Visualisation can enable teachers to identify groups of students that might need assistance and discover new and efficient means of using collaborative systems to achieve group work that can be taught explicitly to students.

However, while acknowledging the incomplete nature of data and complexities associated with data collection, analysis and use, teachers should take caution to avoid bais. Data collected in one context may not be directly applicable to another or have both benefits and cost for individuals or groups from which data was harvested. Therefore, key stakeholders, including teachers, course directors, unit coordinators and researchers must pay proper attention to predictive models and algorithms and take extra care to ensure that the contexts of data analysed are carefully considered. There are also privacy concerns, such as who has access to view analytics relating to a teacher, including other faculty members both directly or indirectly involved in the course, administrators, researchers, future employees of other institutions. It will be useful for institutions to have clear guidelines as to who has access to what and who views what. Other issues around data include how long should data remain accessible ( Siemens 2013 ), with big data technology and infrastructure, data should be kept for as long as it can exist. Pardo and Siemens ( 2014 ) acknowledged that the use of analytics in higher education research has no clear interpretation of the right to privacy. They seem opposed to the need for absolute privacy, on the basis that the use of historical data enhances research with potential rewards for the future of teaching professional development and student outcome.

The review provided in the current article highlighted the significant limitations in the existing literature on teaching analytics. The TAD is proposed to guide teachers, developers, and researchers to understand and optimise teaching and the learning environments. The critical aspect of this review is establishing the link between LA, TA and LD and its value in informing teachers’ inquiry process. Also, the review describes the relationship between LA, TA and LD. Finally, the article proposes TOM, which draws from a research-based approach to guide teachers on how to utilise data to improve teaching. The outcome of this model is a TAD that provides actionable insights for teacher reflection and informed decision-making. Therefore, showing the value that TA brings to pedagogic interventions and teacher reflection.

Theoretical implications

The analysis of data collected from the interaction of teachers with technology and students is a promising approach for advancing our understanding of the teaching process and how it can be supported. Teachers can use data obtained from their teaching to reflect on their pedagogical design and optimise the learning environment to meet students’ diverse needs and expectations.

Teacher-centric learning design can improve the utility of new technologies and subsequent acceptance of the use of these technologies to improve the quality of teaching and enhance students learning experience. TAD is one class of tools that can be designed in such a way that will improve teaching practice.

Research on learning analytics has revealed useful insights about students’ learning and the context in which they learn. While the ability to track, harvest and analyse various forms of learning analytics can reveal useful insights about learners’ engagement with learning environments, our review suggests that there is limited focus on analytics relating to the teacher, their teaching approaches and activities. Also, there has been increasing advances in the design of learner and teaching dashboards. However, many teachers still struggle with understanding and interpreting dashboards partly because they lack data literacy skills, and mostly because most the design of many of the tools does not include teachers as partners.

Although, TAD enable teachers to inspect, and understand the processes and progress relating to their teaching, the current implementations of TAD in general, does not adequately provide teachers with the details they need or want in a readily usable format. Educational technology developers can utilise our proposed model to design better tools for improving teaching practice. For example, a TAD can be designed to perform text analytics on students qualitative comments about a course taught, and results presented to the teacher in the form of themes, sentiments and classification; such that it will support the instructor’s needs and preferences for insight generation and reflection.

Teachers monitor, observe and track both teaching and learning activities to make appropriate decisions. Moreover, it is also important to note that visualisations can be misrepresented, misinterpreted or misused by the viewer [122]. Hence, perception and cognition remain a significant challenge in TAD. Consequently, it becomes necessary to design and write algorithms that extract information visualisation, in such a way that allows adequate understanding by teachers. It is also crucial for dashboards to integrate multiple sources such as combining both the learning and teaching activities into a TAD, to create room for teachers to comprehend, reflect on and act upon the presented information quickly.

Also, the current state of technology shows little progress in taking TA, raising concerns about the accurate validity and scalability of innovations such as predictive analytics and TAD. Furthermore, the ethical issues of data use are not considered sufficient to establish institutional policies which incorporate TA as part of quality education models.

Finally, consideration of the framework’s three layers as a whole raises new questions and opportunities. For example, linking educational performance and satisfaction to specific learning design involves consideration of elements of all three layers. This review has shown that TA is a new and essential area of analytics in education. The study also suggests that the conceptualisation of teaching analytics is still at its infancy. However, the practical and successful use of teaching analytics is highly dependent on the development of conceptual and theoretical foundations into consideration.

Implications for practice

This review has uncovered the value of TA and its role in fostering data literacy skills in teachers to support evidence-based teaching. The purpose of TOM is to guide the development of teaching dashboard, and for researchers to develop strategies that help meaningful ways in which data can be presented to teachers. Teacher dashboards can empower the teachers with tools that create new opportunities to make data-informed strategic decisions, utilising the power of analytics and visualisation techniques. Consequently, increasing the efficiency and effectiveness of the institution, including, improving teaching practice, curriculum development and improvement, active learning engagement and improved students’ success. TOM also presents a platform for teaching academics who may have the best understanding of their course contexts, to provide a significant contribution to a culture of data-informed teaching practice within an institution.

The responsibility for managing the systems that provide the analytics usually falls within the control and supervision of the institution’s information technology (IT) department, and often, they have little to no knowledge of their pedagogical applications to teaching and learning. Likewise, academics and their fields of learning support are often deprived of IT skills and have little to no professional understanding of how software systems work. TOM provides opportunities for the teachers to be involved in the design of TA by providing significant interaction and collaboration between the IT and the other sectors that interpret and act upon the information flow.

Additionally, institutions need to provide teaching staff with the necessary training that fosters the development of data literacy skills, and in the use of data and analytical or visualisation dashboards to monitor their teaching practice. Based on some of the challenges identified in the present review, it is imperative institutions ensure that data is collected transparently, with the awareness of all the stakeholders involved, and informed consent of individuals where appropriate. With the advancements in computing technology, data collection, analysis and use have significantly increased, large amounts of data can be continually pulled from different sources and processed at fast speeds. Big data offers institutions the opportunity to implement big data infrastructures and utilise the full potential of data analytics and visualisation. However, institutions also need to consider implementing a data governance framework to guide the implementation and practice of analytics.

The conceptual framework of TA was established to demonstrate the relationship between LA, TA and LD, which can be useful knowledge to various institutional stakeholders, including the learners, teachers, researchers and administrators. However, there are also issues around data ownership, intellectual property rights, and licensing for data re-use (the students, the instructor, the researcher or the institution). For instance, the same data sources can be shared amongst the various stakeholders, but with different level of access, as such data sharing agreement would be needed to guide sharability without infringing on rights, violating privacy or disadvantaging individuals. The implementation of data sharing agreement would require the building of institutional, group as well as individual trust, which would include guidelines on sharing data within the institution and between third parties, such as external organisations and other institutions. In general, stricter data management policies that guide data collection, analysis and use is essential for every institution.

Limitations and future research

Teaching analytics is an emergent phenomenon in the learning analytics and data science literature, with a limited body of published work in the area, as such conclusions drawn from the review are limited to the databases interrogated and articles reviewed. Further, findings in the review are likely to be influenced by our interpretation of the literature and untestable assumptions. For example, linking LA, TA and LD and their underlying assumptions is not grounded in empirical work. The review serves as an advocacy for teacher data literacy and the ability to work with various forms of data. However, working with a single data point may not be publicly accessible to teachers.

Moreover, the combination of analytics on the several data points may lead to some level of identification, and this would require navigating issues around access, protecting privacy, and obtaining appropriate consents. Therefore, it is almost impossible for individual teachers to comprehend not only the scope of data collected, analysed and used but also the consequences of the different layers of collection, analysis and use. Consequently, making it challenging for teachers to make use of the full potentials of data to make informed choices in learning design. No matter how straightforward or transparent institutional policies around data are, the sheer complexity of the collection, analysis and use has made it impossible, posing a fundamental issue for the stakeholders trying to use analytics to enhance teaching practice and learning outcome across an institution.

In future research, we hope to carry out more extensive empirical research on how TOM could be applied to address issues with regards to ethical and privacy concerns about the utilization of TA. We are currently exploring how teaching analytics dashboards can be used to support teacher data literacy and use analytics to improve teaching practice and learning outcome.

Availability of data and materials

Not applicable.

Abbreviations

Academic analytics

Artificial intelligence

Educational data mining

Higher education

Interactive whiteboard

• Learning analytics

Learning design

Learning management system

Machine learning

Massive open online courses

Natural language processing

Open learners model

Student evaluation of teaching

Social network analysis

• Teaching analytics

Teaching analytics dashboard

Term frequency inverse document frequency

• Teaching and learning analytics
• Teaching outcome model

Technology, pedagogy, and content knowledge

Teacher professional development

Adams, M.J., & Umbach, P.D. (2012). Nonresponse and online student evaluations of teaching: Understanding the influence of salience, fatigue, and academic environments. Research in Higher Education , 53 (5), 576–591.

Article   Google Scholar  

Anderson, T. (2003). Getting the mix right again: An updated and theoretical rationale for interaction. The International Review of Research in Open Distributed Learning , 4 (2).

Asare, S., & Daniel, B.K. (2017). Factors influencing response rates in online student evaluation systems: A systematic review approach. In E-Learn: World Conference on E-Learning in Corporate, Government, Healthcare, and Higher Education . Association for the Advancement of Computing in Education (AACE), (pp. 537–541).

Assunção, M.D., Calheiros, R.N., Bianchi, S., Netto, M.A., Buyya, R. (2015). Big data computing and clouds: Trends and future directions. Journal of Parallel and Distributed Computing , 79 , 3–15.

Bakharia, A., Corrin, L., De Barba, P., Kennedy, G., Gašević, D., Mulder, R., Williams, D., Dawson, S., Lockyer, L. (2016). A conceptual framework linking learning design with learning analytics. In Proceedings of the Sixth International Conference on Learning Analytics & Knowledge . ACM, (pp. 329–338).

Banerjee, A., Bandyopadhyay, T., Acharya, P. (2013). Data analytics: Hyped up aspirations or true potential? Vikalpa , 38 (4), 1–12.

Barmaki, R., & Hughes, C.E. (2015). Providing real-time feedback for student teachers in a virtual rehearsal environment. In Proceedings of the 2015 ACM on International Conference on Multimodal Interaction . ACM, (pp. 531–537).

Beer, C., Jones, D., Clark, K. (2009). The indicators project identifying effective learning: Adoption, activity, grades and external factors. In Ascilite . Citeseer.

Benton, S.L., & Cashin, W.E. (2014). Student ratings of instruction in college and university courses , (pp. 279–326): Springer.

Bihani, P., & Patil, S. (2014). A comparative study of data analysis techniques. International Journal of Emerging Trends & Technology in Computer Science , 3 (2), 95–101.

Google Scholar  

Borgman, C.L., Abelson, H., Dirks, L., Johnson, R., Koedinger, K.R., Linn, M.C., Lynch, C.A., Oblinger, D.G., Pea, R.D., Salen, K. (2008). Fostering learning in the networked world: The cyberlearning opportunity and challenge. a 21st century agenda for the national science foundation. https://doi.org/10.1037/e532532011-001 .

Boring, A. (2015). Gender biases in student evaluation of teachers . Paris. https://doi.org/10.1016/j.jpubeco.2016.11.006 .

Boring, A., Ottoboni, K., Stark, P.B. (2016). Student evaluations of teaching are not only unreliable, they are significantly biased against female instructors. Impact of Social Sciences Blog . The London School of Economics and Political Science.

Bos, N., & Brand-Gruwel, S. (2016). Student differences in regulation strategies and their use of learning resources: implications for educational design. In Proceedings of the Sixth International Conference on Learning Analytics & Knowledge . ACM, (pp. 344–353).

Braga, M., Paccagnella, M., Pellizzari, M. (2014). Evaluating students’ evaluations of professors. Economics of Education Review , 41 , 71–88.

Butson, R., & Daniel, B. (2017). The Rise of Big Data and Analytics in Higher Education , (pp. 127–140): Auerbach Publications.

Charleer, S., Klerkx, J., Odriozola, S., Luis, J., Duval, E. (2013). Improving awareness and reflection through collaborative, interctive visualizations of badges. In ARTEL13: Proceedings of the 3rd Workshop on Awareness and Reflection in Technology-enhanced Learning, vol. 1103 . CEUR Workshop Proceedings, (pp. 69–81).

Chatti, M.A., Dyckhoff, A.L., Schroeder, U., Thüs, H. (2013). A reference model for learning analytics. 5-6 , 4 , 318–331.

Chen, L.-L. (2019). Enhancing teaching with effective data mining protocols. Journal of Educational Technology Systems , 47 (4), 500–512.

Chen, Z., Chudzicki, C., Palumbo, D., Alexandron, G., Choi, Y.-J., Zhou, Q., Pritchard, D.E. (2016). Researching for better instructional methods using ab experiments in moocs: results and challenges. Research and Practice in Technology Enhanced Learning , 11 (1), 9.

Choudhury, S., Hobbs, B., Lorie, M., Flores, N. (2002). A framework for evaluating digital library services. D-Lib magazine , 8 (7/8), 1082–9873.

Chounta, I.-A., McLaren, B.M., Albacete, P.L., Jordan, P.W., Katz, S. (2016). Analysis of human-to-human tutorial dialogues: Insights for teaching analytics. In IWTA@ EC-TEL , (pp. 9–17).

Cumbley, R., & Church, P. (2013). Is “big data” creepy? Computer Law & Security Review , 29 (5), 601–609.

Dana, N.F., & Yendol-Hoppey, D. (2019). The Reflective Educator’s Guide to Classroom Research: Learning to Teach and Teaching to Learn Through Practitioner Inquiry : Corwin.

Daniel, B. (2015). Big data and analytics in higher education: Opportunities and challenges. British Journal of Educational Technology , 46 (5), 904–920. https://doi.org/10.1111/bjet.12230 .

Daniel, B., & Harland, T. (2017). Higher Education Research Methodology: A Step-by-Step Guide to the Research Process . Routledge London. https://doi.org/10.4324/9781315149783 .

Daniel, B.K. (2019). Artificial reality: The practice of analytics and big data in educational research. In: Pedersen, J.S., & Wilkinson, A. (Eds.) In Big data: Promise, application and pitfalls . https://doi.org/10.4337/9781788112352.00018 . Edward Elgar, Cheltenham, (pp. 287–300).

Chapter   Google Scholar  

De Freitas, S., Gibson, D., Du Plessis, C., Halloran, P., Williams, E., Ambrose, M., Dunwell, I., Arnab, S. (2015). Foundations of dynamic learning analytics: Using university student data to increase retention. British Journal of Educational Technology , 46 (6), 1175–1188.

Dix, A.J., & Leavesley, J. (2015). Learning analytics for the academic: An action perspective. J. UCS , 21 (1), 48–65.

Ducheva, Z., Pehlivanova, M., Dineva, S. (2013). Possibilities for students to evaluate and improve electronic courses. In The 8th International Conferemnce on Virtual Learning ICVL .

Edström, K. (2008). Doing course evaluation as if learning matters most. Higher Education Research & Development , 27 (2), 95–106.

Elias, T. (2011). Learning analytics. Learning , 1–22.

Ferguson, R. (2012). Learning analytics: drivers, developments and challenges. International Journal of Technology Enhanced Learning , 4 (5/6), 304–317.

Fischer, F., Wild, F., Sutherland, R., Zirn, L. (2014). Grand Challenge Problems from the Alpine Rendez-Vous , (pp. 3–71): Springer.

Flanders, N.A. (1970). Analyzing Teacher Behavior . Addison-Wesley P.C.

Flavin, M. (2017). Disruptive Technology Enhanced Learning: The Use and Misuse of Digital Technologies in Higher Education : Springer. https://doi.org/10.1057/978-1-137-57284-4 .

Gandomi, A., & Haider, M. (2015). Beyond the hype: Big data concepts, methods, and analytics. International Journal of Information Management , 35 (2), 137–144.

Gauthier, G. (2013). Using teaching analytics to inform assessment practices in technology mediated problem solving tasks. In IWTA@ LAK .

Ginon, B., Johnson, M.D., Turker, A., Kickmeier-Rust, M. (2016). Helping Teachers to Help Students by Using an Open Learner Model. https://doi.org/10.1007/978-3-319-45153-4_69 .

Goggins, S.P. (2012). Group informatics: A multi-domain perspective on the development of teaching analytics. In Proceedings of the TaPTA Workshop at EC-TEL .

