structured literature review deutsch

Kostenlose Rechtschreibprüfung

Kostenlose plagiatsprüfung, korrektur deiner bachelorarbeit.

• Wissensdatenbank
• Aufbau und Gliederung

Wie führt man einen Literatur-Review durch?

Veröffentlicht am 9. November 2015 von Annelien Krul . Aktualisiert am 15. August 2023.

Wenn du eine Abschlussarbeit schreibst, ist es unbedingt nötig, dass du einen Literatur-Review durchführst. Aber was genau bedeutet das? Wie strukturierst du in erster Linie diesen Review und wie integrierst du anschließend die Informationen, die du findest, in deine Arbeit?

Scribbrs kostenlose Rechtschreibprüfung

Fehler kostenlos beheben

Inhaltsverzeichnis

Was ist ein literatur-review, literatur-wegweiser, wie verweist du auf deine quellen, standardsätze die sie verwenden können, häufig gestellte fragen.

Bei einem Literatur-Review handelt es sich um eine Methode, die angewandt wird, um Wissen, das bereits in Bezug auf ein bestimmtes Thema oder Problem besteht, zu sammeln. Diese Informationen können in verschiedenen Quellen, wie Zeitschriftenartikel, Bücher, Papers, Abschlussarbeiten und Archivmaterial, gefunden werden.

Durch die Durchführung eines Literatur-Reviews kannst du dir einen Einblick in bereits bestehende Kenntnisse und Theorien in Bezug auf dein Thema verschaffen. Dies stellt außerdem sicher, dass deine Abschlussarbeit über eine starke wissenschaftliche Fundierung verfügt.

Wenn er richtig durchgeführt wird, entsteht aus einem Literatur-Review nicht nur eine einfache Liste oder Zusammenfassung der verfügbaren Daten. Dein Ziel ist es stattdessen, die relevantesten Ideen und Informationen, die du herausgefunden hast, im Rahmen deines theoretischen Rahmens kritisch zu diskutieren.

Was ist der Unterschied zwischen einem Literatur-Review und einem systematic Review?

Ein systematic Review ist systematischer und formaler als ein Litertur-Review.

Bei einem Literatur-Review werden existierende Arbeiten qualitativ zusammenfasst und evaluiert, ohne dass eine formale, explizite Methode verwendet wird.

Was ist der Unterschied zwischen einem Literatur-Review und einem theoretischen Rahmen?

Der Literatur-Review dient als echter Grundstein für die Analyse des Problems, das du untersuchst. Abhängig vom Aufbau deiner Abschlussarbeit kann er auch als Grundlage verwendet werden, um einen umfassenden theoretischen Rahmen zu entwickeln.

Beispiel theoretische Rahmen

Vorbereitende Erforschung des Problems

Sobald du einen allgemeinen Überblick über das Problem und die Forschungsfragen, die du in deiner Abschlussarbeit ansprechen möchtest, hast, ist der erste Schritt oft, mit einem Literatur-Review zu beginnen. Dies ist eine wertvolle Möglichkeit, um dich selbst innerhalb des Forschungsfeldes besser zu orientieren und dich auf das Problem, das du untersuchen willst, genauer auszurichten.

Diese Einblicke in die bestehenden Kenntnisse und Theorien, die du durch den Literatur-Review gewinnst, werden dir auch dabei helfen, einen starken wissenschaftlichen Ausgangspunkt für den Rest deiner Forschung zu schaffen.

Nachdem du eine klare Problemstellung und Forschungsfrage(n) bestimmt hast, ist der nächste Schritt, dass du dich eingehender in dein Thema und die einschlägige Literatur vertiefst.

Dies kann eine echte Herausforderung sein, angesichts der Menge der verfügbaren Literatur und der begrenzten Zeit, die dir unter Umständen für das Schreiben deiner Abschlussarbeit zur Verfügung steht. Deshalb ist es wichtig, dass du an diesen Prozess so effizient und systematisch wie möglich herangehst.

Kostenlos auf Plagiate prüfen.

Hältst du dich an diesen Vier-Stufen-Wegweiser, wird dir das dabei helfen, deinen Literatur-Review effektiv durchzuführen.

• Vorbereitung
• Literatur sammeln
• Literatur bewerten und auswählen
• Literatur verarbeiten

1. Vorbereitung

Der erste Schritt umfasst, dass du dich selbst auf dein Thema ausrichtest, damit du ein umfassenderes Bild des Untersuchungsgebietes erlangst. Dazu gehört auch die Erstellung einer Liste von Schlüsselbegriffen, die als Grundlage für den nächsten Schritt dient.

• Lies eine aktuelle Publikation zu deinem Thema. Wähle eine Publikation von einem renommierten Autor, in der alle (oder möglichst viele) Facetten deines Themas erörtert werden. Dadurch erhältst du einen allgemeinen Überblick über dein Forschungsfeld. Frage deinen Betreuer, falls Fragen auftauchen.
• Notiere während des Lesens die Begriffe, die dir am wichtigsten/relevantesten erscheinen.
• Identifiziere auch die wichtigsten/relevantesten Begriffe deines (Ausgangs-) Problems und deiner Forschungsfrage(n).
• Nonverbale Kommunikation
• Nonverbale Signale
• Nonverbale Gesten
• Körpersprache

2. Literatur sammeln

Diese Vorbereitungsarbeit macht es nun viel einfacher, nach spezifischer Literatur und nach anderen Quellen zu suchen. Diese Suche beginnt oftmals online. Dabei ist es sehr wichtig, dass die richtigen Schlüsselbegriffe verwendet werden, weshalb der erste Schritt dieses Wegweisers die Zusammenstellung einer Liste umfasst.

Suche nach diesen Begriffen sowohl in deutscher Sprache als auch in allen anderen Sprachen, in denen du über Lesekompetenzen verfügst. Es ist auch hilfreich zu versuchen, Synonyme und verschiedene Kombinationen von Begriffen zu verwenden.

Es gibt viele verschiedene Arten von Datenbanken, die du vielleicht durchforsten möchtest:

• Der Online-Katalog deiner Schule oder der Universitätbibliothek. Die meisten wissenschaftlichen Bibliotheken haben einen großen Bestand an physischen Ressourcen, einschließlich Büchern, Papers, Zeitschriften und Magazinen. Die meisten haben jedoch ihre Angebote exponentiell durch das Abonnieren wissenschaftlicher Ressourcen, einschließlich Fachzeitschriften sowie wissenschaftlicher Datenbanken (siehe unten), erweitert.
• Google Scholar. Über www.scholar.google.de gelangst du zu der speziellen Suchmaschine von Google für wissenschaftliche Literatur. Wenn du auf einen Artikel, an dem du interessiert bist, nicht kostenlos zugreifen kannst, versuche stattdessen über die Bibliothek deines Instituts darauf zuzugreifen.
• Länderspezifische Datenbanken. Manche Datenbanken werden auf nationaler Ebene betrieben. Beispielsweise kann auf die Datenbank PiCarta fast immer über Bibliotheken niederländischer Institute zugegriffen werden. Sie enthält Daten zu allen verfügbaren Publikationen in den Niederlanden, einschließlich Büchern und Magazinen, die nicht in deiner eigenen Bibliothek verfügbar sind.
• Fachübergreifende Datenbanken. Datenbanken wie JSTOR und EBSCO sind digitale Bibliotheken, die Fachzeitschriften, Bücher und Primärliteratur zu einer Vielzahl von Themen beinhalten. Die meisten Institutsbibliotheken abonnieren mehrere davon.
• Fachspezifische Datenbanken. Mehrere Datenbanken konzentrieren sich auf bestimmte Disziplinen (oder Gruppen von verwandten Disziplinen). Ein Beispiel hierfür ist die AGRIS-Datenbank, die ein breites Spektrum an Themen in Bezug auf Landwirtschaft und Umwelt abdeckt.