Goggins, S.P., Galyen, K., Petakovic, E., Laffey, J.M. (2016). Connecting performance to social structure and pedagogy as a pathway to scaling learning analytics in moocs: an exploratory study. Journal of Computer Assisted Learning , 32 (3), 244–266.

Gorham, J. (1988). The relationship between verbal teacher immediacy behaviors and student learning. Communication Education , 37 (1), 40–53.

Griffiths, D. (2017). The use of models in learning design and learning analytics. Interaction Design and Architecture(s) Journal , 33 , 113–133.

Gudivada, V.N., Irfan, M., Fathi, E., Rao, D. (2016). Cognitive analytics: Going beyond big data analytics and machine learning (Vol. 35, pp. 169–205).

Gudivada, V.N., Rao, D.L., Ding, J. (2018). 2. Evolution and Facets of Data Analytics for Educational Data Mining and Learning Analytics , (pp. 16–42). New York. https://doi.org/10.4324/9780203728703-3 .

Harland, T., & Wald, N. (2018). Vanilla teaching as a rational choice: the impact of research and compliance on teacher development. Teaching in Higher Education , 23 (4), 419–434.

Hernández-Leo, D., Martinez-Maldonado, R., Pardo, A., Muñoz-Cristóbal, J.A., Rodríguez-Triana, M.J. (2019). Analytics for learning design: A layered framework and tools. British Journal of Educational Technology , 50 (1), 139–152.

Herodotou, C., Hlosta, M., Boroowa, A., Rienties, B., Zdrahal, Z., Mangafa, C. (2019). Empowering online teachers through predictive learning analytics. British Journal of Educational Technology . https://doi.org/10.1111/bjet.12853 .

Huang, C.-W. (2001). Educlick: A computer-supported formative evaluation system with wireless devices in ordinary classroom. In Proceedings of Int. Conference on Computers in Education, 2010 , (pp. 1462–1469).

Joseph, R.C., & Johnson, N.A. (2013). Big data and transformational government. IT Professional , 15 (6), 43–48.

Kaser, L., & Halbert, J. (2014). Creating and sustaining inquiry spaces for teacher learning and system transformation. European Journal of Education , 49 (2), 206–217.

Kay, D., Korn, N., Oppenheim, C. (2012). Legal, risk and ethical aspects of analytics in higher education. Analytics Series . JISC Cetis (Centre for educational technology and interoperability standards).

KU, O., LIANG, J.-K., CHANG, S.-B., WU, M. (2018). Sokrates teaching analytics system (stas): An automatic teaching behavior analysis system for facilitating teacher professional development. In Proceedings of the 26th International Conference on Computers in Education. Philippines: Asia-Pacific Society for Computers in Education .

Laney, D. (2001). 3d data management: Controlling data volume, velocity and variety. META Group Research Note. META group research note , 6 (70), 1.

Lazar, J., Feng, J.H., Hochheiser, H. (2017). Research Methods in Human-computer Interaction : Morgan Kaufmann.

Leitner, P., Khalil, M., Ebner, M. (2017). Learning analytics in higher education—a literature review , (pp. 1–23): Springer.

Libbrecht, P., Kortenkamp, U., Rebholz, S., Müller, W. (2013). Tales of a companion teacher analytics. In IWTA@ LAK .

Linse, A.R. (2017). Interpreting and using student ratings data: Guidance for faculty serving as administrators and on evaluation committees. Studies in Educational Evaluation , 54 , 94–106.

Loughran, J.J. (2002). Effective reflective practice: In search of meaning in learning about teaching. Journal of teacher education , 53 (1), 33–43.

Macfadyen, L.P., & Dawson, S. (2012). Numbers are not enough. why e-learning analytics failed to inform an institutional strategic plan. Journal of Educational Technology Society , 15 (3).

MacNell, L., Driscoll, A., Hunt, A.N. (2015). What’s in a name: Exposing gender bias in student ratings of teaching. Innovative Higher Education , 40 (4), 291–303.

Mansfield, J. (2019). Pedagogical Equilibrium: The Development of Teachers’ Professional Knowledge : Routledge.

Marlin Jr, J.W., & Niss, J.F. (1980). End-of-course evaluations as indicators of student learning and instructor effectiveness. The Journal of Economic Education , 11 (2), 16–27.

Marsh, H.W. (1987). Students’ evaluations of university teaching: Research findings, methodological issues, and directions for future research. International Journal of Educational Research , 11 (3), 253–388.

McKenney, S., & Mor, Y. (2015). Supporting teachers in data–informed educational design. British journal of educational technology , 46 (2), 265–279.

Menchen-Trevino, E. (2016). Web historian: Enabling multi-method and independent research with real-world web browsing history data. In IConference 2016 Proceedings . https://doi.org/10.9776/16611 .

Michos, K., & Hernández Leo, D. (2016). Towards understanding the potential of teaching analytics within educational communities. In: Vatrapu, R.G.B.B.S., & Kickmeier-Rust, M. (Eds.) In Vatrapu R, Kickmeier-Rust M, Ginon B, Bull S. IWTA 2016 International Workshop on Teaching Analytics. Proceedings of the Fourth International Workshop on Teaching Analytics, in Conjunction with EC-TEL 2016; 2016 Sept 16; Lyon, France.[place Unknown]: CEUR Workshop Proceedings , (pp. 1–8).

Moore, B.L. (2018). 6. The Role of Data Analytics in Education Possibilities and Limitations, 1st edn.https://doi.org/10.4324/9780203728703-8.

Mor, Y., Ferguson, R., Wasson, B. (2015). Learning design, teacher inquiry into student learning and learning analytics: A call for action. British Journal of Educational Technology , 46 (2), 221–229.

Müller, W., Rebholz, S., Libbrecht, P. (2016). Automatic inspection of e-portfolios for improving formative and summative assessment. In International Symposium on Emerging Technologies for Education . Springer, (pp. 480–489).

Norris, D., Baer, L., Leonard, J., Pugliese, L., Lefrere, P. (2008). Action analytics: Measuring and improving performance that matters in higher education. EDUCAUSE Review , 43 (1), 42.

Olson, D.L., & Lauhoff, G. (2019). Descriptive data mining , (pp. 129–130): Springer.

Osterman, K.F., & Kottkamp, R.B. (1993). Reflective Practice for Educators: Improving Schooling Through Professional Development : ERIC.

Pantazos, K., & Vatrapu, R. (2016). Enhancing the professional vision of teachers: A physiological study of teaching analytics dashboards of students’ repertory grid exercises in business education. In System Sciences (HICSS), 2016 49th Hawaii International Conference On . IEEE, (pp. 41–50).

Pantazos, K., Vatrapu, R.K., Hussain, A. (2013). Visualizing repertory grid data for formative assessment. In IWTA@ LAK .

Papamitsiou, Z., & Economides, A.A. (2016). Learning analytics for smart learning environments: A meta-analysis of empirical research results from 2009 to 2015. Learning, Design, and Technology: An International Compendium of Theory, Research, Practice, and Policy , 1–23. https://doi.org/10.1007/978-3-319-17727-4_15-1 .

Pardo, A., & Siemens, G. (2014). Ethical and privacy principles for learning analytics. British Journal of Educational Technology , 45 (3), 438–450.

Pascual-Miguel, F., Chaparro-Pelaez, J., Hernandez-Garcia, A., Iglesias-Pradas, S. (2011). A characterisation of passive and active interactions and their influence on students’ achievement using moodle lms logs. International Journal of Technology Enhanced Learning , 3 (4), 403–414.

Pennings, H.J., Brekelmans, M., Wubbels, T., van der Want, A.C., Claessens, L.C., van Tartwijk, J. (2014). A nonlinear dynamical systems approach to real-time teacher behavior: Differences between teachers. Nonlinear Dynamics, Psychology, and Life Sciences , 18 (1), 23–45.

Persico, D., & Pozzi, F. (2015). Informing learning design with learning analytics to improve teacher inquiry. British Journal of Educational Technology , 46 (2), 230–248.

Pozzi, F., & Persico, D. (2013). Sustaining learning design and pedagogical planning in cscl. Research in Learning Technology , 21 . https://doi.org/10.3402/rlt.v21i0.17585 .

Prieto, L.P., Magnuson, P., Dillenbourg, P., Saar, M. (2017). Reflection for action: Designing tools to support teacher reflection on everyday evidence. https://doi.org/10.31219/osf.io/bj2rp .

Prieto, L.P., Rodriguez-Triana, M.J., Kusmin, M., Laanpere, M. (2017). Smart school multimodal dataset and challenges. In Joint Proceedings of the Sixth Multimodal Learning Analytics (MMLA) Workshop and the Second Cross-LAK Workshop Co-located with 7th International Learning Analytics and Knowledge Conference, vol. 1828 . CEUR, (pp. 53–59).

Prieto, L.P., Sharma, K., Dillenbourg, P., Jesús, M. (2016). Teaching analytics: towards automatic extraction of orchestration graphs using wearable sensors. In Proceedings of the Sixth International Conference on Learning Analytics & Knowledge . ACM, (pp. 148–157).

Prieto, L.P., Sharma, K., Kidzinski, Ł., Rodríguez-Triana, M.J., Dillenbourg, P. (2018). Multimodal teaching analytics: Automated extraction of orchestration graphs from wearable sensor data. Journal of Computer Assisted Learning , 34 (2), 193–203.

Ramos, C., & Yudko, E. (2008). "hits"(not "discussion posts") predict student success in online courses: a double cross-validation study. Computers & Education , 50 (4), 1174–1182.

Rayward-Smith, V.J. (2007). Statistics to measure correlation for data mining applications. Computational Statistics & Data Analysis , 51 (8), 3968–3982.

Article   MathSciNet   MATH   Google Scholar  

Remmers, H.H., & Brandenburg, G. (1927). Experimental data on the purdue rating scale for instructors. Educational Administration and Supervision , 13 (6), 399–406.

Rienties, B., Boroowa, A., Cross, S., Farrington-Flint, L., Herodotou, C., Prescott, L., Mayles, K., Olney, T., Toetenel, L., Woodthorpe, J. (2016). Reviewing three case-studies of learning analytics interventions at the open university uk. In Proceedings of the Sixth International Conference on Learning Analytics & Knowledge . ACM, (pp. 534–535).

Roberts, L.D., Chang, V., Gibson, D. (2017). Ethical considerations in adopting a university-and system-wide approach to data and learning analytics , (pp. 89–108): Springer. https://doi.org/10.1007/978-3-319-06520-5_7 .

Romero, C., & Ventura, S. (2010). Educational data mining: a review of the state of the art. IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews) , 40 (6), 601–618.

Saar, M., Kusmin, M., Laanpere, M., Prieto, L.P., Rüütmann, T. (2017). Work in progress–semantic annotations and teaching analytics on lecture videos in engineering education. In Global Engineering Education Conference (EDUCON), 2017 IEEE . IEEE, (pp. 1548–1551).

Saar, M., Prieto, L.P., Rodríguez-Triana, M.J., Kusmin, M. (2018). Personalized, teacher-driven in-action data collection: technology design principles. In 2018 IEEE 18th International Conference on Advanced Learning Technologies (ICALT) . IEEE, (pp. 58–62).

Saric, M., & Steh, B. (2017). Critical reflection in the professional development of teachers: Challenges and possibilities. CEPS Journal , 7 (3), 67–85.

Saye, J.W., & Brush, T. (2007). Using technology-enhanced learning environments to support problem-based historical inquiry in secondary school classrooms. Theory Research in Social Education , 35 (2), 196–230.

Schempp, P., McCullick, B., Pierre, P.S., Woorons, S., You, J., Clark, B. (2004). Expert golf instructors’ student-teacher interaction patterns. Research Quarterly for Exercise and Sport , 75 (1), 60–70.

Schmidlin, K., Clough-Gorr, K.M., Spoerri, A. (2015). Privacy preserving probabilistic record linkage (p3rl): a novel method for linking existing health-related data and maintaining participant confidentiality. BMC Medical Research Methodology , 15 (1), 46.

Sergis, S., & Sampson, D.G. (2016). Towards a teaching analytics tool for supporting reflective educational (re) design in inquiry-based stem education. In Advanced Learning Technologies (ICALT), 2016 IEEE 16th International Conference On . https://doi.org/10.1109/icalt.2016.134 . IEEE, (pp. 314–318).

Sergis, S., & Sampson, D.G. (2017). Teaching and learning analytics to support teacher inquiry: A systematic literature review , (pp. 25–63): Springer. https://doi.org/10.1007/978-3-319-52977-6_2 .

Sergis, S., Sampson, D.G., Rodríguez-Triana, M.J., Gillet, D., Pelliccione, L., de Jong, T. (2017). Using educational data from teaching and learning to inform teachers’ reflective educational design in inquiry-based stem education. Computers in Human Behavior . https://doi.org/10.1016/j.chb.2017.12.014 .

Shen, J., Chen, H., Jiang, J. (2018). A research on techniques for data fusion and analysis of cross-platform mooc data. In 2018 17th International Conference on Information Technology Based Higher Education and Training (ITHET) . IEEE, (pp. 1–8).

Siemens, G. (2013). Learning analytics: The emergence of a discipline. American Behavioral Scientist , 57 (10), 1380–1400.

Stier, S., Breuer, J., Siegers, P., Thorson, K. (2019). Integrating Survey Data and Digital Trace Data: Key Issues in Developing an Emerging Field. https://doi.org/10.1177/0894439319843669 .

Subramanya, S. (2014). Toward a more effective and useful end-of-course evaluation scheme. Journal of Research in Innovative Teaching , 7 (1).

Suehiro, D., Taniguchi, Y., Shimada, A., Ogata, H. (2017). Face-to-face teaching analytics: Extracting teaching activities from e-book logs via time-series analysis. In Advanced Learning Technologies (ICALT), 2017 IEEE 17th International Conference On . IEEE, (pp. 267–268).

Sun, J., Przybylski, R., Johnson, B.J. (2016). A review of research on teachers’ use of student data: From the perspective of school leadership. Educational Assessment, Evaluation and Accountability , 28 (1), 5–33.

Taniguchi, Y., Suehiro, D., Shimada, A., Ogata, H. (2017). Revealing hidden impression topics in students’journals based on nonnegative matrix factorization. In Advanced Learning Technologies (ICALT), 2017 IEEE 17th International Conference On . IEEE, (pp. 298–300).

Thille, C., & Zimmaro, D. (2017). Incorporating learning analytics in the classroom. New Directions for Higher Education , 2017 (179), 19–31.

Thomas, C. (2018). Multimodal teaching and learning analytics for classroom and online educational settings. In Proceedings of the 2018 on International Conference on Multimodal Interaction . ACM, (pp. 542–545).

ur Rehman, M.H., Chang, V., Batool, A., Wah, T.Y. (2016). Big data reduction framework for value creation in sustainable enterprises. International Journal of Information Management , 36 (6), 917–928.

Van Harmelen, M., & Workman, D. (2012). Analytics for learning and teaching. CETIS Analytics Series , 1 (3), 1–40.

van Leeuwen, A., Rummel, N., van Gog, T. (2019). What information should cscl teacher dashboards provide to help teachers interpret cscl situations? International Journal of Computer-Supported Collaborative Learning , 1–29. https://doi.org/10.1007/s11412-019-09299-x .

Vatrapu, R.K. (2012). Towards semiology of teaching analytics. In Workshop Towards Theory and Practice of Teaching Analytics, at the European Conference on Technology Enhanced Learning, TAPTA, vol. 12 .

Vatrapu, R.K., Kocherla, K., Pantazos, K. (2013). iklassroom: Real-time, real-place teaching analytics. In IWTA@ LAK .

Vatrapu, R., Reimann, P., Bull, S., Johnson, M. (2013). An eye-tracking study of notational, informational, and emotional aspects of learning analytics representations. In ACM International Conference Proceeding Series . https://doi.org/10.1145/2460296.2460321 . https://www.scopus.com/inward/record.uri?eid=2-s2.0-84876499638&doi=10.1145%2f2460296.2460321&partnerID=40&md5=e7b4d83a3e33e7a1c3c5b5f56d5ebe7d , (pp. 125–134).

Vatrapu, R., Reimann, P., Hussain, A., Kocherla, K. (2013). Towards teaching analytics: Repertory grids for formative assessment (rgfa). In CSCL 2013 Conference Proceedings, vol 2 , (pp. 422–426).