Wenn du eine nützliche Quelle gefunden hast, prüfe die Biografie dieser Publikation auf andere relevante Quellen (dies wird als „Schneeball“-Forschungsmethode bezeichnet). Taucht ein Autorname immer wieder auf? Dann bedeutet das normalerweise, dass diese Person eine umfangreiche Forschung zu diesem Thema geleistet hat.

Ein Blick auf seine/ihre Webseite oder die Suche nach seinem/ihrem Namen direkt in einem (Online-) Katalog wird wahrscheinlich zu weiteren Ergebnissen führen.

3. Literatur bewerten und auswählen

Es ist wahrscheinlich, dass du eine überwältigende Menge an Literatur entdeckst. Da dir nur ein begrenztes Ausmaß an Zeit zur Verfügung steht, ist es wichtig, dass du dich auf die wichtigsten Quellen konzentrierst. Wir schlagen vor, die Literatur, die du gefunden hast, zuerst in Hinblick auf ihre Relevanz und anschließend auf ihre wissenschaftliche Qualität zu bewerten.

Relevanz Eine relevante Publikation ist eine, die gut zu deinem Thema oder deinem Problem passt. Um die Relevanz eines Buches oder Artikels, ohne es/ihn in seiner Gesamtheit zu lesen, zu bestimmen, beginne nur mit der Einleitung und dem Fazit . Dies wird dir häufig genügend Informationen liefern, um zu beurteilen, ob die Publikation für deine Arbeit relevant ist.

Qualität Die Qualität einer Publikation wird durch eine Reihe von Faktoren bestimmt. Als allgemeine Regel gilt: Versuche nur die Artikel, die in renommierten Fachzeitschriften veröffentlicht wurden, zu verwenden. Rankings wie die Journal Quality List helfen dir dabei herauszufinden, welche Zeitschriften qualitativ sind.

Ein Blick auf die Fachkenntnisse mitwirkender Autoren kann ebenfalls hilfreich sein. Fachkundige Autoren sind normalerweise einem akademischen Institut zugehörig, publizieren umfangreich und werden häufig von anderen zitiert.

Bedenke, dass Informationen von Webseiten, mit Ausnahme von Webseiten, die von wissenschaftlichen, staatlichen oder zwischenstaatlichen Institutionen betrieben werden, häufig nicht zuverlässig sind. Es ist auch wichtig, wenn möglich die neueste Literatur zu nutzen; wenn du das nicht machst, läufst du Gefahr, deine Arbeit auf veralteten Informationen aufzubauen.

4. Literatur verarbeiten

Sobald du die Literatur, auf die du dich konzentrieren willst, bestimmt hast, findet im nächsten Schritt die Verarbeitung der Informationen, die du herausgefunden hast, statt (zum Beispiel durch eine Problemanalyse oder einen theoretischen Rahmen). Es ist natürlich wichtig, damit zu beginnen, die ausgewählten Publikationen gründlich zu studieren. Stell dir dabei die folgenden Fragen:

• Was ist das zu untersuchende Problem und wie nimmt es die Forschung in Angriff?
• Was sind die Schlüsselkonzepte und wie werden diese definiert?
• Welche Theorien und Modelle verwendet der Autor?
• Was sind die Ergebnisse und Schlossfolgerungen der Studie?
• Wie steht diese Publikation verwandten Publikationen innerhalb dieses Forschungsfeldes gegenüber?
• Wie kann ich diese Forschung auf meine eigene anwenden?

Die Analyse all deiner Quellen auf diese Art und Weise wird dir ein klares Bild des Forschungsfeldes und wie deine Forschung dazu passt, vermitteln. Danach bist du dazu in der Lage, die Literatur auf eine kritische und gut fundierte Weise zu diskutieren.

Die Art und Weise, wie du die Ergebnisse deines Literatur-Reviews präsentieren sollst, variiert je nach Programm. Die entsprechenden Leitlinien deines Fachgebiets sollten dir zur Verfügung gestellt werden. Wenn du den Literatur-Review verwendest, um beispielweise einen theoretischen Rahmen vorzubereiten, würde der Schwerpunkt auf der Definition und Analyse von Theorien und Modellen liegen.

Die Zitate in deinem Literatur-Review sollten sehr genau sein. Viele Schulen und Universitäten nutzen die Zitierweise der American Pscyhological Association ( APA ). Nutze den kostenlosen Scribbr-Zitiergenerator , der dir dabei hilft, Zitate schnell und korrekt zu erstellen.

Wenn du deine Quellen nicht richtig zitierst, werden deine verwendeten Informationen als Plagiate gewertet. Plagiarismus ist eine schwere Form des Betrugs, der schwerwiegende Folgen hat. Hast du dabei Zweifel oder brauchst du Hilfe? Führe eine Plagiatsprüfung durch – Vorsicht ist immer besser als Nachsicht.

• Aus vorgehenden Untersuchungen wurde ersichtlich, dass…
• In mehreren Studien (Smith, 1988; Driessen, 2007) wurde aufgezeigt, dass…
• Durch Studien (Smith, 1988; Driessen, 2007) zu X wird die Wichtigkeit von…hervorgehoben
• Frühere Untersuchungen (Smith, 1988) zum Vergleich von X und Y zeigen, dass…
• Studien, wie jene von Smith (1988), zeigen, dass…

Mehr standardsätze »

• Previous studies have shown that…
• Several studies (Smith, 1988; Driessen, 2007) have looked at…
• Studies (Smith, 1988; Driessen, 2007) of X shows the importance of…
• Previous research (Smith, 1988) comparing X and Y has found…
• Research carried out by Smith (1988) indicated that …

Ist deine Bachelorarbeit fehlerfrei?

Durchschnittlich enthält eine Bachelorarbeit 150 Fehler pro 1.000 Wörter .

Neugierig? Bewege den Regler von links nach rechts!

Zu deiner Korrektur

Beachte diese vier Schritte, um einen Literatur-Review zu schreiben:

Für Inhalte von Theorien, Definitionen und Fakten benutzt du im Literatur-Review den Indikativ Präsens. Abgeschlossene Studien und Ergebnisse von anderen Untersuchern werden im Indikativ Perfekt geschrieben.

Diesen Scribbr-Artikel zitieren

Wenn du diese Quelle zitieren möchtest, kannst du die Quellenangabe kopieren und einfügen oder auf die Schaltfläche „Diesen Artikel zitieren“ klicken, um die Quellenangabe automatisch zu unserem kostenlosen Zitier-Generator hinzuzufügen.

Krul, A. (2023, 15. August). Wie führt man einen Literatur-Review durch?. Scribbr. Abgerufen am 23. April 2024, von https://www.scribbr.de/aufbau-und-gliederung/wie-fuehrt-man-einen-literatur-review-durch/

War dieser Artikel hilfreich?