Vatrapu, R., Tanveer, U., Hussain, A. (2012). Towards teaching analytics: communication and negotiation tool (coneto). In Proceedings of the 7th Nordic Conference on Human-Computer Interaction: Making Sense Through Design . ACM, (pp. 775–776).

Vatrapu, R., Teplovs, C., Fujita, N., Bull, S. (2011). Towards visual analytics for teachers’ dynamic diagnostic pedagogical decision-making. In Proceedings of the 1st International Conference on Learning Analytics and Knowledge . ACM, (pp. 93–98).

Verbert, K., Duval, E., Klerkx, J., Govaerts, S., Santos, J.L. (2013). Learning analytics dashboard applications. American Behavioral Scientist , 57 (10), 1500–1509.

Voithofer, R., & Golan, A.M. (2018). 5. Data Sources for Educators, 1st edn.https://doi.org/10.4324/9780203728703-7, (p. 18).

Waller, M.A., & Fawcett, S.E. (2013). Data science, predictive analytics, and big data: a revolution that will transform supply chain design and management. Journal of Business Logistics , 34 (2), 77–84.

Wang, F., & Hannafin, M.J. (2005). Design-based research and technology-enhanced learning environments. Educational Technology Research and Development , 53 (4), 5–23.

Williamson, B. (2016). Digital education governance: data visualization, predictive analytics, and ’real-time’ policy instruments. Journal of Education Policy , 31 (2), 123–141.

Winkler, R., & Söllner, M. (2018). Unleashing the potential of chatbots in education: A state-of-the-art analysis. In Academy of Management Annual Meeting (AOM) . https://doi.org/10.5465/ambpp.2018.15903abstract .

Xu, B., & Recker, M. (2012). Teaching analytics: A clustering and triangulation study of digital library user data. Educational Technology & Society , 15 (3), 103–115.

Yigitbasioglu, O.M., & Velcu, O. (2012). A review of dashboards in performance management: Implications for design and research. International Journal of Accounting Information Systems , 13 (1), 41–59.

Download references

Acknowledgements

The research reported is part of an ongoing PhD research study in the area of Big Data Analytics in Higher Education. We also want to thank members of the Technology Enhanced Learning and Teaching (TELT) Committee of the University of Otago, New Zealand for support and for providing constructive feedback.

This research project was fully sponsored by Higher Education Development Centre, University of Otago, New Zealand.

Author information

Authors and affiliations.

Higher Education Development Centre, University of Otago, Dunedin, New Zealand

Ifeanyi Glory Ndukwe & Ben Kei Daniel

You can also search for this author in PubMed   Google Scholar

Contributions

IGN conceived and presented the Conceptualisation of Teaching Analytics and Teachingv Outcome Model. BKD developed the Tripartite Approach that was utilised in this research. BKD encouraged IGN to perform a systematic review of teaching analytics that was guided by the Tripartite Approach. BKD supervised the findings of this work. IGN took the lead in writing the manuscript. All authors discussed the results, provided critical feedback and contributed to the final manuscript.

Corresponding author

Correspondence to Ifeanyi Glory Ndukwe .

Ethics declarations

Competing interests.

The authors declare that they have no competing interests. All authors have approved the manuscript and agree with its submission to the International Journal of Education Technology in Higher Education. This manuscript has not been published and is not under consideration for publication elsewhere.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Ndukwe, I.G., Daniel, B.K. Teaching analytics, value and tools for teacher data literacy: a systematic and tripartite approach. Int J Educ Technol High Educ 17 , 22 (2020). https://doi.org/10.1186/s41239-020-00201-6

Download citation

Received : 29 October 2019

Accepted : 01 April 2020

Published : 22 June 2020

DOI : https://doi.org/10.1186/s41239-020-00201-6

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

The Different Types of Data in Education: A Complete Guide

As the field of education continues to evolve, it’s become increasingly clear that data plays a vital role in shaping how we approach instruction. But not all data is created equal. Without a comprehensive understanding of the different types of data in education, you could end up missing crucial pieces of information that might be putting your students—and school—at a disadvantage.

In this blog post, we’ll explore some of the most common types of educational data—and what they can tell you about student performance. By the end of this blog post, you’ll have a strong grip on which types of data in education you should be analyzing to drive success in your organization.

Here’s what we’ll cover:

• What is data in education?

Demographic Information

Academic performance, behavioral records, engagement indicators.

• Which types of data should you collect?

1.  What is data in education?

Data in education refers to quantitative and qualitative information collected from students, teachers, parents, administrators, or other stakeholders that help inform decision-making within an educational setting.

In today’s rapidly evolving educational landscape, understanding the different types of data in education is crucial to stay ahead of the curve. Advances in tech and data analysis tools have made it easier to gather and analyze educational data—which school leaders can use to better understand individual students’ needs and develop targeted strategies to enhance student achievement.

Most types of data in education fall into four main categories:

• Demographic information , which covers student background details.
• Academic performance , which includes test scores and grades.
• Behavioral records , which refer to attendance and discipline incidents.
• Engagement indicators , which measure participation in school activities.

In addition to the above four types of data in education, you also have qualitative and quantitative data.

Quantitative data uses statistical and mathematical analysis to measure variables (i.e., student achievement, attendance, and demographic characteristics).  Educational quantitative data is collected through numerical methods, like surveys with closed-ended questions, standardized tests, and administrative records.

Qualitative data , on the other hand, is all about understanding nuances and complexities through in-depth exploration. Qualitative educational data is collected through non-numerical methods, like observations, interviews, focus groups, surveys with open-ended questions, case studies, and document analysis. This type of data can help you interpret the experiences, attitudes, behaviors, and perceptions of both students and teachers.

2.  Types of Educational Data

Now we’ve got a handle on what data in education is, why it’s important, and a high-level overview of what falls under the umbrella of educational data—let’s take a closer look at the different types of data in education, and what they can tell us about school performance.

Demographic information , such as age, gender, race/ethnicity, socioeconomic status, and language proficiency, is essential for understanding the diverse needs of individual students. School staff can effectively tailor their teaching methods by analyzing demographic data to better serve all students.

Demographic information includes:

• Age , to identify age-related patterns and trends in education.
• Gender , including the socially constructed roles, behaviors, and identities associated with gender—and how that influences access to educational opportunities and outcomes.
• Race/Ethnicity: The social and cultural groups a student belongs to based on their ancestry, heritage, or cultural traditions. Race/ethnicity can influence educational experiences, opportunities, and outcomes.
• Socioeconomic Status (SES): Students’ economic and social conditions, including household income, occupation, education, and social status. SES can affect access to resources and opportunities that impact educational outcomes.
• Language: The spoken and written communication systems used by students, including their primary language and proficiency in other languages. Language fluency and comprehension levels can impact access to educational opportunities and outcomes.
• Disability Status: The physical, sensory, intellectual, or emotional impairments that impact a student’s ability to engage in educational activities. Disability status can affect access to accommodations and services that support academic success.
• Geographic Location: The physical location of a student’s home or community, which can influence access to educational resources and opportunities, as well as cultural and social factors that impact educational outcomes.

Academic performance data measures students’ academic achievement and progress across different skills and subjects. Tracking educational performance data means educators can better allocate resources to improve student outcomes and meet accountability requirements.

Here are some examples of academic performance data:

• Grades: The scores or letters that students receive on assignments, tests, and courses, which reflect their level of understanding and mastery of the subject matter.
• Standardized Test Scores: The scores students receive on standardized tests, such as the SAT, ACT, or state assessments, which measure their performance in specific content areas and compare their performance to other students.
• Course Completion Rates: The percentage of students who complete a course or program, which measures their persistence and success in meeting academic requirements.
• Attendance Rates: The percentage of time that students are present in school or class, which indicates their engagement and participation in the learning process. Attendance records allow schools to identify patterns requiring intervention or support services.
• Graduation Rates: The percentage of students who complete their schooling within a specified timeframe, which measures their overall academic achievement.
• Progress Monitoring Data: Data collected over time to track students’ growth and progress in specific skills or areas, such as reading fluency or math computation, highlighting their strengths and areas for improvement.
• Teacher Observations and Evaluations: The data that teachers collect and report on students’ performance in different classroom activities, like participation, behavior, and homework completion, which provides a qualitative indication of their academic performance.

Behavioral data in education refers to information that documents students’ behavior in and out of the classroom. Behavioral records data can provide insights into students’ behavioral patterns and needs, informing the development of interventions and support services. With this information, school leaders can implement training to support teachers in effectively managing classroom behavior.

Here are some examples of different educational data types that fall under behavioral records.

• Discipline Incidents: Data documenting a student’s infractions and disciplinary actions taken in response.
• Suspensions and Expulsions: The number of documented times a student has been removed from school due to behavioral issues.
• Behavior Checklists: Data documenting a student’s behavior using a standardized checklist.
• Peer and Parent Surveys: Documentations of a student’s behavior reported by peers or parents.
• Counseling and Mental Health Records: Qualitative documentation of a student’s counseling or mental health visits.

Engagement indicators provide insights into how actively and meaningfully students participate in learning. When educators understand students’ motivation, interest, and investment in their learning individually, they can tailor instruction and support services for maximum impact.

Here are some examples of different types of engagement indicators:

• Class Participation: The degree to which students contribute to class discussions and activities—measured using participation rubrics or documented observations.
• Homework Completion: The percentage of assigned homework students complete, indicating their engagement with the subject matter outside the classroom.
• Learning Time: The amount of time students spend engaged in learning activities, including classroom instruction, homework, and independent study.
• Assessment Performance: The degree to which students demonstrate understanding and mastery of content on quizzes, exams, or other assessments.
• Technology Usage: The extent to which students use technology to support their learning, including online resources, interactive tools, and learning management systems.
• Extracurricular Involvement: Participation in school clubs, organizations, sports, and other extracurricular activities.
• Social and Emotional Learning: The development of student’s social and emotional competencies, including self-awareness, self-management, social awareness, relationship skills, and responsible decision-making.
• Mobility Rates: How often students change schools during a given period, so school administrators can address whether lack of stability affects their engagement.

3.  Which types of data in education should you collect?

Now that we’ve explored the different types of data in education, the ultimate question remains: Which types of data should you be collecting and analyzing?

Of course, there’s no right or wrong answer to this question. It depends entirely on your organization’s goals and the challenges your school or educational institution currently faces. All types of educational data are valuable in their own right, and understanding their differences can help you make more informed decisions about how best to support your students.

The wider the range of data you collect, the better your outcomes will be. For example, qualitative data—like mental health records and emotional learning—can provide rich and detailed insights into the complex social and cultural aspects of school life that quantitative data (i.e., test scores and attendance) alone can’t capture. A diverse data set can help educators understand how students construct their knowledge, how teachers make decisions in the classroom, how schools create a sense of community, and how educational policies are implemented and experienced.

The bottom line? By gathering and analyzing both qualitative and quantitative data across all four categories of educational data, you can gain an in-depth insight into your student’s needs, strengths, and areas for growth—ultimately helping them succeed in the classroom and beyond.

Want to learn more about how data can supercharge your school’s success? You’re in the right place! Check out our Inno™ page to learn more.

Thank you for sharing!

You may also be interested in

Why interactive data visualization is the key to better student outcomes.

Discover the power of interactive data visualization for student data with this in-depth guide.

by Innovare | Dec 4, 2023 | Data in K-12 Schools

Data in K-12 Schools

What is Data in Education? The Ultimate Guide

Discover the power of data in education with our comprehensive guide.

by Innovare | Sep 18, 2023 | Data in K-12 Schools

This website uses cookies to improve your experience. See our Privacy Policy to learn more. Accept

DIGITAL TOOLS FOR REAL-TIME DATA COLLECTION IN EDUCATION

By Emily Gustafsson-Wright, Sarah Osborne, and Muskan Aggarwal

Real-time performance data in education enable critically needed tracking of program activities and instructors’ and learners’ progress to better inform adaptations along the way. In this report, we provide an overview of some of the key considerations related to real-time data collection in education, with a particular focus on the digital tools that enable their collection. Importantly, we have developed a typology of tools and selection criteria to support policymakers, practitioners, and researchers around the globe in either developing new tools or selecting from the landscape of existing ones. Our aim is to help initiate dialogue around the use of real-time data for adaptive management in education and contribute to data-informed decisions toward ensuring that all children have access to quality teaching and learning experiences.

Introduction

The availability of—and access to—real-time performance data from classrooms and other learning spaces will be critical in addressing the increasingly dire learning crisis facing today’s children and youth. Real-time performance data focus on the ongoing tracking of program activities and progress of instructors and learners in order to better inform adaptations to program inputs, activities, and outputs along the way. Administrators (such as school or government leaders or program facilitators) can use information on performance to adaptively manage resources and tailor programs according to the needs of students and instructors. Similarly, instructors can use student data to ensure that they are teaching at the right level and customize their instruction based on individual learning patterns and needs. Students and families can use real-time data to track progress and use this information to advocate for their needs.

Collecting data in real time is inherently challenging. Paper-based systems of data collection can be slow, administratively burdensome, and prone to human error. Digital technologies potentially offer more efficient collection and analysis of real-time data, allow for more flexibility and customizability, and can provide functionalities such as automatically generated visualizations and ongoing recommendations.

In this report, we provide an overview of the critical factors related to real-time data collection in education to support policymakers, practitioners, and researchers across the globe in either developing new tools or selecting from the landscape of existing ones. After providing brief background on the state of data and measurement in the global education sector, we explore the who, why, and for whom of real-time data collection and use in education systems. Next, we focus our attention on one part of the puzzle—digital tools for data collection. Informed by survey responses from the developers of over 20 such tools, we classify tools based on a typology of their purpose. We elaborate on important factors such as tool usability, functions, and context, which can help decisionmakers select the most appropriate tool for their setting (potentially reducing the need to develop new ones where they already exist), or inform the design of a new tool. These sections are accompanied by an interactive database of digital tools for real-time education data. We conclude with key findings and policy recommendations.

Tackling the learning crisis is made more difficult due to the global data gap on learning outcomes.

The purpose of this report is not to assess the impact that a particular tool or type of tool may have on learning outcomes. Rather, the value-add of the report and the framework presented within—as the first of their kind—is to help initiate dialogue around the use of real-time data for adaptive management in education and to scope the wide array of digital tools for real-time data collection currently available in the education sector. This will help address the complex challenges faced in the education ecosystem and is just the beginning of an iterative process of action-oriented research. As such, new criteria and dimensions will likely be added as this work evolves—along with additional tools—with the help of crowdsourcing from the education community.

Tracking learning progress

Over the past decade, there has been a shift in the discourse within the global education sector from one of access to one of access plus learning. The “Global Compact on Learning” report (Perlman Robinson, 2011) called for, among other things, an emphasis on measurement as one of the six core principles to achieve learning for all. This jumpstarted a discussion that led to the establishment of the Learning Metrics Taskforce that was instrumental in catalyzing a shift in the global education conversation and building consensus on global learning indicators and actions to improve the measurement of learning in all countries (UNESCO & Center for Universal Education, 2013). The taskforce identified and recommended a set of key skills and competencies that all children should learn globally and developed a framework of indicators to track this learning at the global level. Finally, with the codification of child learning outcomes in Sustainable Development Goal (SDG) 4, quality of schooling and learning became the central focus of all global education targets, representing a significant advancement from the Millennium Development Goals’ focus on access to education. Furthermore, SDG 4 expands the scope of what is considered a “learning outcome” to include not only literacy and numeracy but also socio-emotional “skills needed to promote global citizenship and sustainable development” (Target 4.7).

Recent advancements in measuring learning and establishing comparability have been instrumental in the development of global indicators that may be used to evaluate learning systems and students’ learning levels across countries. When the World Bank and UNESCO introduced the Learning Poverty indicator in October 2019, it became even more evident that the world was in the throes of a global learning crisis. The metrics indicated that 53 percent of all children in low- and middle-income countries were not attaining the minimum reading proficiency by the age of 10 or by the end of primary school—and given the slow pace of progress—it would be impossible to reduce learning poverty rates to zero by 2030 (World Bank, 2021; Azevedo, 2020).

The COVID-19 pandemic’s unparalleled effect on global education—and its impact on learning has been particularly severe among marginalized populations—has only deepened the global learning crisis. In terms of learning poverty levels, low- and middle-income countries may potentially see a sharp increase of up to 70 percent. While the costs and magnitude of the crisis are yet to be fully realized, it is estimated that this current generation of students could lose up to $21 trillion in lifetime earnings (in present value) due to the learning loss—a significant increase from the projected $10 trillion loss pre-COVID (World Bank, UNICEF, & UNESCO, 2022).

Tackling the learning crisis is made more difficult due to the global data gap on learning outcomes. Many countries lack sufficient data and/or the capacity to measure and monitor learning outcomes, which is impeding the advancement of evidence-based policies (UNESCO & CUE, 2013). While substantial losses in both reading and math skills have been documented across countries to varying extents, data on learning loss—not to mention data on socio-emotional well-being—remain scarce (World Bank, UNICEF & UNESCO, 2021). This makes it more difficult to estimate the full scale of the crisis and to provide a precise estimate of students’ learning levels and well-being. Robust data to inform policymaking, track progress, and hold policymakers, practitioners, and funders accountable are crucial. A report from the Center for Global Development posits that institutionalized evaluation of interventions can act as an “economic safeguard” that can ensure that only effective interventions are scaled (Kaufman et al., 2022).

While it is true that measurement of outcomes is essential to improving learning, measurement should also be in a form that makes it actionable, at the right frequency, and accessible. Data must be usable through a range of tools serving different purposes and by various stakeholders—for example, ranging from formative classroom assessments to inform teacher instruction, to national and international levels of assessment for policy prioritization and standards benchmarking. Lack of systematic measurement and data collection not only makes it difficult to assess current learning levels, but also makes it hard to know which actions are making a difference, and where and how to promote action.

Real-time data

Real-time data for development.

With the growing emphasis on measurement in education systems, the education and international development communities in recent years have placed greater emphasis on continual data collection that can support data-driven adaptive management of programs. In contrast to final program or intervention evaluation data (which saw a boom in the early 2000s with the randomized controlled trial movement), real-time data are gathered along the way and analyzed quickly, which enable timely decisions and adaptation. In multiyear social programs or policy efforts, meaningful results for program participants hinge on understanding what is and is not working—and for whom—during the cycle of implementation. This allows improvements to be made, as opposed to after the program or policy cycle has ended, when it is too late to make changes.

Real-time data are gathered along the way and analyzed quickly, which enable timely decisions and adaptation.

A range of efforts are underway to raise the profile and rigor of real-time data collection and use across the field of international education and development more broadly. For example, a consortium led by Results for Development is building evidence on adaptive program implementation through the rapid feedback monitoring, evaluation, research, and learning as part of a collaboration with USAID (Benson et al., 2019). Also at USAID, “adaptive management” is one of the core principles of the program cycle and is understood as “an intentional approach to making decisions and adjustments in response to new information and changes in context.” Notably, adaptive management practices are much more likely to thrive with the right enabling conditions: culture, processes, and resources. As a result, considering the context’s enabling conditions and working to address the “disabling” elements is critical for adaptive management adoption (USAID, 2018).

A rise in interest in outcomes- and results-based financing (including social and development impact bonds and outcomes funds) among policymakers, funders, and practitioners has drawn further attention to real-time data and adaptive management. The focus on results within this form of innovative finance necessitates the collection of data in real-time to inform feedback loops for better service delivery to achieve the outcomes on which the contracts are based (Tomkinson, 2015). Establishing and refining systems of adaptive management has thus become a core task for service providers engaging in outcomes-based contracting, including improving data management capacity and employing technological solutions to enable efficient data use (World Bank, 2014).

This adaptive management approach prioritizes “learning by doing” (Ramalingam et al, 2017), but by firmly relying on systematically collected data. Such adjustments can be either tactical⁠—that is, targeting immediate operational processes⁠—or strategic—addressing systemic shortcomings in the mechanics of the program itself. One well-cited example of this is the Educate Girls Development Impact Bond (Gustafsson-Wright, Osborne, & Massey, 2020), where a strong system of adaptive management was developed within the project, allowing for timely course correction and ultimately a positive impact on student outcomes.

Real-time data for education

Within the education sector, the conversations about timely data provision are centered around the focus on equitable learning outcomes and as part of a broader movement toward data-driven decisionmaking at all levels of the education system. In fact, many education decisionmakers have highlighted the need for more timely and accessible data. For example, in a 2017 survey of education leaders from around the world, timeliness and accessibility were ranked as the top two areas in need of improvements to make data in the education sector more helpful (Custer et al., 2018).

We have seen a response to this need at the global policy level. As a follow-up to the 2018 World Development Report (World Bank, 2018) recommendations, for example, the World Bank—supported by the Bill and Melinda Gates Foundation and the U.K. Department for International Development—announced the establishment of the Global Education Policy Dashboard, which aims to provide policymakers with actionable data on indicators associated with learning outcomes (inputs and infrastructure, teaching, learners, school management, and learning poverty) in basic education (World Bank, 2019).

This and similar types of policy responses are critical because real-time data have the potential to achieve equitable positive learning outcomes by facilitating timely decisionmaking and course corrections and represent an important shift from program monitoring data, which primarily track how much an education program’s activities and outputs align with intended milestones, as well as possible unintended consequences. While program monitoring data are important, they rely on critical assumptions about the links among activities, outputs, and the proposed final outcomes, which may or may not occur. Real-time performance data, on the other hand, track ongoing program activities and instructors’ and learners’ progress to better inform data-driven adaptations to program inputs, activities, and outputs along the way (Black & Wiliam, 2008). In this way, real-time data can inform the design and management of a variety of education programs and initiatives (GIZ, 2017).

As shown in Table 1, there are a variety of reasons to collect real-time data in education, which correlate with the types of data collected and the users of the data and tools. The collection of these data can encompass a variety of domains, including organization of instruction, personnel management, resource management, and administrative planning. Across these domains, data are primarily used for tracking and responding to student and instructor behavior and performance with the goal of ensuring student learning. Broadly speaking, the types of real-time data needed to achieve these goals can include: child (or student) attendance and enrollment, child (or student) performance data (or assessment), instructor (or other service provider) attendance, and instructor (or other service provider) performance, which can be self- or objective assessment. These data can be collected and used by many actors in the education system, ranging from children and their parents, to instructors, coaches, and school administrators, to high-level government administrators.

TABLE 1. A taxonomy of real-time education data

Assessing student performance can take multiple forms—for example, conducting examinations to track individual student progress and inform learning strategies. Formative classroom assessments (for example periodic quizzes, assignments, or learning checks) are one approach to track student progress toward learning, provide immediate feedback to inform classroom instruction, and guide teacher professional development. Data can also include detailed information on students’ social and behavioral experiences, which can help instructors and students build stronger relationships and change the way children learn. Just as formative assessment can be used to improve students’ learning, tracking data on educator performance can be used to inform professional learning and development (Dillaha & Haren, 2017).

National and international level large-scale summative assessments are instrumental as monitoring and evaluation strategies and provide feedback and data on general trends and the state of the education system (World Bank, 2020; World Bank, 2019).

For any data to be effective at improving teaching and learning outcomes, they must reach the proper personnel, or those who can make and enforce decisions and take action. As shown in Box 2, each education decisionmaker has a different role in determining the ultimate outcomes of student learning—from the national or ministerial level and local government to school administrators to classroom or non-center-based instructors/coaches, to children and their parents/families. Furthermore, for data-informed decisionmaking to take place, it is not enough for stakeholders to have access to the data—they also need to have the capacity, time, and authority to collect and use it. The following section discusses some of the factors that facilitate real-time data collection and use, as well as some of the potential pitfalls.

Real-time data are valuable at all stages of the program lifecycle from design to implementation, and their analysis can drive both immediate operational adaptations at the frontline of educational instruction by teachers, school administrators, and mentors, and at high-level tactical adaptations at a systemic level (USAID, 2017).

Broadly, government administrators (in education ministries and subnational ministries) use education data for policy design and strategic planning. This may include monitoring national, subnational, or school-level attendance, attainment, and achievement levels, as well as measures of equity. This information feeds into decisions about allocation of resources and development and revision of standards and goals. It may also include decisions around the employment and training of administrators, teachers, and other staff.

At the school administrator level, education data are used to track, evaluate, and support student, teacher, and staff performance and progress. These data can also be used to develop school action plans and guide-school level practices.

Instructors and coaches use education data to assess the performance, progress, and needs of students in order to develop and revise classroom instruction. These data can come through a variety of forms including formative and summative assessments.

Source: Author adaptation based on Gill et al. 2014 (Mathematica).

Principles and pitfalls of real-time data collection and use

The academic literature identifies a number of principles that can help ensure accuracy, efficiency, effectiveness, and equity in the collection and use of data and that stakeholders are employing on the ground. The Right-Fit Evidence unit at Innovations for Poverty Action, for instance, works to implement the CART principles identified by Gugerty and Karlan (2018) to support data collection for monitoring and evaluation. According to these principles, data should be credible and actionable, data collection should be responsible, and findings should be transportable (IPA, n.d.). These principles also feed into the “nimble evaluation” approach—whereby data are quickly collected on short-term outcomes (Karlan, 2017)—which the Strategic Impact Evaluation Fund (SIEF) at the World Bank is currently testing (World Bank, 2018). USAID also outlines five key standards to improve the consistency and accuracy of collected data. These include validity of data, integrity of the data collection process, precision (or a small margin of error on performance indicators), reliability or rigor in analysis, and timeliness in using data in decisionmaking.

Some potential pitfalls are also highlighted in the literature. Specific to the education sector, for instance, Pritchett (2018) argues that efforts to collect real-time data must be grounded in a cohesive theory of change framework to avoid being overwhelmed with unstructured data and potentially lose sight of the larger picture of child well-being. In other words, because education systems tend to be non-linear and interactive, understanding causal connections between different inputs is essential to making sense of real-time data. Another potential pitfall is that real-time data can create an illusion of information if decisionmakers are unable to detect relevant trends from the noise generated by continuous tracking in a complex environment. For example, across four inventory management experiments, Lurie et al. (2009) found that when provided with real-time data, managers focused asymmetrically on the most recent data and failed to adequately compare data from multiple time periods.

Drivers of successful data collection and use

There are a number of potential drivers that can facilitate or hinder the collection and use of data. These can be split broadly into three categories: capacity, logistics, and incentives. The capacity category includes understanding which data are relevant for responsive decisionmaking and when they should be collected, the ability to analyze data and apply data-driven insights to teaching practices and support, and the ability to act on data. The logistics category includes practical issues around data collection and analysis, such as the tools used to collect and analyze data and the format in which data are accessed or displayed. In the final category, incentives are related to the intrinsic or extrinsic motivation to collect and analyze data and apply the findings (Author adaptation from multiple sources including USAID, 2010; Gugerty & Karlan, 2018; Gill et al., 2018, and Kaufman et al., 2022).

Furthermore, these categories can interact with one another. Research on data usage finds that instructors are more likely to use data when schools develop a culture of data use or when there is support, training, time, and/or systems in place for data use. For instructors, data timeliness, perceived validity, and capacity for instructional adaptation all affect self-reported data use; instructors are most likely to use formative data to learn about new students, group students for instruction, and assess weaknesses and strengths at the class level (Tyler, 2011).

This study focuses on one aspect of the data ecosystem: the tools to collect and analyze data—specifically, real-time data for education. A comprehensive deliberation of this category requires that we also take into account other factors mentioned under the costs, incentives, and capacity associated with the process of turning data into decisions. Limited exploration of these factors aside, we restrict ourselves to a detailed description of the nature of digital tools developed for use in education systems. This includes, for example, “how to collect the data” and “how to analyze the data” in the capacity category and “cost of data collection” and “analysis and application” in the incentives category. The other drivers listed above are outside the scope of this study, as are attempts to analyze whether real-time data and/or the factors associated with their collection and use have any impact on learning outcomes. While related and of interest, such literature should be seen as complementary to this report and framework. In the following sections, we provide a brief overview of the tools for data collection before zooming in on digital tools for collection and analysis of real-time data.

Tool typology

Tools to collect data can take many forms, ranging from a simple pen and paper survey to the more current technologies that can collect data remotely on and offline—for example, using social media like WhatsApp and Facebook. In low- and middle-income countries in particular, education data have historically often been collected manually and on paper. The lack of standardization of the data pipeline, susceptibility to human error make it more challenging to analyze and/or use this data to change an intervention. However, in recent years, a boost in the availability, affordability, and accessibility of technology—and education technology specifically—has led to an increase in the digitization of data collection globally. Digital technologies can be used at different stages of the data life cycle, starting from data collection to analysis and dissemination of results to decisionmakers. Digital tools may use different approaches to carry out these functions, depending on the context for which they are designed. The following Table 2 is an indicative, non-exhaustive list of the methods generally used to collect, analyze, and disseminate data (Author adaptation from USAID, 2021).

TABLE 2. Broad methods to collect, analyze, and disseminate data

In high-income countries in particular, the use of digital tools is commonplace; instructors, administrators, and families typically have access to this information via dashboards, allowing them to track student participation and progress, and identify where students need additional support. Historically, due to various constraints (lack of access to technology, software, and internet, as well as limited capacity) in lower-income countries, there are fewer of these technologies available or utilized. However, as the demand for real-time data increases, their use is expanding rapidly.

Digital technologies have an increasing presence in many components of the education system and offer the potential to transform the management of education systems to overcome geographic disparities, personalize learning, and make information available in digital forms (West, 2012). Digital technologies can also specifically support instructors with teaching and reach: engaging with learners in new ways; broadening access to learning materials; reducing administrative burdens; and creating tools for formative and summative assessment (Trucano, 2013). Furthermore, data digitalization enables “closed-loop” experimentation which allows for iterative impact evaluation of an intervention (Kaufman et al., 2022).

Real-time data systems can range across the spectrum of digital technologies—from those that depend on face-to-face surveys and observations to sensor-based automated computational systems that rely on web- or application-based tools for computers, tablets, and phones (smart and basic feature)—for collection, sharing, managing, and reporting data (GIZ, 2017). Integrating more complex digital data systems into education practice requires robust infrastructural support, along with training of personnel involved in the data collection and use (The Economist, 2021).

Innovation in development and application of technology for education has led to remarkable progress. However, it has also led to a crowded field in which many innovations are unnecessarily recreated due to lack of knowledge of their existence, and digital tools are no different.

These requirements often put the sustainable use of digital tools outside the capacity of decisionmakers operating in resource-strained environments. COVID-19 has made an internet/power connection essential to accessing education and has raised concerns whether technology actually reaches the students who need it the most. Globally, 1.3 billion children in the age group 3-17—two-thirds of this population—do not have access to the internet at home. Low-income countries bear the brunt of this digital divide: Only 1 in 20 school-going children in low-income countries have an internet connection at home, while in high-income countries, 9 out of 10 children have internet access. Even within low-income countries, rural areas are disproportionately less likely to have a reliable internet connection (UNICEF, 2020). Access to energy also is similarly inhibited along income and geography lines, with 759 million people living without electricity. While the number of people without electricity is slowly decreasing in the majority of the world, in sub-Saharan Africa, this number increased for the first time in six years. Digital tools must be designed to be resilient to the challenges posed by difficult settings in order to be inclusive and equitable.

With the growing global focus on continuous measurement and adaptive management, many programs are seeking to design intricate data and accountability systems that often hinge on bespoke data collection technologies. Innovation in development and application of technology for education has led to remarkable progress. However, it has also led to a crowded field in which many innovations are unnecessarily recreated due to lack of knowledge of their existence, and digital tools are no different. To aid selection of tools by decisionmakers, in the next section of this report, we categorize tools into a typology based on their functions and purposes.

Multipurpose tools across different education settings can simultaneously collect data on learner performance or analytics in the background while the tool is in use. Given the overlapping functionalities and many different purposes or goals of tools, we created a typology of tools and breakdown of their various functions, as well as why one might choose a certain tool based on in-depth surveys. For a detailed description of the research methodology and survey instrument, please see the Appendix.

Since there are many other types of digital education technology tools available, such as tools for communications, collaboration, and/or education administration that are not as relevant to real-time data collection for education sector or program performance, they are excluded from this typology.

Along with data collection, many of the tools we surveyed had features related to data analysis and dissemination using technology such as offline surveys, data analysis software, etc. listed earlier. Through inductive research, we have also included a third, often related, type of tool which specializes in adaptive content delivery, and may also employ technology used for data collection, analysis, and learning.