Annelien Krul

Annelien ist Kunsthistorikerin mit einer Leidenschaft für die Sprache. Sie hat ein eigenes Redaktionsbüro und arbeitet nebenbei mit viel Freude für Scribbr, um Studenten mit deren Abschlussarbeiten zu helfen.

Das hat anderen Studierenden noch gefallen

Theoretischer rahmen einer abschlussarbeit, beispiel: theoretischer rahmen einer abschlussarbeit, qualitative forschung und quantitative forschung, aus versehen plagiiert finde kostenlos heraus.

Structured literature reviews – A guide for students

This is a step-by-step guide aimed at Master's students undertaking a structured literature review as part of their Master's thesis.

There are several different kinds of literature reviews, but any literature review typically includes an extensive literature search. Whenever a systematic approach is used, the literature search features a methodical step-by-step procedure. However, as a Master's student, it might not be possible to fulfill all the criteria of a systematic review when writing a literature review-based thesis; you should rather do a structured literature review, which will include only certain aspects of the systematic review methodology.

In this guide we will go through the different steps of a structured literature review and provide tips on how to make your search strategy more structured and extensive. Additionally, make sure to follow any programme and course specific requirements.

Step 1: Formulate and delimit your research question

• It will be much easier for you to perform a structured information search if you first define and delimit your research question in a clear way.
• One way to define and structure your question is to break it down into different parts .
• PICO and PEO are two different frameworks that can be used for breaking down a research question into different parts.
• You also need to define the most important key concepts of your research question.

The formulation of your research question is partly connected to what kind of literature review you are doing. This article by Maria J. Grant and Andrew Booth usefully compares different kinds of reviews . While a systematic literature review is usually grounded in a clearly delimited and structured question, a scoping review may, for instance, feature a wider problem formulation. The wider a research question is, the larger number of search hits it tends to generate.

To be able to perform a literature review, you need to consider a subject area in which there seems to be a sufficiently large number of original research studies. Therefore, it may be a good idea to test search a database for previous research on the subject while you are trying to formulate and delimit your research question.

One way to structure your research question is to break it down into different parts. A well-delimited question often consists of three to four different parts. PICO and PEO are two examples of frameworks that can help you identify and define your research question.

• PICO ( P opulation, I ntervention, C omparison, O utcome) is primarily used for quantitative research questions.
• PEO ( P opulation, E xposure, O utcome) is primarily used for qualitative research questions.

Structuring your research question in accordance with a framework, such as PICO or PEO, will also help you decide on the inclusion and exclusion criteria of your literature review.

PICO & PEO

After you have delimited your research question, you also need to identify the key concepts that make up your question. Based on these key concepts, you will create " search blocks " that you will use to organise your search terms.

Step 2: Find search terms and create search blocks

• Test searching is a good way to investigate the terminology of a subject area and find search terms.
• Reading key articles can help you gather additional search terms for your final search strategy.
• Find subject headings  for PubMed with the help of the US National Library of Medicine's MeSH database.
• Find& free-text search terms by investigating what words that occur in the title and abstract of relevant articles.
• A good way of achieving a structured final search query is to arrange your search terms into search blocks ; these blocks should arise from the key concepts of your research question.

While working on a literature review-based thesis, you will need to search for articles on several occasions. In the beginning of your project, it is often good to do a couple of unstructured and simple search queries, so-called test searches, in academic databases. This way you are off to a good start, as test searching helps you investigate the terminology of your subject area and find relevant search terms. While the final search strategy is typically reported in full, you don't need to present your test queries in your thesis.

Try to find a couple of key articles, that is, articles that correspond to the type of studies that you are planning to include in your review. Use key articles to gather additional search terms for your final search strategy. Analyse the terminology of your key articles by examining what subject headings (MeSH terms, etc.) that the articles have been tagged with and what words that occur in the titles and abstracts.

Test search - find keywords and narrow down your topic

Test search in pubmed & cinahl.

To retrieve as many relevant studies as possible, you will need to include free-text search terms as well as subject headings in your final search strategy. Free-text search terms are words that occur in the article's title and abstract – words used by the authors themselves. Subject headings are subject-related words that an article is tagged with when the article is added to the database.

• In PubMed , articles are tagged with MeSH terms ( Me dical  S ubject  H eadings). You can look up and browse MeSH terms in the US National Library of Medicine's  MeSH database .
• Databases such as CINAHL , PsycInfo , ERIC , and Sociological Abstracts  have their own subject heading lists; look up subject headings in each database's subject heading list.
• There are also so-called free-text databases, such as Web of Science . These databases lack subject heading lists. Hence, when searching a free-text database, you can only use free-text search terms.

Find subject headings

An effective way to increase the structure of your final search strategy is to arrange your search terms in so-called  search blocks . Create your search blocks based on the key concepts of your research question.

Create search blocks

This search strategy worksheet might help you document and organise your search terms.

Download worksheet

• Worksheet for search terms (Word, 30.54 KB)

Step 3: Search in a structured way

• To get a comprehensive search result, you will need to search for articles in  several different databases .
• Your search strategy should be as uniform as possible in every database, but you may have to  adapt your use of subject headings .
• As you search the databases, combine your search terms and blocks with the help of  AND  and  OR .
• Save time by  documenting your search queries .

When doing a literature review-based thesis it is often wise to use at least two different databases. Many databases overlap, but may also contain unique content. At KI it is common for Master's students to use PubMed and Web of Science when doing a structured literature review as part of their Master's thesis. Depending on your research question, other databases may also be appropriate and useful. Read more about the most frequently used databases at KI .

Your search strategy should be as uniform as possible in every database. However, as mentioned in Step 2, databases may use different subject headings, and some databases only let you use free-text search terms. This means that you need to adapt your use of subject headings depending on the database.

Example: How subject headings may differ between databases 

If you want to search for articles about day surgery in PubMed, you should use the MeSH term Ambulatory Surgical Procedures . However, if you also want to perform your search in a database such as CINAHL, you need to use the corresponding CINAHL Headings term instead: Ambulatory Surgery .

There are many different ways of searching databases. Most databases have one simple, basic Google-like search box and one advanced search form. One advantage of the latter is that combining search terms with AND and OR is usually easier in an advanced search form, especially if you will be using both AND and OR within the same search query. However, you can often combine search terms with AND and OR in a basic search box too, and in that case, you often isolate your different search blocks from each other by enclosing each block in parentheses.

Example: A search query that contains AND, OR, and parentheses

( inflammatory bowel diseases OR ulcerative colitis OR crohn disease) AND (adolescent OR child OR young adult OR teenager) AND (self-management OR self care OR self efficacy )

By choosing the advanced search form you will also be able to exert more control over your search process. The advanced search form lets you specify more closely and decide exactly how you want the database to interpret your search terms; this way you can make your search query more precise.

You should always document your search strategy in order to remember what search terms you have used, how these search terms have been combined, and whether you have applied any limits to your search. The easiest way to do this is to copy and paste your search history from the database into a text document. Also, academic databases often let you create a personal account, so that you may save your searches online.

How to do a structured search in PubMed

Step 4: narrowing or broadening your search.

• Briefly examine your search results to see if you need to narrow or broaden your search query.
• Investigate whether your key articles are present in the search results.
• By using the advanced search form you can improve your search.

Prepare yourself for having to modify and redo your search query several times, before deciding on your final search strategy. After you have combined all your search terms and made your very first database search, you should examine the search results and analyse whether your search query is able to generate the type of search hits that you are looking for.

Analyse your search results

• Are all your key articles present in the search results, or are there some key articles that your search query is unable to retrieve?
• Are you getting too few search hits ? Investigate why. Perhaps you need to remove one of your search blocks, add one or several synonyms within a search block, or search for parts of words by truncating one or several of your free-text search terms, in order to broaden your search ?
• Does your search strategy generate too many non-relevant search hits that have nothing to do with your research question? Investigate why. Perhaps you need to add another search block, remove one of the synonyms from one of your search blocks, or search for phrases by enclosing one or several of your free-text terms in quotation marks, in order to narrow your search ?
• More tips on how to improve your search strategy .

It is important to remember that there is nothing wrong, per se, if your search query generates irrelevant hits. This is quite normal when performing a structured literature search. What's important is that your search strategy is able to retrieve the type of articles that you are looking for, and that you are not overwhelmed by the total number of hits (given the time frame of your thesis project).

We recommend that you use the advanced search form when improving your search strategy. By using the advanced search form, you will for example be able to specify which search fields your search terms must be present in.

Narrowing your search 

Broaden your search, how to specify the field you would like to search in pubmed, step 5: select and review articles.

• After you have completed your search, you will need to go through all your search hits and select which articles to include in your review.
• When selecting articles, read through the titles and abstracts of each article to decide its relevancy .
• Check the quality of each study that you include in your review.
• When checking the quality of articles, it is common to use critical appraisal worksheets or checklists .

Selecting articles

When you have completed and feel satisfied with your search, it is time to go through all the search hits and select which studies to include in your review. All relevant studies, that is, those studies that correspond to your research question and your previously set inclusion criteria, should be included. You decide on the relevancy of a study primarily by reading through the title and abstract. If you feel unsure, go through the whole article. You can describe your selection process with the help of a flow chart, such as the frequently used PRISMA flow diagram .

One of the challenges of systematic literature searches is that the search strategy should be exhaustive, but at the same time the number of search hits also needs to be kept within reasonable boundaries. A search query needs to be broad enough to retrieve all relevant studies, but on the other hand, this also means that a large portion of the search results will be irrelevant. Hence, even though your search strategy may have generated hundreds of hits, it is fine to only include ten to twenty articles in your review in the end.

Saving articles

If you create a personal account in a database it will be easier for you to save any references that you may find there. Another way of saving and organising article references is to use reference management software. There are several different reference management software, for example Endnote Online and Zotero.

Read more about reference management and see software guides.

Reviewing articles

When you have made your selection, you should critically examine the quality of all articles included in your review. The assessment is typically performed with the help of a critical appraisal guide or checklist. The purpose is to assess the reliability of the study results and whether there are any methodological flaws that may have impacted the results. Qualitative research articles are often reviewed with a focus on authenticity, credibility, and validity.

There are many different critical appraisal worksheets and checklists. Some examples are the SBU checklists for assessing the quality of randomized studies, observational studies, and qualitative research. In the course book How to do a systematic review in nursing there is a review guide that can be used for assessing different kinds of studies (both qualitative and quantitative); the original source is Caldwell, Henshaw & Taylor, 2011 .

Review worksheets and checklists contain criteria and questions that may help you identify flaws, errors, or bias. Sometimes different aspects of the study are scored separately. Later, all scores make up a final score that indicates whether the study is of high, medium, or low quality.

Many programmes and courses provide instructions on which checklists to use when reviewing articles, so check your course guidelines.

Step 6: Report your search strategy

• Describe your search strategy in a manner that makes it possible for your readers to replicate the search and get the same results.
• The search strategy is often presented in the form of a table .
• Look at the search history to see what words and limits that you have used when searching a database.

An important aspect of doing a structured literature review is transparency. It has to be easy for your readers to follow what you did when you searched and selected the articles that you have included in your review. In the method section of your literature review you should describe how you searched different databases. This is also where you describe any manual searches that you did. Search strategies are commonly reported in the form of tables. Present one table for each database.

You can examine your search terms and any limits you have applied when searching a database by visiting its search history.

Read more about how to report your search strategy and view examples.

Checklist for search strategies

Here is a checklist to help you review your own or someone else's search strategy.

If you would like us to get back to you, please submit your contact information in the form below along with your feeback.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Database (Oxford)
• v.2020; 2020

Structured reviews for data and knowledge-driven research

Núria queralt-rosinach.

1 Department of Integrative Structural and Computational Biology, Scripps Research, 10550 N Torrey Pines Rd. La Jolla, CA 92037, USA

Gregory S Stupp

Tong shu li, michael mayers, maureen e hoatlin.

2 Department of Biochemistry and Molecular Biology, Oregon Health and Science University, 3181 SW Sam Jackson Parkway, Portland, OR 97239, USA

Matthew Might

3 Department of Medicine, Hugh Kaul Precision Medicine Institute, University of Alabama at Birmingham, 510 20th St S, Birmingham, AL 35210, USA

Benjamin M Good

Andrew i su.

Hypothesis generation is a critical step in research and a cornerstone in the rare disease field. Research is most efficient when those hypotheses are based on the entirety of knowledge known to date. Systematic review articles are commonly used in biomedicine to summarize existing knowledge and contextualize experimental data. But the information contained within review articles is typically only expressed as free-text, which is difficult to use computationally. Researchers struggle to navigate, collect and remix prior knowledge as it is scattered in several silos without seamless integration and access. This lack of a structured information framework hinders research by both experimental and computational scientists. To better organize knowledge and data, we built a structured review article that is specifically focused on NGLY1 Deficiency, an ultra-rare genetic disease first reported in 2012. We represented this structured review as a knowledge graph and then stored this knowledge graph in a Neo4j database to simplify dissemination, querying and visualization of the network. Relative to free-text, this structured review better promotes the principles of findability, accessibility, interoperability and reusability (FAIR). In collaboration with domain experts in NGLY1 Deficiency, we demonstrate how this resource can improve the efficiency and comprehensiveness of hypothesis generation. We also developed a read–write interface that allows domain experts to contribute FAIR structured knowledge to this community resource. In contrast to traditional free-text review articles, this structured review exists as a living knowledge graph that is curated by humans and accessible to computational analyses. Finally, we have generalized this workflow into modular and repurposable components that can be applied to other domain areas. This NGLY1 Deficiency-focused network is publicly available at http://ngly1graph.org/ .

Availability and implementation

Database URL: http://ngly1graph.org/ . Network data files are at: https://github.com/SuLab/ngly1-graph and source code at: https://github.com/SuLab/bioknowledge-reviewer .

ude.sppircs@usa

Introduction

Science progresses via an iterative loop between hypothesis generation, experimentation and interpretation. Interpretation and generation of hypothesis relies on putting new data in context with existing relevant knowledge. Researchers typically need to access the relevant knowledge to their research question and hypothesis. One method we have for accessing existing relevant knowledge are review articles, as they summarize the current knowledge of a particular topic. In the context of hypothesis generation, reviews are designed to collect all evidence that answers a specific question. These evidence focuses on information that directly relate to the research question, background knowledge specific to the research question domain such as a disease and experimental data. All these different data and knowledge are synthesized from structured distributed knowledge bases or unstructured scientific papers and experimental datasets. These systematic review articles are common in biomedicine, but the content is typically expressed only as free-text in scholarly papers, which is not easily queryable and computable. Leveraging the knowledge and data contained in a review by researchers is currently not possible in a computer-accessible and automatic way. Consequently, the community does not benefit from the full value of review articles for hypothesis generation.