• Data collection: Broadly speaking, these are tools that collect data, which may include inputs into the tool (e.g., entered by tool user) or data collected as users interact with a program or application (i.e., data collected passively on the backend of an educational app).
• Data analysis and visualization: These are tools with analysis and visualization capacities based on the collected data, such as calculations, graphs, and more, which could come in the form of a dashboard, for example.
• Learning content delivery: These tools deliver educational or training content, such as guided instruction or games, to the tool user who may be a student or an instructor. They may also assist instructors in adapting and improving instructional practices.

In practice, tools can have multiple functions at once, and it is the combinations of these functions that help to determine the larger purpose and goal of the tool. For the purposes of our research, we have narrowed in on four permutations of these functions and therefore classified results in four main permutations of typologies, or “types” of tool purposes, as shown in Figure 3 below. In the next section, we elaborate on each type and provide an illustrative example.

It is also the case that what appears to the user as a single tool may in fact be multiple lower-level methods and applications integrating into one unified system or platform for the user. For example, the World Bank’s Teach platform utilizes SurveyCTO for its digital application. Some tools may include multiple underlying systems for different types of data collection, which may be analyzed and presented together as a dashboard for program monitoring. For the purposes of this study, our typology considers both the individual applications and the integrated platforms to be “tools” for real-time data collection.

Type A: Data collection

Purpose: These tools collect and store data that are likely to be analyzed with an external statistical program or spreadsheet application. The data could be used to track attendance, administer assessments, deliver household or individual surveys, conduct classroom observations, or more. They differ from other types of tools in that they do not have analysis or content delivery features, but they could be linked or integrated with other tools and systems that do possess these features.

*Note: In our survey, we did not come across a tool that performed the sole function of data collection. This type of tool may have been more common in earlier stages of the digitization process, whereas more sophisticated recent tools often have analysis functions embedded within.

Type B: Data collection + analysis and visualization

Purpose: These data collection + analysis and visualization tools (Type B) collect and store data, and also provide data analysis and visuals of the result. These could be more complex surveys or observational, formative, or summative assessment tools for children or instructors, school or classroom management, and more. The data collected can then be analyzed or visualized (often automatically), for example in the form of an integrated dashboard, and likely would be used for program monitoring and evaluation activities, such as identifying trends, measuring progress, and/or devising potential ways to address challenges.

Example: Waliku is a suite of digital tools for student learning and well-being. Along with tracking enrollment and attendance for more than 30,000 students in Indonesia and Guatemala, it also records health metrics to aid child welfare management.

Type C: Data collection + learning content delivery

Purpose: These tools (Type C) usually function as game-based education technology applications, student learning platforms, or as a lesson/pedagogy delivery tool aimed at instructors or school leaders. With the included data collection functions, these tools are geared toward gaining a deeper understanding of the tool’s usage and users. Content delivery is usually the primary function of these tools, and could include educational content targeted at learners, instructional or coaching content targeted at instructors, family engagement content targeted at caregivers, or others. The data collected alongside content delivery are nearly always collected automatically.

Example: Chimple is an open-source mobile application for children to learn mathematics, reading, and writing in informal environments without instructor supervision. It features an adaptive learning algorithm and a gamified environment to facilitate dynamic learning.

Type D: Data collection + analysis and visualization + learning content delivery

Purpose: These tools (Type D) are typically quite similar to Type C, but with the addition of data analysis and visualization, these applications can provide a one stop shop for delivering content for interventions, as well as measuring impact and presenting that data to decisionmakers in real time. These are typically more complex system-level tools that address the holistic engagement of students, learners, and education systems, and may include education technology applications with adaptive learning capabilities.

Example: Tangerine is an open-source tool that was previously used primarily for offline, oral student assessment and timed data collection required for early grade reading and math assessments. It has evolved to include content delivery through its “Teach” and “Coach” applications.

We have sought to establish a clear framework for understanding and evaluating existing tools, so that decisionmakers can anticipate features that their program might need and what tools may be available to fit these needs, thereby better utilizing limited funding.

Due to a crowded field and lack of clear understanding of what tools are available and how they function, education decisionmakers may have a difficult time identifying existing tools that meet their needs. We noticed the lack of a cohesive framework outlining considerations that might help decisionmakers identify an appropriate digital tool for a given purpose, especially for use in low- and middle-income countries. Thus, we have sought to establish a clear framework for understanding and evaluating existing tools, so that decisionmakers can anticipate features that their program might need and what tools may be available to fit these needs, thereby better utilizing limited funding.

Tool selection criteria

In addition to considering the above typology, there are many other factors to consider when selecting or developing a digital tool for real-time data collection in education. These considerations can range from highly contextual to broadly purposed goals, and can cover everything from programmatic structure or goals to the limitations of the physical environment. For example, some tools are focused on early childhood years, others are more open-ended, some are available only in certain languages, and still others are suitable (or not) for low-power or internet contexts.

These elements all come together in the evaluation of a program or organization’s needs for data collection. As depicted in Figure 4, we evaluate these considerations across three main categories: functions, usability, and context.

There are a number of sub-elements in each of these categories, related to their varied features and applications, which can make them better or less suited for particular uses. These sub-elements were based on a survey that we administered to tool developers (see Appendix for further details), and examples of tools from this survey administration can be found in our interactive Tool Finder database .

The components and sub-components are summarized in Table 2 below, and further described in the section which follows.

TABLE 3. Detailed selection criteria for digital tools

Selection framework

This section describes the various categories of consideration and how they may influence the selection of a tool. To view tables of our sampled tools across these different categories, please see the Appendix.

As described in the previous section, the primary typology of tools includes their functions for data collection, data analysis and visualization, and learning content delivery. However, each of these three elements also have sub-components contained within them, which may impact tool selection.

Data collection: When we think of data collection, one of the first considerations for decisionmakers is what data are being collected. While this can vary greatly across different tools and designs, we can often think of this data in terms of what it measures, in the categories of student (child) or provider (instructor) data, and whether they include enrollment, attendance, completion rates, performance (self or externally assessed), or usage information. This data could be in the form of assessments (typically formative), classroom observations, or enrollment/attendance trackers.

Another important consideration is how a tool collects data, which provides a lot of information about how the tool functions. This can primarily be thought of as active data collection, in which a user is interfacing with the tool to enter information (such as a survey), or passive (such as when the tool automatically collects data while it is functioning, as with application analytics). Furthermore, active data collection could be through self-administration (such as a self-reporting questionnaire) or external administration (such as a proctor-delivered assessment). Whether passive or active, data can be collected (or intended to be collected) at different periodic frequencies (e.g., daily, weekly, monthly, quarterly, annually, or at beginning/end of the program) or in real time.

Directly related to the how and what is who is collecting the data and who is using it. In contrast to passive tools, which collect data automatically, active tools could be administered by instructors, coaches, school or center administrators, researchers or M&E officials, families, or children. These data —in either individual or aggregate format—are often used by decisionmakers at different levels of the education system, ranging from parents and families to instructors and school officials, to INGO or research/evaluation staff, to government officials at local or national levels.

Data analysis and visualization: The goal of data in this format is often to inform the action of decisionmakers (such as instructors, practitioners, and families), and analyzing and visualizing data is often key to achieving that outcome. These decisionmakers can occupy any level of the education system, including students, parents, instructors, school leaders, program staff, or government officials, and their actions can be informed through a number of different approaches. Dashboards are a common feature of such tools, where real-time and aggregated data are presented to track program or learner status and identify strategies for further action.

While the format of the analysis and output can take many forms, one important function is their automation and level of user input. Some tools provide (or have the ability to provide) automatic data analysis, visualizations, and / or recommendations across a set of predetermined metrics that the tool designers or users can set themselves, including intervals (such as daily, weekly, monthly, quarterly, or annually). Other tools provide data analysis, visualizations, and/or recommendations with the input or manipulation of the user; while this approach may not be as quick as automatic recommendations, it may allow the user more insights into which data to consider and how—questions which may only emerge once the data are collected. Whether automatic or manual, data analysis and recommendations can be provided at various time intervals—from real time / continuous to annually.

Learning content delivery: This includes tools designed to deliver content to users—often students or instructors—and then gather data in the background—often passive data about the tool use or user performance. This may include a wide range of functions—for example, child-focused, game-based learning / education applications, tools which focus on tech-enabled lessons / content delivery (e.g., synchronous or asynchronous virtual lessons), or computer-based adaptive learning.

Since the focus of our research is on tools that contribute to data analysis, we do not seek to further categorize types of content delivery programs. Instead, we direct readers to the wealth of resources on education technology, such as the Brookings education-technology playbook .

There are numerous usability considerations for specific programmatic needs and contexts, which can aid in determining which tools are best suited.

Platform: The platform that deploys a digital tool can play a significant role in determining its selection. This may include the tool’s codebase, as well as the hardware (technology device), software (interface and program), and operating system with which the digital tool is compatible. Hardware refers to the tool’s compatibility with a tablet, computer, smartphone, and/or basic/analog phone, while the operating system may be designed for Android, iOS (Apple), Windows, and/or other software.

Connectivity: Different tools and their accompanying hardware require different power and internet connectivity functions to operate, which may be particularly relevant for a given context. Power and internet connectivity could be needed at all times, only occasionally (such as for charging or for data transmission), or not needed at all for the operation of the tool.

Integrative capabilities: Tool integration can primarily be viewed in two dimensions: integration with other tools and integration within existing systems—both of which can be critical to effective uptake and usage. Tool-to-tool linkages can be configured in different ways, including database exports and imports, database connections, and APIs, or they can be customized. Tool-to-system integrations, such as with education management information systems (EMIS), can be direct or indirect and may be possible functions even if not currently utilized in that way.

Adaptive capabilities: Adaptive capabilities are one of the most important considerations when looking to adopt an existing tool. These include whether the tool can be updated or customized in a general sense, as well as whether the user can modify specific fields or settings. In addition, some tools may need periodic maintenance and upgrades, which may impact their capacity for adaptation. Another important consideration of adaptive capabilities is whether and how much training is available and / or required of users. This aspect may also encompass a tool’s adaptability to different types of user needs, including potential disabilities.

Data security: Data security is a key concern of many seeking to utilize tools for data collection. This could include those related to software, such as encryption of the data (and how), protections available, and who has access and how, as well as geographic or political considerations. This can also relate to the geographic location of the servers, whether the tool and/or data storage are GDPR compliant, and whether or not the tool is open source.

Tool costs: When considering the use of an already existing tool or building a custom tool, cost can be a key consideration. The cost of developing new tools can range widely but does run quite high. While adapting or adopting an existing tool is typically much less expensive, there are important factors to consider, such as the cost of use or licensing by another organization, the cost of ongoing tool maintenance, and the cost of training tool users. Some tools offer various options, which can impact cost; open-source solutions, for instance, are typically “free” to install on one’s own server and manage there (but could incur a cost for server backup space), as well as have software-as-a-service options, which include a fee for use.

Some tools are developed for specific locations or contexts, which may enable or restrict their use for some applications or cases. While it can be useful to consider the contexts for which a tool was developed or how it has been used in the past, it is also important to keep in mind its adaptability, as an investment in adaptation may make a tool suitable beyond its current use or design.

Setting/intervention: The tool may have been developed or only used thus far in specific settings or interventions. The interventions could include, for example, primary or early childhood education and could be targeted at parent support, early childhood development, preprimary, primary, or secondary school. The settings of these tools may be directly related to the type of intervention, such as at schools (public or private), care centers, or home. Given the importance of different data for each of these settings or interventions, some tools may be better suited for a specific program or use.

Target population: Certain populations have specific needs that could be well suited to a tool’s design. This could include the age of the children, as well as other characteristics, such as gender (tools with data disaggregation functions) or children with disabilities. In addition, some tools may be well suited to other contexts, in particular crisis and conflict settings, which have specific needs to consider in terms of tool functionality.

Geography/language: A key constraint for tool adoption could be the language(s) the tool interface and tool delivery have available. While translation may be possible, it is important to keep in mind cost considerations. Additionally, some tools may be more contextually appropriate to certain regions or geographies. Reasons for this may include cultural considerations of content (such as characters depicted and relevance of examples used), regional educational standards, and more.

Subject(s): Some tools for teaching and learning may have been developed thematically (such as social-emotional learning) or subject-specific (such as math and language). While it may be possible to expand tools to other subjects, it is important to keep in mind the cost considerations.

Decisionmakers, armed with rich contextual knowledge of their intervention and looking to use a digital tool, can use our framework to anticipate needs when they choose to adopt a new tool for their program.

Decisionmakers , armed with rich contextual knowledge of their intervention and looking to use a digital tool, can use our framework to anticipate needs when they choose to adopt a new tool for their program. It may also serve as a guide for those looking to develop their own tools from scratch, so that they can create original platforms that solve previously unaddressed problem areas. The interactive Tool Finder database complements the framework by identifying and providing details of digital tools that can help determine whether an existing tool can be adopted and/or adapted, or whether it makes sense to spend limited programmatic funding to develop a custom tool.

With the global education crisis and compounding challenges of COVID-19, it is now more critical than ever to make data-informed decisions toward achieving a range of equitable learning outcomes. Real-time data can help instructors, administrators, and even parents ensure that classrooms and learning spaces are managed adaptively to best serve learners. To accomplish this, it is necessary for data to be collected and presented: at the right time , with sufficient frequency , and to the right people in a way that is understandable and actionable .

For each of these elements, it is essential to have effective and efficient real-time data systems and tools . Such tools are available in a variety of forms, with many different features and purposes. In this report, we establish clear frameworks and criteria for appraising digital tools and understanding their key functions (data collection, data analysis and visualization, and learning content delivery) and we analyze the critical factors to consider when selecting or developing a tool.

Although digital tools may not be applicable for data collection in all contexts, especially in rural areas without power or internet, we highlight potential benefits of using digital tools for data gathering and analysis: timeliness, accuracy, dependability, cost effectiveness, and the ease of presenting and sharing data.

To ensure that the potential benefits of real-time data are maximized, stakeholders in the education sector could benefit from increased dialogue regarding data requirements, obstacles in their utilization, and data-driven learning decisions. Despite the importance, few countries have in place the policies, infrastructure, capacity, and/or tools required to generate regular and actionable performance data (World Bank, 2013). This is especially true for low-income countries, which stand to benefit the most from the implementation of robust systems to measure learning outcomes and a systems approach to education analysis. Future attention and research should focus on exploring these enabling factors, as well as on innovations in technology as the field evolves. It is our aim that this report, the frameworks provided within, and the accompanying interactive Tool Finder database can act as a valuable resource in understanding and taking better advantage of digital tools for real-time data collection as we continue to work toward ensuring that all children have access to quality teaching and learning experiences.

file-pdf References and appendix

Acknowledgments

We are grateful for the contributions to the research by Izzy Boggild-Jones, Paroma Mitra, Arushi Sharma, and Will Powers. In addition, we would like to thank Judith-Ann Walker and Emily Morris for their peer review of the report, as well as Katherine Portnoy for her editorial review. We are also greatly appreciative of Carmen Strigel and Abby Carlson for their thoughtful review of the research methods, survey and typology. In addition, we would like to thank the advisory panel members to this project who provided feedback in the early stages of the research: Ariam Mogos, Michael Trucano, Peter Blair, Zak Kaufman, Carmen Strigel, Abby Carlson, Chavaughn Brown, Nicolas De Borman, Sharath Jeevan, Rein Terwindt, Brian Komar, Elena Aria Sortiz and Sonja Giese. Finally, we are greatly to the respondents to our survey on digital tools for taking the time to fill in the survey and for their willingness to share the details of the tools.

The Brookings Institution is a nonprofit organization devoted to independent research and policy solutions. Its mission is to conduct high-quality, independent research and, based on that research, to provide innovative, practical recommendations for policymakers and the public. The conclusions and recommendations of any Brookings publication are solely those of its author(s), and do not reflect the views of the Institution, its management, or its other scholars.

Brookings gratefully acknowledges the support provided by the BHP Foundation.

Brookings recognizes that the value it provides is in its commitment to quality, independence, and impact. Activities supported by its donors reflect this commitment.

About the Authors

Emily gustafsson-wright, senior fellow – global economy and development, center for universal education, sarah osborne, center manager – global economy and development, center for universal education, muskan aggarwal, former intern – global economy and development, center for universal education.

• Media Relations
• Terms and Conditions
• Privacy Policy

• Top Courses
• Online Degrees
• Find your New Career
• Join for Free

What Degree Do I Need to Become a Data Analyst?

Do you need a degree to become a data analyst? If so, which one? Learn these answers and more.

Getting an in-demand job as a data analyst often starts with getting the right skills and qualifications. For many, this might mean a degree. In this article, we’ll discuss whether you need a degree to become a data analyst, which degree to get, and how a higher-level degree could help you advance your career.

Read more: What Does a Data Analyst Do? A Career Guide

Do I need a degree to become a data analyst?

Most entry-level data analyst jobs require a bachelor’s degree, according to the US Bureau of Labor Statistics [ 1 ]. It’s possible to develop your data analysis skills —and potentially land a job—without a degree. But earning one gives you a structured way to build skills and network with professionals in the field. You could also find more job opportunities with a degree than without one.  

Data analyst degrees: What should I major in?

Not all universities offer a bachelor’s degree in data analytics. So what should you major in if you want to pursue a career as a data analyst? Even if your university doesn’t have this specific degree, it likely offers other majors with overlapping skills.

Possible majors for data analysts

Here are some degree options that typically teach common data analysis skills. If you’re looking toward a career as a data analyst, these majors could be a good fit.

Data science: In response to the increasing demand for data professionals, more and more schools are offering bachelor’s degrees in data science. In this degree program, you’ll typically take courses in computer science, statistics, and mathematics. Some programs let you specialize in fields like economics, finance, business, or health care.

Computer science: The emphasis on statistical and analytical skills in many computer science programs makes them a good fit for aspiring data analysts. This degree is also widely available. Depending on the program, you might study artificial intelligence concepts, algorithm design, and programming languages that you can use in your future career.

Applied mathematics, or statistics: Traditional mathematics degrees generally prepare learners for careers in academia. Applied mathematics and statistics degrees shift the focus to real-world applications, like helping businesses make data-driven decisions. The curriculum might include other important skills, like programming languages or statistical software packages.

Finance/economics: If you think you might be interested in working as a financial or business analyst, consider getting your degree in finance or economics. Many of these degree programs include coursework in statistics and analysis, and some offer concentrations in business analytics . 

Psychology: It might not seem obvious at first glance, but psychologists use data to describe, explain, and even predict human behavior all the time. A Bachelor of Science in Psychology might expose you to math and statistical analysis coursework.

Management information systems (MIS): With this degree, you can get a behind-the-scenes look at databases and how they work. This could prove useful as a data analyst. MIS coursework typically covers topics like database design, data management, and business theory. With some programs, you can specialize in data analytics, business intelligence , or data management.

No matter what you choose to get your degree in, be sure to take classes in statistics, calculus, and linear algebra, as well as some computer science classes that cover database and statistical software. If you already have an industry in mind, it can help to take some industry-specific coursework (finance, health care, or business, for example).

Learn more about earning your bachelor’s degree online through Coursera.

Is a master’s in data analytics worth it?

While a bachelor’s degree is the most common entry-level qualification, some companies look for candidates with a master’s degree in data analytics or a related field. A 2017 study by IBM found that six percent of data analyst job descriptions required a master’s or doctoral degree [ 2 ]. That number jumps to 11 percent for analytics managers and 39 percent for data scientists and advanced analysts. 

In general, higher-level degrees tend to come with bigger salaries. In the US, employees across all occupations with a master’s degree earn a median weekly salary of $1,497 compared with $1,248 for those with a bachelor’s degree [ 3 ]. That difference translates into $12,948 more each year.

If you’re looking to advance your career in data analytics or move into data science, earning your master’s degree could set you up for success.

Read more: Going Back to School: 7 Things to Consider

Get started as a data analyst

Take the next step toward a career in data analysis, no matter where you are in your learning journey.

Build a foundation of job-ready skills with a Professional Certificate in data analytics from industry leaders like Google , Meta , or IBM .

Request more information about earning your data analytics degree online.

Try a popular data analytics course to see for yourself if it’s a good fit.

Article sources

US Bureau of Labor Statistics. " Operations Research Analysts , https://www.bls.gov/ooh/math/operations-research-analysts.htm#tab-1." Accessed June 7, 2022.

IBM. " The Quant Crunch: How the Demand for Data Science Skills Is Disrupting the Job Market , https://www.ibm.com/downloads/cas/3RL3VXGA." Accessed June 7, 2022.

US Bureau of Labor Statistics. " Learn more, earn more: Education leads to higher wages, lower unemployment , https://www.bls.gov/careeroutlook/2020/data-on-display/education-pays.htm." Accessed June 7, 2022.

Keep reading

Coursera staff.

Editorial Team

Coursera’s editorial team is comprised of highly experienced professional editors, writers, and fact...

This content has been made available for informational purposes only. Learners are advised to conduct additional research to ensure that courses and other credentials pursued meet their personal, professional, and financial goals.

We use cookies on our website to support technical features that enhance your user experience, and to help us improve our website. By continuing to use this website, you accept our privacy policy .

• Student Login
• No-Cost Professional Certificates
• Call Us: 888-549-6755
• 888-559-6763
• Search site Search our site Search Now Close
• Request Info

Skip to Content (Press Enter)

What is Data Analytics? A Simple Breakdown

By Brianna Flavin on 05/02/2024

Da ta is everywhere these days. Businesses want to make data-driven decisions. And pretty much every industry wants to utilize the new capacities we have to gather information and turn it into better results, more efficiency and more profit.

But that much information, that much raw data is basically useless without data analytics. It's like dumping an ocean on someone and asking them to pull a dose of cough syrup out of it.

Data analysis allows you to navigate an ocean of information, identify trends, pull select insights, and most importantly—learn.

So, what is data analytics?

Data analytics is the action of collecting, organizing and finding ways to use huge amounts of information. It's simple to say. But data analysis is so versatile and can apply in so many different ways that data analytics isn't a job so much as an entire industry of jobs.

There's honestly so much going on in data analytics that it's overwhelming to just get a basic grip on the field. But when you do dive in, you’ll see how exciting the industry can be.

“Data is lit AF,” laughs Nathan Keysser, senior data analyst at BI Worldwide . “The ability to take large amounts of meaningless data and turn it into meaning, into something useful and interesting... it really satisfies that problem-solving itch.”

This is a really good field for people who enjoy puzzles, according to Keysser. If that’s you, keep reading. See how it works.

What does data analysis look like?

If you start from square one, there's a pretty straightforward process to data analytics. You can break all of these steps down into more steps if you want to get really detailed. But for a general understanding of how to analyze data, this is what you need to know.

Step one: Data collection

Data analysis can't happen without data collection—so that's the first step in the data analysis process. Companies, data analysts and data scientists need to figure out how to source information relevant to their needs.

This could look like your favorite shoe company sending out an automated survey after every purchase and collecting the results. Or it could be web cookies that record information about who visits a webpage, what they click on, how long they linger, etc... Data mining is a more-automated aspect of data science that's all about how to gather mass amounts of raw data for useful analysis.

Hospitals might use data from their electronic health records, law firms might analyze historical data on case studies, and environmental organizations might collect data about the soil, wind or solar conditions in hundreds of locations. Basically, you need a way to compile useful information before you can make use of it.

Step two: Data processing

Once you have lots of information, you have to organize it into a format you can easily sort or play around with. This might mean digitizing paper records, porting information from assorted files into one spreadsheet or hundreds of other methods.

This part of analyzing raw data is usually very time-consuming because you not only have to do a lot of manual, individual work—you also have to consider very carefully which aspects of information might be relevant. Otherwise, you’ll wind up doing the process all over again.

People used to ask is data analytics important enough to really splash out on expensive data analytics tools? But as people understand the power of data, it’s often more of a priority area. For a lot of companies, technology that can automate this time-consuming process and assist with data management might well pay for itself before too long.

Step three: Data analysis

Now we've reached the name of the game! After data is stored somewhere secure, accessible and sort-able, you're ready to actually analyze. How you analyze will vary a lot based on your organization, the needs of the moment and even how you prefer to work.

“Generally, I know what I want to find,” Keysser says. “But I frequently don't know how the data is structured, and I have to do a fair amount of problem solving to figure out exactly how I want to pull the data.” Keysser adds that since data analytics sounds so technical, many people don't realize how creative the work can be.

Data scientists and data analysts can approach analysis from many different angles. You could try to find the answer to a specific question. Ex: Will buyers in Illinois spend 15$ on this type of shirt?

Or you could pinpoint something and try to pull as much data related to that as possible. Ex: What is everything we know about people who've purchased a shirt from this website.

The options in data analytics are as wide as your imagination here. But there are defined types of data analytics that can give you some ideas.

Descriptive analytics

This type of analysis can be pretty simple. Descriptive analytics is all about answering the what, when, how types of questions in a given data set. For example, descriptive analytics could mean doing a simple statistical analysis to find the median number in a huge group of statistics.

Or to be more specific, a telehealth provider could keep track of when their clients with young children schedule consultations for strep throat. Descriptive analysis could help them track the months of the year with the most consultations over the last decade and isolate the peak times for strep throat consultations in children.

Descriptive analytics is a good first step to perform data analysis because it can help you identify errors like typos in your data, as well as find commonalities you may not have initially considered.

Diagnostic analytics

Diagnostic analytics goes a little deeper, into the "why" questions we can bring to data. This can range from technical questions (why does our website keep crashing) to more social/emotional questions (why are employees leaving our company).

For example, a company might wonder why they are experiencing higher turnover in their employees. Collecting information from exit interviews, they could use diagnostic analytics to pinpoint the problem areas employees most commonly cite and find ways to weigh the value of each reason.

Predictive analytics

Predictive analytics does exactly what it sounds like—predict things. If you analyze data on every tornado to ever hit the U.S., for example, you can use predictive analytics to help you determine when tornadoes might occur again in the future.

Since having some insight into future trends is highly-profitable for pretty much everyone, predictive analysis is an excellent way to draw valuable insights.

Prescriptive analytics

Prescriptive analytics is all about using data to choose the next course of action. It relies on predictive analytics, but then creates a recommended next step . This can be very helpful when there are lots of predictive variables involved. In the tornado example, cities could use a combination of data on when tornadoes are most likely to occur and what weather conditions will likely occur first and make a prescriptive decision about when they should run tornado warning sirens.

This type of data analytics often involves algorithms, machine learning or automation to help deal with all the raw data. Think about the way a credit card company might monitor transactions—that's way too much info for data analysts to wrangle. So, the company might create a prescriptive analytics algorithm to pay attention to spending patterns, flag anomalies and recommend a fraud or theft alert.

Then, if a customer buys something highly unusual, they will automatically receive a fraud alert and a hold on their card to limit potential damage if the card were stolen.

Step four: Data interpretation

As you can see from the types of data analytics, the whole point of all this is to draw meaningful conclusions from data.

After data is analyzed, those insights are now available. But it still takes some work to explain and translate. This is where data visualization and data modeling come in! You could have the highest quality data analysis in the world, but if people can't understand the results of your analysis, that hard work won't go anywhere.

One of the most important data analytics techniques is to create visual, clear ways to explain the data analytics process (and the results and recommendations) to people outside the field.

How do organizations use data analytics?

You're probably getting the picture already. An organization could use data analytics to answer questions about....

• What they are doing (who are our customers or clients, how much do we produce, where are we most successful)
• Why something is happening (why isn't this working, why is this working)
• What might happen in the future (when will we be busiest, how often is something likely to happen)
• What should happen next (at what point should we react, what is our best course of action when ___ happens)

It's hard to overstate the potential in all that.

It goes way beyond making more money too. Data analytics can help with healthcare, education, governing, economics, politics and all of the “this is too big to wrap my head around” elements of society.

“Knowledge is power,” Keysser says. “I know it's kind of cheesy to say that, but the more we can find truth in data, the better choices we can make about our future.”

The people who puzzle it out

Data analysts are one type of professional in the field of data analytics. If you like to solve problems, you might be curious to learn a little more about what it's actually like to work as a data analyst. All of this macro-level information is great for a general overview—but it gets way more interesting when you zoom in on the actual job.

Get those details at What Does a Data Analyst Do? Exploring the Day-to-Day of This Tech Career .

• Share on Facebook
• Share on Twitter
• Share on Pinterest
• Share on LinkedIn

Request More Information

Talk with an admissions advisor today. Fill out the form to receive information about:

• Program Details and Applying for Classes
• Financial Aid and FAFSA (for those who qualify)
• Customized Support Services
• Detailed Program Plan

There are some errors in the form. Please correct the errors and submit again.

Please enter your first name.

Please enter your last name.

There is an error in email. Make sure your answer has:

• An "@" symbol
• A suffix such as ".com", ".edu", etc.

There is an error in phone number. Make sure your answer has:

• 10 digits with no dashes or spaces
• No country code (e.g. "1" for USA)

There is an error in ZIP code. Make sure your answer has only 5 digits.

Please choose a School of study.

Please choose a program.

Please choose a degree.

The program you have selected is not available in your ZIP code. Please select another program or contact an Admissions Advisor (877.530.9600) for help.

The program you have selected requires a nursing license. Please select another program or contact an Admissions Advisor (877.530.9600) for help.

Rasmussen University is not enrolling students in your state at this time.

By selecting "Submit," I authorize Rasmussen University to contact me by email, phone or text message at the number provided. There is no obligation to enroll. This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

About the author

Brianna Flavin

Brianna is a senior content manager who writes student-focused articles for Rasmussen University. She holds an MFA in poetry and worked as an English Professor before diving into the world of online content. 

Posted in Data Analytics

• business careers
• tech careers

Related Content

Kirsten Slyter | 09.19.2022

Brianna Flavin | 06.28.2022

Carrie Mesrobian | 04.25.2022

Carrie Mesrobian | 02.14.2022

This piece of ad content was created by Rasmussen University to support its educational programs. Rasmussen University may not prepare students for all positions featured within this content. Please visit www.rasmussen.edu/degrees for a list of programs offered. External links provided on rasmussen.edu are for reference only. Rasmussen University does not guarantee, approve, control, or specifically endorse the information or products available on websites linked to, and is not endorsed by website owners, authors and/or organizations referenced. Rasmussen University is accredited by the Higher Learning Commission, an institutional accreditation agency recognized by the U.S. Department of Education.

• Career Guidence
• IoT Resources
• Other Resources
• Power BI Resources
• Press Releases
• Tableau Resources
• Data Science Courses
• Visit Datamites
• Data Analyst Courses
• AI Resources
• ML Resources
• MLOPs Resources
• Python Resources
• Deep Learning Resources

Data Analytics Resources

• Data Engineer Resources
• Data Science Resources
• Data Science Success Stories
• Press Release

Join Our Newsletter

Join our subscribers list to get the latest news, updates and special offers directly in your inbox

How is Data Analytics Influencing the Educational Sector?

The emergence of data analytics in education has completely changed the nature of the field and unlocked a world of cutting-edge technological innovations that have fundamentally altered the way that traditional teaching and learning are conducted. The use of  data analytics in education  is currently having a profound impact on student achievement. The innovative changes made to the infrastructure and system of the previously prevalent educational technique, this has increased students’ creative abilities as well as their excitement for learning.

Do you think that data analytic trends actually work that well, and can they genuinely change the educational technology sector?

What is data analytics?

To proceed with our explorative journey into data analytics, it is fundamentally an act of looking at datasets in order to draw inferences. This could entail spotting trends, patterns, and even connections to make future event predictions. As a result, data analytics can help you analyze client behavior, make smarter business decisions and even enhance operations!

It’s true to say that the pandemic has served as a warning for many firms all over the world, thanks to the knowledge gained from information, new possibilities have arisen. Now, business owners are starting to recognize the value of this wealth of knowledge and ways to promote success. The power of data technology is now visible to different industries, so it is no longer just the domain of big businesses.

Educational institutions have also chipped in, adopting data analytics and leveraging predictive modeling, which has had a huge impact on how the institutes work. New methods of learning, teaching, and assessment are made possible by  data analytics trends  that are altering the EdTech industry at a large scale.

Refer this article:  Support Vector Machine Algorithm (SVM) – Understanding Kernel Trick

Data Analytics in Education

In response, the educational technology sector is repetitively transforming and changing as new tools and techniques are getting introduced. One of the most recent breakthroughs in the domain is data analytics, which can assist teachers in determining areas where students are having difficulty and require support. Additionally, data analytics may be utilized to monitor student motivation and engagement, which can assist teachers in optimizing their teaching techniques.

The use of data analytics in education is getting bigger, and over the next two years, it is anticipated to soar higher, be widespread, and be adopted as per  Gartner’s 2021 Hype Cycle of Higher Education . Subsequently, organizations today have a rare chance to learn how to implement analytics effectively inside their own structures and establish the foundation for success as the technology is taken up by more and more businesses.