To harness computationally the content summarized in reviews, researchers typically must spend much of their time and effort searching, pre-processing and integrating biomedical information. The landscape of biomedical information resources is heterogeneous and broad. As a result, the informatics community is very often faced with the challenge of integrating data across many knowledge resources ( Figure 1 ). Most of these knowledge resources organize information for a relatively limited scope of information types, and across a wide range of biological domains. For example, BioGRID focuses on integrating data on physical and genetic interactions ( 1 ), and the Gene Ontology Consortium annotates functions of gene products ( 2 ). In addition, information aggregation platforms like Monarch ( 3 ), the EBI RDF platform ( 4 ) and Open PHACTS ( 5 ) integrate data for several information types into a single data infrastructure such as knowledge graphs. While these meta-databases are valuable, their objective is typically different from the goals of a domain-specific review article, which usually focuses on a richly heterogeneous network of knowledge in a relatively limited scope of biology.

Conceptual overview of structured review articles. This figure represents the distribution of knowledge in databases accessible to the community in terms of domains compiled (X axis) and information structured (Y axis). Gray squares indicate knowledge focus of a database with regards to the domain(s) and information structured.

Knowledge graphs are computer-readable semantic representations of relational information, where concepts are encoded as nodes, and the relationships between those concepts are represented as edges. Knowledge graphs make it easy to integrate information from many sources, to explore heterogeneous information within a single data model and to infer new relationships via efficient queries. Knowledge graphs have been used to organize background knowledge for data interpretation and hypothesis generation in a wide variety of contexts ( 6–9 , 4 , 10–12 ).

This work focuses on the complementary challenge of organizing a diverse and heterogeneous graph of knowledge relevant to a research question in a relatively defined domain area for hypothesis generation, and performing this integration in a way that is amenable to computational analysis. We call this type of effort a ‘structured review article’. We propose structured review articles as knowledge graphs focused on specific domains and research questions. The goal of a structured review article is to organize relevant knowledge to make it interpretable and queryable by humans and computers. A structured review is similar to a systematic review in that it attempts to summarize current knowledge and evidence relevant to a research question, but it is different in that the knowledge is assembled in a modern and computable fashion. The benefit of this transformation is that the knowledge is computationally accessible and efficiently processable. This allows for the application of graph and artificial intelligence algorithms and tools and better promotes the FAIR principles ( 13 , 14 ). As a proof-of-concept, we worked with researchers studying NGLY1 Deficiency (DOID:0060728), an ultra-rare disease first reported in 2012 that affects less than 100 patients worldwide with no treatment ( 15 ). They previously found evidence of a genetic association between NGLY1 and aquaporins ( 16 ) where multiple aquaporins’ transcription was reduced in NGLY1-deficient cells by an unknown mechanism. Here, we explored the use of knowledge graphs as structured review articles to identify plausible regulatory mechanisms to explain these observations.

To demonstrate the value of this concept, we created a structured review article of NGLY1 Deficiency. In contrast to traditional free-text review articles, this structured review exists as a living knowledge graph that is curated by humans and accessible to computational analyses. Construction of this structured review was based on an iterative cycle of defining a research question, ingesting relevant data resources and querying the resulting knowledge graph. Essential to this process was a close collaboration and iterative design with domain experts in NGLY1 Deficiency. To generalize this process to any domain area, we created tools to assist the creation and exploration of focused knowledge graphs. We also created a tool for community curation and contribution. We show how structured reviews are efficient knowledge structures for access and usability for humans and computers that may be particularly useful in rare disease research to springboard hypothesis building, to identify potential collaborations and to suggest potential testable hypotheses for drug discovery or repurposing.

Biomedical data

The criteria to select our resources were (i) containing substantive information to answer the research question, (ii) open and commonly used and (iii) curated and maintained. To construct our NGLY1 Deficiency-focused knowledge graph, we utilized the following structured knowledge and data resources:

• We did biocuration to collect background knowledge specific to NGLY1 Deficiency and NGLY1 and structured it as a semantic network; see more details in the Results section. We used the 20180118 network version ( https://github.com/NuriaQueralt/bioknowledge-reviewer/tree/master/bioknowledge_reviewer/curation/data/v20180118 ).
• We used Wikidata ( http://wikidata.org/ ) to retrieve metadata for our biocurated network such as identifiers (IDs) from different vocabularies or entity cross-references, human readable labels, synonyms and descriptions. Wikidata is a project of the Wikimedia Foundation that enables the collaborative construction of a centralized graph database. Wikidata contains biomedical knowledge populated automatically from trusted authorities such as NCBI’s Entrez Gene, PubChem and the Human Disease Ontology ( 17 ). Using the Wikidata SPARQL API ( http://query.wikidata.org ), we retrieved data from the 201703 version.
• We used the Monarch Initiative platform ( 3 ) to retrieve human and animal model biological data and metadata for node and edge attributes in the structured review. The Monarch Initiative is developing a Knowledge Graph devoted to semantically integrating genomic, phenomic and related information from several species, tracking the evidence of the relationships. This integration is done with a clear emphasis to translate biomedical curated knowledge from animal models to human biology. Using the Monarch Biolink API ( https://api.monarchinitiative.org/api/ ) we retrieved data from the 201 901 version.
• We used the tftargets R package ( https://github.com/slowkow/tftargets ) and the Molecular Signature Database (MSigDb) ( 18 , 19 ) to retrieve human transcription factors (TFs) and their associated target genes data and metadata for node and edge attributes in the structured review. tftargets aggregates experimental and curated gene regulatory information from the TRED ( 20 ), ENCODE ( 21 ), Neph2012 ( 22 ) and TRRUST ( 23 ) databases. We also retrieved regulatory relationships from MSigDB, a collection of annotated gene sets (the C3:TFT sub-collection v6.1) ( 24 ).
• We included one RNA-Seq dataset on a Drosophila model of NGLY1 Deficiency ( 25 ). To create lists of differentially expressed genes, we filtered for absolute fold change >1.5 and false discovery rate (FDR) < 0.05. We represented that Pngl (fly ortholog of NGLY1) is associated with misregulated processes by relating it to each differentially expressed gene using the ‘interacts with’ semantics.
• We used the BioThings MyGene.info API ( http://mygene.info/ ) ( 26 , 27 ) to annotate synonyms, name and description node attributes. We queried mygene.info services on 2019-01.
• We used the Human Metabolome Database (HMDB) ( 28 ) to manually extract facts related to the GlcNAc metabolite. HMDB is a freely available electronic database containing detailed chemical, clinical and biological information about small molecule metabolites found in the human body. We used the version 2017-05.

ID normalization

To normalize entities and relations from different data models, we used a variety of methods. For normalization of curated data to Monarch model we used the Wikidata SPARQL endpoint to retrieve cross-references to map entities. Diseases were linked to MONDO IDs by adding an extra ‘skos:exactMatch’ relationship. We used the MONDO ontology ( http://obofoundry.org/ontology/mondo.html ) to link disease IDs and retrieve node metadata. We used the OWL file version 2018-04-15 ( http://purl.obolibrary.org/obo/mondo/releases/2018-04-15/mondo.owl ). Genes were normalized to HGNC by using BioThings MyGene.info API. We queried mygene.info services on 2019-01. The semantics of all relationships were manually mapped to the ontologies used in Monarch model. Manual inspection of a sample of these mappings confirmed the quality of this process.

The BioKnowledge Reviewer Library

We created a library using the Python 3 programmatic language to reproduce the creation protocol of the structured review in a workflow. Functionality was guided by knowledge and reasoning of NGLY1 researchers. Under the research question, experts formulated a priori the specific questions they wanted to explore in the graph. The library is designed to give flexibility on the construction of the review network by choosing, concatenating and merging topic-specific networks. It allows to build reviews in a modular way by steps in the workflow and by different topics. We adopted a modular approach for its management to facilitate its reusability, update and consistency checking. The architecture of the system has four components ( Figure 2 ). This programming library is versioned in GitHub to enable the community to add new functionality required to review new pieces of knowledge and apply it to answer more diverse types of questions. Availability at

Library architecture. Architecture of the system based on four components. The edges component contains libraries with functions to collect, normalize and format the information and data resources we want to integrate as individual networks. The graph component contains functions to integrate and create the knowledge graph. The Neo4j component contains the module to import the graph into Neo4j. Finally, the hypothesis-generation component contains the modules to query the graph, structure the resulting semantic paths and extract summaries to analyse connections and the evidence.

Library: https://github.com/SuLab/bioknowledge-reviewer

Workflow notebook: https://github.com/SuLab/bioknowledge-reviewer/blob/master/bioknowledge_reviewer/graph_v3.2_v20190616.ipynb

Data storage and mining

We used the Neo4j graph database framework for storage, management and mining of structured data. The graph database approach has been shown to facilitate management and exploration of biomedical knowledge ( 29 ). Neo4j enables users to query the graph using the Cypher query language, either through an API or a graphical user interface. All data were imported into the Neo4j Community Server v3.5.3.

To evaluate the disease-based biocuration, we utilized the Semantic MEDLINE Database (SemMedDB) ( 30 ). SemMedDB is a repository of semantic relations sentence-based extracted from the titles and abstracts of all PubMed citations by a general knowledge-based text mining system called SemRep ( 31 ). We used version semmedVER31_R, which contains information about approximately 94.0 million relations from all of PubMed citations (about 27.9 million citations).

Construction of the knowledge graph

The overarching goal of our study was to introduce structured review articles as a new way of processing scientific knowledge in the context of hypothesis generation. With this aim, we assessed structured reviews to generate mechanistic hypotheses for recent experimental observations. Specifically, we sought to explain the phenotypic effects of aquaporins on cellular phenotypes of NGLY1 Deficiency. Researchers found a transcriptional regulation link between NGLY1, ENGASE and AQP1 ( 16 ). However, the mechanism of this effect on a molecular level was not clear. Therefore, to identify plausible potential mechanisms to explain this observation and others like it, we iteratively constructed a knowledge graph that focused on information relevant to the NGLY1 gene, NGLY1 Deficiency and aquaporins.

Domain expert knowledge

We seeded our graph with nine key concepts as initial nodes—seven genes (NGLY1 human (HGNC:17646), AQP1 human (HGNC:633), AQP1 mouse (MGI:103201), ENGASE human (HGNC:24622), NFE2L1 human (HGNC:7781), AQP3 human (HGNC:636), AQP11 human (HGNC:19940)), one metabolite (GlcNAc (HMDB:HMDB00215)) and the disease NGLY1 Deficiency (OMIM:615273). NFE2L1 is a TF recently discovered to be dependent on NGLY1 ( 32–34 ). Using these nodes as seeds, we then retrieved as much biomedical information as possible on these entities. We first consulted the knowledge base created by the Monarch Initiative ( 3 ). We expanded our network to include all concepts with an explicit relationship to one of our nine seed nodes. This expanded network included 713 nodes, including 234 genes, 80 diseases/phenotypes, 174 pathways, 111 tissues, 49 gene variants and 65 genotypes (genetic backgrounds associated with individuals, cell lines or mouse models) that were connected via 6756 edges. Although the Monarch-derived network provided an important foundation for our work, we found that Monarch expansion alone did not represent several key facets that were important to our query, including edges related to the GlcNAc metabolite, protein function and domain information and specific transcriptional regulatory relationships. Therefore, we then performed a targeted expansion of our network via three strategies.

Biocuration

To include the most recent findings described in the literature around the molecular basis and clinical description of the disease, we curated two scientific papers that together compiled biomedical knowledge relevant to NGLY1 Deficiency. Together, the papers by Enns et al . in 2014 ( 35 ) and Lam et al . in 2017 ( 36 ) captured the known molecular biology involved in the disorder and the most recent and complete characteristics of the clinical phenotypic spectrum. Based on these two papers, we added to our knowledge network 101 phenotypes linked to NGLY1 Deficiency and an additional 142 biological relationships joining genes, variants and functions. This work also directly led to the creation of a term for NGLY1 Deficiency in the Human Disease Ontology ( 37 ), and the creation of 45 new phenotype terms in the Human Phenotype Ontology ( 38 ).

To evaluate the relevance of this curation effort, we compared our results with text mined statements in SemMedDB ( 30 ). Our manual curation effort resulted in more relationships (243 curated versus 11 text mined statements) and with more precise expressivity than via text mining. For instance, SemMedDB identified only four statements related to NGLY1 Deficiency phenotypes, and the information content was much less precise than our manual biocuration.

In addition, Owings et al. showed that GlcNAc is a metabolite with a potential key role in the molecular basis of NGLY1 Deficiency ( 25 ). As Monarch does not include metabolites in its knowledge base, we extracted edges related to GlcNAc from the Human Metabolome Database ( 28 ), KEGG ( 39 ), ChEBI ( 40 ) and ChEMBL ( 41 ). This work resulted in the addition of 362 edges and 302 nodes to our knowledge graph.

Ortholog phenotypes

To increase the connectivity around NGLY1 and aquaporins, we included animal model information since the conserved biology could help explain the pathology in humans. We first added to our graph the orthologs for the seven seed genes, as well as the orthologs for all genes connected to any of the nine seed nodes. From these gene–ortholog edges, we then added all ortholog–phenotype relationships from Monarch. As a result, we expanded our network with 246 new ortholog nodes, 570 new phenotype nodes and 4930 new edges.

TF regulation

To test the hypothesis that NGLY1 and aquaporins are mechanistically related through altered transcriptional regulation, we looked for data sources of known experimentally determined TF–gene relations. We also integrated a recently published NGLY1 Deficiency fly model transcriptomic profile data set (25) into the review and merged the collected TF–gene data. As a result, we expanded our network with 9723 TF–gene edges and 386 expression edges (including 4226 new genes from which 640 are known TFs).

Finally, we again used the Monarch database to retrieve pairwise relationships between network nodes. The final knowledge graph contained 9361 nodes, including 6152 genes or proteins, 2486 diseases or phenotypes, 355 pathways, 193 genotypes, 117 tissues, 50 variants, 7 chemical compounds and 1 organism and contained 234 717 edges across 29 relationship types. Our final NGLY1-focused knowledge graph integrated data and knowledge derived from scientific literature, domain experts, databases and gene expression data.

The amount of effort required to develop the structured review depended on the structure of the resource added, being the curation process of the scientific literature the most demanding. Conceptually, the creation of structured review articles is as subjective as a regular review. The objective part is that researchers should focus on collecting relevant information, background knowledge and experimental data that are key to the hypothesis and research question. Hence, the construction workflow of a review article is generalizable to (i) collect, (ii) integrate as a knowledge graph, (iii) import into a graph database and (iv) query the review for hypothesis generation.