The onset of data analytics has markedly changed the nature of the educational industry and unlocked a world of cutting-edge technological paradigms that have fundamentally altered the manner in which traditional teaching was conducted. Data analytics are already permeating every aspect of the educational system, producing ground-breaking outcomes. The success of educational interventions, personalized learning for students, and finding and filling in learning gaps are all advantages of  data analytics in education . Let’s get to some of them in detail.

Read this article:  Understanding Auc Roc Curve

Arming students

Teachers can execute more effective lesson plans that are participatory and beneficial if they have an organized overview of structured as well as unstructured data. By promptly and consistently attending to the student’s needs, teachers can empower learners to make connections across various courses and boost their learning opportunities.

Educational data mining

With this strategy, we can create predictive models for recognizing students who are at risk of failing in any subject and support teachers in delivering interventions to help them succeed.

Adaptive content and intelligent curriculum

We can create as many curricula as we have pupils by using their data. It is feasible to create a recommendation system based on the preferences (and abilities) of the students, allowing them to choose from a variety of approaches to learning the same material.

Assessing student success

Make a prediction regarding the student’s ability to learn about any course or general facts about your university.

Enhance Learning Results with Custom Modules

With tailored courses,  data analytics  can improve students’ academic progress and personal growth.

Facilitates Teachers’ Understanding of Their Students

Teachers can better understand the student’s interests and try to improve their learning experience with the use of data acquired from behavior analysis and other evaluations. A data-driven education system aids institutional mentors in planning and creating academic experiences and study curricula that are tailored to each student’s abilities, learning styles, preferences, and academic achievement. The teachers are able to receive individualized feedback on each student’s performance as well as that of the entire class, and they can change their teaching style to better meet the needs of each student.

Classifying students

Find out which learning format results in the most effective and productive pupils using the pertinent data that has been obtained. Additionally, knowing the information can help students have a better academic experience overall.

Statistical models

Are used to predict the class grade averages of the pupils. Denoting to the instructor that the students will need to put in more effort to achieve the desired grade, the model predicts a student’s CGPA based on the data it has collected. This model also helps teachers take a closer look at the level of competence a student has in a particular subject.

Also read this article:  10 Python Built-in Functions

Higher Efficiency

By assisting them in identifying areas for improvement, data analytics can help EdTech organizations become more productive in their daily operations. Data analytics, for instance, can assist businesses in determining the most lucrative aspects of their operations and those that require further development.

Elevated Decisiveness

By supplying them with knowledge of the behavior of their clients, data analytics can also assist EdTech companies in making better decisions. Making judgments regarding The services to offer and how to sell them can be aided by this.

Better Customer Insights

In order to better understand their clients, EdTech companies can benefit from data analytics thereby being able to produce those services and goods that are likely to suit the wants of their clients.

Greater Safety

By guiding them in locating potential security flaws and addressing them, data analytics can assist EdTech organizations in enhancing their security systems.

Higher revenue

Data analytics  can benefit the EdTech industry by raising its revenue by facilitating the initiation of new potential growth opportunities.

Better perception of students

By giving them perspectives on how students learn and act, data analytics can aid EdTech companies in getting a clearer picture of the student’s needs. This will facilitate the growth of businesses structurally and financially through the development of students by meeting their needs.

Data Analytics to Identify the Best Teaching Methods

To monitor and display each student’s educational and behavioral patterns, data analytics can be effectively used. Teachers can evaluate the data gathered and utilize it to learn more about how the most successful students at the institution behave and operate. The data is evaluated to identify the most effective teaching strategies, and the remaining students are then urged to use these strategies and approach their studies similarly in conducive to pulling off the top-drawer academic success.

Learn Data Analysis with the Help of Python

Learning outcomes

Student learning outcomes are among the core areas in which analytics may be used to promote success. Data analytics software can be used to find engagement possibilities for student assistance, establish individualized development plans, and provide important insights into how the curricula can be made effective.

Enrollment Process

The enrollment process is a critical area in which data analytics thrives in education.  Data analytics  enables universities to be more strategic in their enrollment process by linking information gathered with the success of their marketing and recruiting initiatives in luring new students.

Even in a time of intense competition and low enrollment, institutions employ predictive analytics to support retention efforts, improving conversion rates and raising retention.

Operational Effectiveness

The application of data analytics can significantly augment the efficiency of an education institution/EdTech company as a whole. Utilizing data analytics insights can help get a better understanding of the pupils and alter the company’s financial outlook. The time of logging in and out for each student as well as the teacher can be accurately and excellently monitored with data analytics. It is even possible to deploy a tool based on intelligent data analytics to track student’s attendance in the library, gym, and other spaces.

Predicting future

Data analytics-driven education programs help mentors and organizations acquire a thorough understanding of a student’s academic development as well as a determination of the student’s strengths and weaknesses. On the basis of the performance statistics collected through data analytics, it is possible to effectively use data analytics to identify students who are at risk of failing and then guide them toward choosing a course of action that will lead to academic success.

What is Exploratory Data Analysis

Participatory pedagogy and Data Analytics

Data analytics-driven educational environments make it easier to keep professors, teachers, students, assistants, and everyone else working together on various projects on the same platform. Utilizing data analytics effectively and efficiently for the integration of pedagogy and technology aids in the creation of a cohesive, analytical, and forcefully driven ecosystem that allows for the real-time interchange of skills and knowledge.

Evaluating academic performance

Data analytics can be effectively used to monitor how technology, devices, hardware, and software are used throughout the course of the day. The results can then be evaluated to see which technology performs the best. The most efficient way to help instructors is using data analytics, which can be used to understand and analyze how pupils acquire and process information. Utilizing data analytics, competency-based learning may be created so that each student can work at their own speed and achieve academic success.

As per  ResearchAndMarkets , the global market for big data analytics in the educational sector is projected to raise to  $47.82 billion  by the year 2027 with a market growth of  20.79% CAGR  during the forecast period. Predictions are that whilst data analytics is still in its early stages, educational institutions ought to welcome it with open arms since, this domain is ultimately going to revolutionize the way that education is delivered.

Now you may learn and master data analytics from the basics through DataMites  Data Analytics Training  which is offered in collaboration with internationally recognized  IABAC ,  JainX , and  NASSCOM . DataMites also provides courses in  data science  and related technologies like artificial intelligence, data engineer, and python. Learn data analytics from professionals in the field through  DataMites  to get ready for this fascinating, expanding career!

What is HR analytics?

Previous Article

What are the Top IT Companies in Australia?

Next Article

Machine Learning Course Fee in Hyderabad

Datamites Team

DataMites Team publishes articles on Data Science, Machine Learning, and Artificial Intelligence periodically. These articles are meant for Data Science aspirants (Beginners) and for those who are experts in the field. It highlights the latest industry trends that will help keep you updated on the job opportunities, salaries and demand statistics for the professionals in the field. You can share your opinion in the comments section. Datamites Institute provides industry-oriented courses on Data Science, Artificial Intelligence and Machine Learning. Some of the courses include Python for data science, Machine learning expert, Artificial Intelligence Expert, Statistics for data science, Artificial Intelligence Engineer, Data Mining, Deep Learning, Tableau Foundation, Time Series Foundation, Model deployment (Flask-API) etc.

Related Posts

How to Become a Data Analyst in Bangladesh?

Datamites Team Jul 17, 2023 0 45

How much is the Data Analytics Course Fee in Gurgaon?

Datamites Team Apr 1, 2023 0 15

How to Become a Data Analyst in Pakistan?

Datamites Team Jul 12, 2023 0 58

How to Become a Data Analyst in Trivandrum?

Datamites Team Sep 11, 2023 0 17

Data Analyst Course Fee in Hyderabad

Datamites Team Mar 19, 2024 0 41

How to Become a Data Analyst in Guwahati?

Datamites Team Sep 27, 2023 0 8

DROP US A QUERY

Popular posts, support vector machine algorithm (svm) – understanding....

Datamites Team Sep 7, 2019 0 380

What are the Top IT Companies in Manila?

Datamites Team Mar 28, 2023 0 307

What is the Salary of a Data Scientist in Oceania?

Datamites Team May 25, 2021 0 240

Datamites Team Nov 2, 2023 0 224

Resampling Methods in Machine Learning

Datamites Team Mar 5, 2021 0 212

What Are The Top IT Companies In Germany?

Datamites Team Dec 8, 2022 0 190

Logistic Regression with Example

Datamites Team Sep 11, 2021 0 184

Artificial Intelligence Perform on Travel and Tourism?

Datamites Team Jun 1, 2022 0 142

Recommended Posts

Datamites Team May 4, 2024 0 67

Machine Learning Course Fee in Pune

Datamites Team May 3, 2024 0 57

What is Certified Data Engineer?

Datamites Team May 3, 2024 0 66

Machine Learning Course Fee in Chennai

Datamites Team May 2, 2024 0 45

What is Datamites Certified Data Analyst Certification...

Datamites Team May 2, 2024 0 64

Random Posts

Machine learning course fee in bangalore.

Datamites Team Apr 30, 2024 0 54

What would be the Data Engineer Course Fee in India?

Datamites Team Apr 16, 2024 0 107

Data Science Course Fee in Pune

Datamites Team Mar 14, 2024 0 38

How to Become a MLOps Engineer in Pune?

Datamites Team Sep 14, 2023 0 27

How Much is the MLOps Course Fee in Pune?

Datamites Team Jul 15, 2023 0 27

Home News Highlights Best of the Week Exclusive data analysis powers Education Reporting Network collaboration with dozens of newsrooms

Best of the Week

Exclusive data analysis powers Education Reporting Network collaboration with dozens of newsrooms

Adriane Burnett reads to her son Karter Robinson on Saturday, April 14, 2024 in Birmingham, Ala. Women's participation in the American workforce has reached a high point, but challenges around child care are holding back many working class parents. When women without college degrees face an interruption in child care arrangements – whether it's at a relative's home, a preschool or a daycare center – they are more likely to have to take unpaid time or to be forced to leave their jobs altogether, according to an Associated Press analysis. AP PHOTO / BUTCH DILL

By SHARON LUYRE AND MORIAH BALINGIT

New Orleans-based data reporter Sharon Luyre analyzed hard-to-use Census microdata to show the crisis is uniquely devastating to moms without college degrees. For those women, a day without work is often a day without pay, so every child care disruption is debilitating — and routine, in the country’s hobbled care economy.

The education team partnered with six other local and nonprofit newsrooms to report the story. Washington, D.C.-based early education reporter Moriah Balingit expertly wove together powerful feeds of mothers whose careers have been upended by the broken child care system.

A week before the story published, AP shared its exclusive data analysis with 30 more newsrooms in states that stood out in the analysis. About a dozen of those newsrooms ran AP’s story and some later produced their own using AP’s guidance and data. More are in the works.

Submission Links

By The Associated Press

MAY 3, 2024

Explore related services

Access the worlds stories at your fingertips

Related articles

America’s Catholic Church seeing shift toward more conservative ways

MAY 10, 2024

AP delivers leading coverage of destructive weather in Texas and Oklahoma

Story shows what it means for refugee system to rebound 

• Breaking news
• Business and finance
• Entertainment
• Religion and faith
• Paris summer games
• AP Newsroom
• AP Stylebook
• AP StoryShare
• Global Media Solutions
• AP Productions
• Branded content and advertising
• Workflow Solutions
• AP’s local news AI initiative
• Metadata services
• AP Daybooks
• How AP counts the vote
• How we declare winners
• AP VoteCast: how we survey the electorate
• By industry
• Awards and recognition
• Media center
• News values
• Supporting AP
• Corporate Archives
• AP Customer Zone

Understanding data analysis: A beginner's guide

Before data can be used to tell a story, it must go through a process that makes it usable. Explore the role of data analysis in decision-making.

What is data analysis?

Data analysis is the process of gathering, cleaning, and modeling data to reveal meaningful insights. This data is then crafted into reports that support the strategic decision-making process.

Types of data analysis

There are many different types of data analysis. Each type can be used to answer a different question.

Descriptive analytics

Descriptive analytics refers to the process of analyzing historical data to understand trends and patterns. For example, success or failure to achieve key performance indicators like return on investment.

An example of descriptive analytics is generating reports to provide an overview of an organization's sales and financial data, offering valuable insights into past activities and outcomes.

Predictive analytics

Predictive analytics uses historical data to help predict what might happen in the future, such as identifying past trends in data to determine if they’re likely to recur.

Methods include a range of statistical and machine learning techniques, including neural networks, decision trees, and regression analysis.

Diagnostic analytics

Diagnostic analytics helps answer questions about what caused certain events by looking at performance indicators. Diagnostic analytics techniques supplement basic descriptive analysis.

Generally, diagnostic analytics involves spotting anomalies in data (like an unexpected shift in a metric), gathering data related to these anomalies, and using statistical techniques to identify potential explanations.

Cognitive analytics

Cognitive analytics is a sophisticated form of data analysis that goes beyond traditional methods. This method uses machine learning and natural language processing to understand, reason, and learn from data in a way that resembles human thought processes.

The goal of cognitive analytics is to simulate human-like thinking to provide deeper insights, recognize patterns, and make predictions.

Prescriptive analytics

Prescriptive analytics helps answer questions about what needs to happen next to achieve a certain goal or target. By using insights from prescriptive analytics, organizations can make data-driven decisions in the face of uncertainty.

Data analysts performing prescriptive analysis often rely on machine learning to find patterns in large semantic models and estimate the likelihood of various outcomes.

analyticsText analytics

Text analytics is a way to teach computers to understand human language. It involves using algorithms and other techniques to extract information from large amounts of text data, such as social media posts or customer previews.

Text analytics helps data analysts make sense of what people are saying, find patterns, and gain insights that can be used to make better decisions in fields like business, marketing, and research.

The data analysis process

Compiling and interpreting data so it can be used in decision making is a detailed process and requires a systematic approach. Here are the steps that data analysts follow:

1. Define your objectives.

Clearly define the purpose of your analysis. What specific question are you trying to answer? What problem do you want to solve? Identify your core objectives. This will guide the entire process.

2. Collect and consolidate your data.

Gather your data from all relevant sources using  data analysis software . Ensure that the data is representative and actually covers the variables you want to analyze.

3. Select your analytical methods.

Investigate the various data analysis methods and select the technique that best aligns with your objectives. Many free data analysis software solutions offer built-in algorithms and methods to facilitate this selection process.

4. Clean your data.

Scrutinize your data for errors, missing values, or inconsistencies using the cleansing features already built into your data analysis software. Cleaning the data ensures accuracy and reliability in your analysis and is an important part of data analytics.

5. Uncover valuable insights.

Delve into your data to uncover patterns, trends, and relationships. Use statistical methods, machine learning algorithms, or other analytical techniques that are aligned with your goals. This step transforms raw data into valuable insights.

6. Interpret and visualize the results.

Examine the results of your analyses to understand their implications. Connect these findings with your initial objectives. Then, leverage the visualization tools within free data analysis software to present your insights in a more digestible format.

7. Make an informed decision.

Use the insights gained from your analysis to inform your next steps. Think about how these findings can be utilized to enhance processes, optimize strategies, or improve overall performance.

By following these steps, analysts can systematically approach large sets of data, breaking down the complexities and ensuring the results are actionable for decision makers.

The importance of data analysis

Data analysis is critical because it helps business decision makers make sense of the information they collect in our increasingly data-driven world. Imagine you have a massive pile of puzzle pieces (data), and you want to see the bigger picture (insights). Data analysis is like putting those puzzle pieces together—turning that data into knowledge—to reveal what’s important.

Whether you’re a business decision maker trying to make sense of customer preferences or a scientist studying trends, data analysis is an important tool that helps us understand the world and make informed choices.

Primary data analysis methods

Quantitative analysis

Quantitative analysis deals with numbers and measurements (for example, looking at survey results captured through ratings). When performing quantitative analysis, you’ll use mathematical and statistical methods exclusively and answer questions like ‘how much’ or ‘how many.’ 

Qualitative analysis

Qualitative analysis is about understanding the subjective meaning behind non-numerical data. For example, analyzing interview responses or looking at pictures to understand emotions. Qualitative analysis looks for patterns, themes, or insights, and is mainly concerned with depth and detail.

Data analysis solutions and resources

Turn your data into actionable insights and visualize the results with ease.

Microsoft 365

Process data and turn ideas into reality with innovative apps, including Excel.

Importance of backing up data

Learn how to back up your data and devices for peace of mind—and added security. 