The data model represents heterogeneous biomedical knowledge using common controlled vocabularies and ontologies in the Life Sciences to identify nodes and edges. Regarding node and edge human readable descriptions, we put special emphasis on being sufficient and efficient, i.e. the minimum but useful amount of information for our biologist users to understand the relational information, the entities involved and the supporting evidence backing each edge when accessible. For nodes, we included an ID, a label, a name and a description. For edges, we included a property ID, a label, a property Uniform Resource Identifier (URI) to link to a more detailed description of the semantics of the relation, a sentence supporting the relation extracted by biocuration from the original scientific publication or stating the source of provenance, and a reference URI to trace back the description of the relation in the original research study. A detailed description of the data model is available at ngly1graph.org . This data model allows users to query the graph by any of these node and edge attributes.

Implementation and availability

Review creation by bioinformatic workflows.

To create and analyse the structured review, we created a Python library and a jupyter notebook based on the knowledge graph construction protocol described above. The library allows any researcher to automate and reproduce the ingestion, integration of all data sources and the creation of the knowledge graph in bioinformatic workflows. The workflow design meets the organic and iterative process performed with users. Every review can be created by mixing and remixing evolving knowledge on different topics, and it is versioned for use and reuse. Using the library, we can also derive structured hypotheses alongside each structured review, in the same workflow. This enables researchers: (i) to update the review iteratively avoiding redundant effort, and with own private datasets, (ii) to reuse reviews to revisit the hypothesis with new findings and (iii) to share with other researchers for reproduction, discussion or curation or even as a citation. In addition to automating the exact work described in this manuscript, these programmatic tools also allow this general strategy to be generalized to other disease areas and enable other developers to extend our work to new knowledge resources. To see a figure of a canonical workflow, please refer to the ‘Library architecture’ section in the README in GitHub ( https://github.com/NuriaQueralt/bioknowledge-reviewer ). The README provides documentation and a jupyter notebook with the creation workflow used to build the review.

Data access

We stored the resulting graph in a Web-based application for dissemination, querying, visualization and curation. This application is a hybrid between two technologies—Neo4j and Wikibase—with complementary strengths: knowledge navigation and knowledge contribution. Neo4j is a graph database with several useful features. First, Neo4j offers a powerful graph query language (‘Cypher’) that enables any researcher to mine the database on our central server without having to set up any computer hardware or software of their own. Second, Neo4j also provides a Web-based graphical user interface for interactive database access and exploration. To provide simple starting points, we created Neo4j ‘guides’ as basic tutorials, which contain templates for representative queries that can be extended and customized. Third, Neo4j utilizes a simple structured data format for both data import and export. This data structure facilitates the downstream reuse of the network using external tools and custom analyses.

The second component of our application is based on Wikibase, a system that enables living community curation and contribution of structured data. One of our key motivating design features was the ability for the NGLY1 community to contribute to a centralized community resource, but Neo4j does not have an easy mechanism for this purpose. Wikibase is the open source software that powers Wikidata, a crowdsourced effort to curate, manage and share structured data ( 42 ). The Wikibase software offers manual editing via a Web-based interface, as well as automated editing via an API. Wikibase also includes detailed change tracking, and a SPARQL endpoint as the RDF query service a key technology for connecting with Linked Data and the Semantic Web.

Joining the Neo4j and Wikibase components in a single hybrid system is a continuously running, real-time synchronization engine. This synchronization allows to perform updates of underlying databases through the library and keep track of these changes in the state of the review in the history of Wikibase. This hybrid system combines the complementary strengths of each component—community contributions via Web-based and programmatic interfaces, a Web-based interface for graph visualization and powerful query capabilities for discovery through two widely used query languages. Finally, the structured review can be accessed via download of code and data, data as CSV or the interoperable RDF format.

Applications

Community curation.

The NGLY1 community (and other rare disease communities in general) do not have the resources to sustain focused biocuration efforts. Therefore, we turned to community curation as a mechanism to continue the maintenance and expansion of the NGLY1 knowledge graph to keep it up-to-date. Our interface for community curation is based on Wikibase, the software underlying Wikidata. This Wikibase extension allows anyone in the community to directly add and edit information to the NGLY1 knowledge graph through an online, graphical interface. Importantly, community curated data is fully structured at time of submission, making it fully integrated into the knowledge network. Every contribution made in the graph is tracked (time, curator and change) in the history link of each entity page, either node or edge, and the structured review can be dumped and versioned from either Wikibase or Neo4j implementations.

This centralized, community-maintained resource will facilitate the exchange of knowledge across the entire NGLY1 community. Wikibase also makes these data interoperable with the broader ecosystem of Linked Data resources, which in turn will facilitate further reuse and additional data integration.

Hypothesis generation and exploration

A structured review brings together different topics that open the possibility to unveil hypotheses by mining existing knowledge, as well as by offering additional ways to use existing knowledge to contextualize newly generated data. Now the user can easily interrogate multi-dimensional and multi-domain knowledge and data in the structured review from a single endpoint. The user can traverse the NGLY1 Deficiency gene expression data in Drosophila, human TF–gene relationships and the heterogeneous content on genes, pathways, phenotypes or gene–gene interactions and their crosstalk between different species to mine the gap of knowledge to generate hypotheses around the research question. Some example queries that explore potential mechanistic links between NGLY1 and AQP1 include the following:

• (i) Do NGLY1 and AQP1 (or related genes) share any phenotypes in NGLY1 and AQP1 knockout animals?
• (ii) Do NGLY1 interacts with TFs related to AQP1 transcriptional regulation?
• (iii) Do AQP1 and phenotypes of NGLY1 Deficiency share associated genes?

To illustrate a possible use of the NGLY1 Deficiency knowledge graph, we developed the research question of whether the AQP1 gene may be important in the NGLY1 Deficiency phenotypes. To answer this question, biologists can interrogate the review as a graph with only two simple queries: a first query to find mechanistic links between NGLY1 and AQP1 gene expression and a second query to find links between these mechanisms and the disease phenotypes. These queries are detailed in a Neo4j guide available at ngly1graph.org .

First, to identify mechanistic links between these two genes, we formulated a query template based on the hypothesis that they are related via transcriptional regulation looking for an NGLY1 dependent TF which regulates AQP1 or what we called herein the regulatory hypothesis . We found no direct regulatory links between the human NGLY1 gene and AQP1 gene. Therefore, we expanded our query to include regulatory links that were described in Drosophila orthologs (25) ( Figure 3A ).

Exploration of mechanistic paths between NGLY1 and AQP1 based on the regulatory hypothesis. ( A ) First query topology for the regulatory hypothesis. We defined a path topology based on gene pathways of length four linking the NGLY1 ortholog in Drosophila (Pngl) with the human AQP1 gene. The bridging nodes and edges were based on transcriptional regulatory relationships in both Drosophila and human, plus orthology relationships between human and fly genes. ( B ) Mechanistic hypotheses resulted from the first query.