Copilot in Excel

Go deeper with your data using Microsoft Copilot—your AI assistant.

Excel expense template

Organize and track your business expenses using Excel.

Excel templates

Boost your productivity with free, customizable Excel templates for all types of documents.

Chart designs

Enhance presentations, research, and other materials with customizable chart templates.

Follow Microsoft

You are using an outdated browser. Upgrade your browser today or install Google Chrome Frame to better experience this site.

• Professional learning

Teach. Learn. Grow.

Teach. learn. grow. the education blog.

Summer learning loss: What we know and what we’re learning

Concerns about students losing ground academically during summer break go back at least a century, with early evidence suggesting that summer contributed to large disparities in students’ outcomes. This narrative spurred expansion of a variety of summer programs and interventions aimed at stemming summer learning loss.

However, in the last five years, there has been a spirited debate about two long-standing questions about students’ summers: 1) the degree to which test scores actually drop during the summer and 2) the degree to which summer break contributes to educational inequities. A new layer to this conversation is the response to the learning disruptions caused by the COVID-19 pandemic. School leaders and policymakers have used the summer break as a potential time for academic recovery. Summer programs have emerged as one of the most popular recovery strategies offered by school districts, with an estimated $5.8 billion of ESSER funds expected to be spent on summer programs by September 2024.

With more focus on the impact of summer on students’ learning and the potential to extend the school year, it is essential for educators, policymakers, and families to have an up-to-date understanding of the impact of summer breaks on students’ learning patterns. In this post, we aim to highlight what is known about summer learning loss by quickly summarizing recent research and posing some questions that remain unanswered about the role of summers on students’ learning.

Students’ test scores flatten or drop during the summer

While our initial understanding of summer learning loss dates back to studies conducted in the 70s and 80s , a recent collection of studies in the last six years provides a fresh look at students’ learning across summers using four modern assessments ( ECLS-K direct cognitive tests , MAP® Growth™, Star, and i-Ready) with large national (though not typically nationally representative) samples. See “School’s out: The role of summers in understanding achievement disparities,” “When does inequality grow? School, summer, and achievement gaps,” “Evidence of ‘summer learning loss’ on the i-Ready diagnostic assessment,” “Findings on summer learning loss often fail to replicate, even in recent data,” and “Inequality in reading and math skills forms mainly before kindergarten: A replication, and partial correction, of ‘Are schools the great equalizer?’”

Figure 1 compares the test score patterns across four different studies. Three important patterns stand out:

• On average, test scores flatten or drop during the summer , with larger drops typically in math than reading.
• Studies using test scores from ECLS-K:2011 show that student learning slows down but does not drop over the summers after kindergarten and first grade. However, research using interim and diagnostic assessments ( MAP Growth , Star, and i-Ready ) has found far larger summer drops across a range of grade levels.
• Given the sizable differences in the magnitude of test score drops across tests, it remains uncertain whether summer slide should be considered a trivial issue or a serious educational challenge.

Figure 1. Comparison of summer slide estimates across datasets

Note: All estimates are reported as the total average summer test score change in standard deviation (SD) units relative to the prior spring test score. Whenever possible, we report the estimate that adjusted scores for time in school prior/after testing in the fall and spring. Sources: Author calculations based on data reported in ECLS-K:20210-11 , MAP Growth , i-Ready , and Star .  

Who is most likely to show summer learning loss.

While all three diagnostic assessments show some degree of summer slide in grades 3–8 on average, the research community lacks consensus about whether summers disproportionately impact certain students. Paul von Hippel and colleagues have pointed out that whether and how much summers contribute to educational inequalities (across students of different income levels, races, ethnicities, and genders) depends on the test used to study students’ learning patterns. Nonetheless, we can present a few key patterns from this line of research:

• Learning rates are more variable during the summer than during the school year. See “School’s out: The role of summers in understanding achievement disparities,”   “When does inequality grow? School, summer, and achievement gaps,”  and  “Inequality in reading and math skills forms mainly before kindergarten: A replication, and partial correction, of ‘Are schools the great equalizer?’”
• Gaps between students attending low- and high-poverty schools do not consistently widen during the summer. See “Is summer learning loss real, and does it widen test score gaps by family income?”  and  “Is summer learning loss real?”
• Test score differences between Black and white students hold steady or narrow during the summer. See “Do test score gaps grow before, during, or between the school years? Measurement artifacts and what we know in spite of them”  and  “When does inequality grow? School, summer, and achievement gaps,” though results can be sensitive to the metric and test used. See also  “Black-white summer learning gaps: Interpreting the variability of estimates across representations” and “Findings on summer learning loss often fail to replicate, even in recent data.”
• The field cannot really explain why differences in students’ summer learning occur. See “Rethinking summer slide: The more you gain, the more you lose”  and  “Inequality in reading and math skills forms mainly before kindergarten: A replication, and partial correction, of ‘Are schools the great equalizer?’”

Planning effective summer programming

It is clear across recent studies that summer is a particularly variable time for students. Summer break is also increasingly a time in which districts are offering a range of academic offerings.

During summer 2022, an estimated 90% of school districts offered summer programs with an academic focus. However, evidence on the effectiveness of academic summer programs during and after the COVID-19 pandemic is limited. One study of eight summer programs in summer 2022 found a small positive impact on math test scores (0.03 SD), but not on reading. The improvements in math were largely driven by elementary students compared to middle schoolers. However, the effectiveness of these programs remained consistent across student groups, including race/ethnicity, poverty, and English learner status.

It is crucial to recognize the challenges associated with scaling up summer programs. In the districts studied, only 13% of students participated in the summer programs , which only lasted for an average of three to four weeks. Prior research indicates that for summer programs to yield measurable academic benefits, they should run at least five weeks with at least three hours of instruction a day. Additionally, getting students to regularly attend summer programs remains a significant hurdle. To address this issue, districts should actively recruit families to participate and offer a mix of academic instruction and engaging extracurricular activities. By adopting these strategies, districts can maximize the effectiveness of their summer programs and better support student learning during the break.

If you’re interested in learning more about effective summer programs, we encourage you to read the following:

• “Effective summer programming: What educators and policymakers should know”
• “Investing in successful summer programs: A review of evidence under the Every Student Succeeds Act”
• “Analysis: Summer learning is more popular than ever. How to make sure your district’s program is effective”
• “The impact of summer learning programs on low-income children’s mathematics achievement: A meta-analysis”
• “The effects of summer reading on low-income children’s literacy achievement from kindergarten to grade 8: A meta-analysis of classroom and home interventions”

Recommended for you

From the pages of  Literacy Today : How to boost reading achievement using dyslexia research

NWEA research snapshot: Insights on academic recovery strategies

How AI can improve digital accessibility in math

White Paper

Worth the investment: Why research and innovation matter for improving outcomes for all kids

Making sound, student-focused decisions amid constant change can feel impossible. Partnering with a research-based assessment provider can help.

Read the white paper

STAY CURRENT by subscribing to our newsletter

You are now signed up to receive our newsletter containing the latest news, blogs, and resources from nwea..

Manufacturing Design Engineer - Machine Learning/Data Analysis

Add a favorite.

Don’t have an Apple ID?

• Create one now
• Forgot your Apple ID or password?

Key Qualifications

• Key expertise: Machine Learning/ computer science / data analysis / manufacturing process
• - Experienced in building, deploying and running Machine Learning applications or services
• - Experience working with very large scale of data, familiar with data processing framework like Spark
• - Strong software development skills with proficiency in Python / C++
• - Hands-on experience working with deep learning toolkits such as PyTorch or TensorFlow/JAX
• - Deep technical skills in machine learning, computer vision, natural language processing, or artificial intelligence
• - A real passion for making simple, robust, and scalable platforms used by other engineering teams
• - Experiences of process technical development for general manufacturing process (or very familiar with some of),;
• - Familiar with GD&T, DOE, SPC.
• - Good written and oral communication skills, fluent in English.
• - On-site build support; must be willing to work from OEM factories.
• - Ability and willingness to work independently and travel frequently (mainly in Asia).
• - A team player who is self-motivated, results oriented and detailed oriented.

Description

Education & experience, additional requirements.

UMass Donahue Institute

Massbenchmarks.

A program of the Economic and Public Policy Research Group at the UMass Donahue Institute in collaboration with the Federal Reserve Bank of Boston

• About MassBenchmarks
• MassBenchmarks Journal

Current and Leading Indexes

• Notes from the Board
• Special Releases
• Subscribe to MassBenchmarks
• Contact Us: MassBenchmarks

State economic growth slows in Q1, UMass journal reports

Continued strong labor market conditions and income growth reflect resilience of state economy

»  Download the detailed report

In the first quarter of 2024, Massachusetts real gross state product (GDP) increased at an annual rate of 1.8 percent, according to MassBenchmarks, while U.S. GDP increased at an annual rate of 1.6 percent, according to the U.S. Bureau of Economic Analysis (BEA). According to the BEA, in the fourth quarter of last year, Massachusetts GDP grew at an annual rate of 3.0 percent while U.S. GDP grew at a 3.4 percent rate.

The annual benchmark revisions to payroll employment indicate a much slower pace of job growth in the Commonwealth as compared to the nation during 2023. For the period December 2022 to December 2023, the Bureau of Labor Statistics (BLS) revised job growth in Massachusetts down from 1.9 percent to 0.7 percent. This compares to 2.0 percent national job growth during the same period.

Consistent with those revisions, the Bureau of Economic Analysis’ (BEA) estimate of growth in economic output in 2023 entails notably slower GDP growth for Massachusetts than for the U.S. From the fourth quarter of 2022 to the fourth quarter of 2023, the BEA reports that Massachusetts GDP grew 1.9 percent while U.S. GDP grew 3.1 percent. Although the state and the nation grew at about the same pace in the first quarter of 2024, the growth figures for that quarter reflect a slowdown for the U.S. and a continuation of relatively moderate growth in Massachusetts.

Nevertheless, the indicators for the first quarter show that the state’s economy seems fairly healthy on balance. In that quarter payroll employment in Massachusetts grew 2.2 percent on an annualized basis, just above the national growth rate of 2.0 percent, and well up from the state’s (annualized) job growth pace of 0.7 percent in the fourth quarter of 2023. Relative to the first quarter of 2023, the number of jobs in the state was just 0.6 percent higher one year later. For the U.S., annualized job growth in the fourth quarter was 1.6 percent, and the number of jobs increased 1.8 percent on net from the first quarter of 2023 through the first quarter of 2024.

State tax revenues are consistent with robust income growth and spending. Based on seasonally adjusted personal withholding taxes, wage and salary income grew at an annual rate of 12.3 percent in the first quarter of 2024 and ended the quarter 10.1 percent higher than a year earlier. Month-to-month and even quarter-to-quarter in recent years, withholding tax revenues have been volatile.

“The strength in the first quarter in part reflects timing; withholding taxes in both the fourth quarter of 2023 and the first quarter of 2023 were both relatively weak,” noted Alan Clayton-Matthews, Senior Contributing Editor and Professor Emeritus of Economics and Public Policy at Northeastern University, who compiles and analyzes the Current and Leading Indexes for MassBenchmarks. “Those low denominators mean that the pace of income growth, while strong, should be interpreted cautiously,” Clayton-Matthews added.   

The level of withholding tax revenues in the first quarter of this year is consistent with continuing state income growth. The BEA estimates that Massachusetts wage and salary income grew 3.6 percent in the fourth quarter of 2023 and offers a clearer view of recent trends in state income growth.

U.S. wage and salary income also continued to grow robustly in the first quarter, at a 6.4 percent annualized rate. In the fourth quarter of last year, this income grew 4.5 percent, and growth from the first quarter of last year to the first quarter of this year was 5.8 percent.

Spending on items subject to Massachusetts regular and motor vehicle sales taxes grew at a robust 8.6 percent annual rate in the first quarter, after growing 12.7 percent in the fourth quarter of last year. Relative to the first quarter of last year, this spending grew a mere 0.1 percent. This spending largely reflects spending on durable goods.

The labor market continues to look strong, with low unemployment rates and low initial unemployment claims. The headline (U-3) unemployment rate in Massachusetts in March was 2.9 percent, as compared to 3.8 percent nationally. The lower rate for Massachusetts largely reflects demographics. Educational attainment is higher in Massachusetts relative to the U.S., and unemployment rates tend to be lower in more highly educated states.

One indicator suggests some potential softening in the labor market. The U-6 unemployment rate in Massachusetts ticked up to 8.0 percent in March from 6.8 percent in December and 6.2 percent in March of last year. This measure counts more people as unemployed than the U-3 measure by including those who only have part-time work but want full-time work, and those who have given up looking for work but want to work. In Massachusetts, the rise reflects an increase in this latter group. For the U.S., the U-6 rate in March was 7.3 percent, up from 6.7 percent in March 2023.

Inflation remains stubbornly high in the Boston metro area. According to the BLS’ headline consumer price index (CPI-U), prices rose at a 4.3 percent rate in the first quarter of this year relative to the fourth quarter of last year, with core prices – which exclude food and energy – up 6.4 percent in the first quarter. This represents an acceleration in inflation from the fourth quarter. From the first quarter of last year, prices were up 2.7 percent for all items, and up 3.7 percent for core items. The U.S. shows a similar pattern in the acceleration of inflation, though in the first quarter national inflation was lower than in Boston: 3.8 percent for all items, and 4.2 percent for core items.

The outlook calls for steady, but slow growth in the state in the next two quarters. The Massachusetts leading index is predicting annualized growth of 3.1 percent in the second quarter and 0.5 percent in the third quarter of 2024. Economists surveyed by the Wall Street Journal in early April have an average expectation of U.S. growth in the second quarter of 1.6 percent and in the third quarter of 1.4 percent.

Business Group(s)

Contracting with us.

The UMass Donahue Institute is pre-approved to offer a variety of options for streamlined, simplified contracting both within Massachusetts and at the federal level.

University of Massachusetts Amherst

100 Venture Way, Suite 9

Hadley, Massachusetts 01035

(413) 545-0001

Contact us by email.

Related Links

• Connected Beginnings Training Institute
• MassHire Greater Brockton Career Center
• UMass Donahue Institute GSA Services

Campus Links

• UMass System
• UMass Amherst
• UMass Boston
• UMass Dartmouth
• UMass Lowell
• Medical School
• UMass Online

Receive our Newsletter

Connect with us.

• Formerly Twitter

Advertisement

Where Protesters on U.S. Campuses Have Been Arrested or Detained

By The New York Times

Police officers and university administrators have clashed with pro-Palestinian protesters on a growing number of college campuses in recent weeks, arresting students, removing encampments and threatening academic consequences. More than 2,700 people have been arrested or detained on campuses across the country.

Campus protests where arrests and detainments have taken place since April 18

The fresh wave of student activism against the war in Gaza was sparked by the arrests of at least 108 protesters at Columbia University on April 18, after administrators appeared before Congress and promised a crackdown. Since then, tensions between protesters, universities and the police have risen, prompting law enforcement to take action in some of America’s largest cities.

An earlier version of this article misstated the number of arrests at Princeton University. There have been 15 arrests, not 14.

• Share full article

Our Coverage of the U.S. Campus Protests

News and Analysis

U.S.C. : The University of Southern California’s academic senate voted to censure Carol Folt , the school’s president, after several tumultuous weeks, including canceling the valedictory address of a Muslim student, clearing a protest encampment and calling in police to arrest dozens of protesters.

G.W.U. : Hours before the mayor of Washington, D.C., was scheduled to testify on Capitol Hill about the city’s handling of a pro-Palestinian encampment at George Washington University, police moved to break up the encampment .

U.C.L.A. : A police consulting firm will review a violent confrontation  at the University of California, Los Angeles, in which a group of counterprotesters attacked demonstrators  at a pro-Palestinian encampment while security guards and police officers failed to intervene.

An Agreement to Divest :  Students who oppose the war in Gaza began dismantling their protest camp  at Trinity College Dublin in Ireland, after the institution agreed to divest from three Israeli companies.

Republican Hypocrisy:  Prominent Republicans have seized on campus protests to assail what they say is antisemitism on the left. But for years they have mainstreamed anti-Jewish rhetoric .

Remembering the 1968 Protests:  As Chicago prepares to host the Democratic National Convention , it wants to shed memories of chaos from half a century ago even as the campus protests are growing.

Outside Agitators:  Officials in New York City have blamed “external actors” for escalating demonstrations at Columbia, but student protesters reject the claim .