This query returned 19 paths, each of which represents a potential mechanistic hypothesis of how NGLY1 and AQP1 are related ( Figure 3B ). The genes TP53 (HGNC:11998), TBP (HGNC:11588) and LEF1 (HGNC:6551) are candidate regulators identified by the query. TP53 interacts via the proteasome complex (PSMA1 (HGNC:9530), PSMA4, PSMC3 (HGNC:9549)), SQSTM1 (HGNC:11280), a multifunctional protein that binds ubiquitin, and MYC (HGNC:7553), which is a phosphoprotein also related to TBP and LEF1. Further network exploration unveiled that TP53 interacts with TBP ( 43–47 ) and, interestingly, that TBP is an interactor of MEF2A (HGNC:6993) ( 31 ), a known transcriptional regulator of AQP1 ( 24 ) and a member of a TF protein family recently associated with AQP1 transcription regulation ( 48 ).

To further explore this regulatory hypothesis, we queried the knowledge graph for relationships between phenotypes associated with NGLY1 Deficiency and the candidate genes identified in the previous query ( Figure 4A ). This query was designed to prioritize those candidate genes based on available prior evidence linking them to NGLY1 Deficiency phenotypes.

Exploration of the evidence relating candidate regulators of AQP1 to NGLY1 Deficiency phenotypes. ( A ) Second query topology for the AQP1 regulation-disease phenotypes shared genetic basis hypothesis. ( B ) Hypotheses resulted from the second query. All edges are of type ‘has phenotype’.

This query returned 30 paths that link AQP1 to 14 NGLY1 Deficiency phenotypes through genes involved in the regulatory hypothesis ( Figure 4B ). These paths highlighted a potential mechanistic role for four candidate genes: SQSTM1 (HGNC:11280), MYC (HGNC:7553), TP53 (HGNC:11998), TBP (HGNC:11588). For instance, SQSTM1 encodes a protein with regulatory activity on the inflammatory/immune responses related to the nuclear factor kappa-B signalling pathway, which is linked to cancer and nervous system processes such as synaptic plasticity and learning ( 49–52 ). The query results show that SQSTM1 has previously been linked to several phenotypes associated with NGLY1 Deficiency, including ‘Cerebellum atrophy’, ‘peripheral nervous system disease’ or interestingly ‘Dysmetria’, a phenotype caused by lesions in the cerebellum or proprioceptive nerves that lead to the cerebellum that coordinate visual, spatial and other sensory information with motor control ( 53 ). From the first query, we can see that SQSTM1 interacts with TP53, which in turn is also linked to ‘Dysmetria’.

In the Neo4j interface, users can interactively visualize any of the paths identified in these queries, check entity attributes such as a human readable description and explore the evidence and context of each statement accessing the supporting reference through its URI. Users can also extract a table summarizing interesting path features like the most common genes from the transcriptome or the most common TFs (see an example in the Neo4j guide).

The examples described here are just two representative queries that demonstrate the power of mining a semantically precise knowledge graph. The Cypher query language offers powerful capabilities to harness the knowledge in our structured review. Users can formulate Cypher query templates corresponding to their biological question. Cypher templates can handle a broad spectrum of queries, from very precise queries that correspond to specific mechanistic hypotheses, to open-ended queries that flexibly retrieve paths that incorporate arbitrary types of connectivity. More example queries using the Cypher query language are provided in the advanced Neo4j guide available from ngly1graph.org .

To interrogate the review, researchers with no previous experience in interacting with network tools started with the template queries in the guide. Templates walk beginning users through the process of creating a simple case. The definition in collaboration with them of the data model eased both interrogation and understanding of results. To explore this research question without the NGLY1 graph, researchers would have to explore several databases and make time-consuming pre-processing steps and analysis on the results of these searches. Instead, the knowledge graph we have created enables researchers to query the graph with complex questions, to traverse different domains and databases in one single query and to explore disease-specific hypotheses and evidence. These extracted NGLY1-AQP1 regulatory mechanistic links provide a knowledge foundation for bench researchers to create an informed regulatory hypothesis to evaluate in the laboratory. The knowledge sources and versions used to generate the hypothesis are available and described online in the ‘Data’ section of the Neo4j guide, and they were selected jointly with our collaborators. This is important since we assessed that additions of new sources can affect the explanations obtained if they are relevant to answer the question, such as the new deep phenotyping dataset added.

To enable knowledge exploration and exploitation for researchers working on a specific question, we explored the use of knowledge graphs as structured review articles. With our approach, we built the first review article for NGLY1 Deficiency rare disease-specific topic, and we demonstrated that it is now an actionable knowledge resource for the whole community. This paper describes how the resource supports knowledge discovery and dissemination, and it supports and facilitates collaboration between experimental researchers and bioinformaticians.

This work was motivated by the goal of identifying mechanistic hypotheses for an experimental observation. While this general procedure is not unique to our effort, we have incorporated two features that we believe make this work a notable contribution in this area. First, we have focused on creating a structured review article, which is distinct from other review articles in that the output is computationally accessible, and also distinct from other structured data integration efforts because of its relatively narrow and deep focus on a particular domain area. Second, we have published this structured review article to the community in the form of a centralized resource (accessible at http://ngly1graph.org/contribute/ ), a living knowledge graph that can be continually refined by community curation.

To put in context our new way of processing scientific knowledge for comparison, typically structured knowledge resources are created by curation efforts. On the one hand, there are homogeneous edge-specific resources such as GO, BioGRID, STRING, HPRD, UniProt, Reactome, KEGG or MP ( 2 , 1 , 54–57 , 39 , 58 ). On the other hand, there are heterogeneous data integration knowledge bases such as the Monarch knowledge graph ( 3 ). Curated resources are vast but incomplete because the majority of the wealth of knowledge is unstructured since expert curation cannot keep the pace of scientific production.

To organize this knowledge, several text mined heterogeneous knowledge bases have been developed ( 30 , 59 ). Text mined resources are comprehensive but without the semantic specificity required to be useful for the rare disease field. Also, heterogeneous knowledge bases typically have a broad scope to get a systems level understanding, but lack a focus on domain-specific knowledge to address a specific question.

A structured review article helps to mine the gap of knowledge where other resources are incomplete or not expressive enough for the domain or question to solve. Here, we created a Neo4j-Wikibase framework and a Python library to aid in the construction of a new knowledge resource to synthesize information focused on a specific research question. These tools facilitate the integration of data from diverse heterogeneous resources: from manual curation to biomedical databases, to experimental data, to expert knowledge. This knowledge integration library produces complete research objects, i.e. a workflow with the data, the code, the graph and structured hypotheses. These research objects promote more efficient research and reproducible science. Also, relative to traditional review articles, structured review articles are more Findable, Accessible, Interoperable and Reusable ( 60 , 61 ).

This work focused on NGLY1 Deficiency, an ultra-rare disease that has been diagnosed in fewer than 100 children worldwide. However, the principles and tools developed herein are generalizable to other domain areas, and we believe they will be of particular interest and utility to the rare disease community. The community curation application enables living structured review and promotes the creation of FAIR content from the time of its inception. In areas where investments in data infrastructure are modest, these tools will facilitate synergy between experimental and computational biologists and between data curators and data miners.

Acknowledgements

The authors thank Dr Mitali Tambe and Prof. Hudson H. Freeze from the Freeze Lab in the Human Genetics Program at Sanford-Burnham-Prebys Medical Discovery Institute for their input, collaboration and help. We acknowledge the NCATS Biomedical Data Translator Hackathon 2018 in La Jolla for the inspiration and input. The FAIRness of a prototype of the library was evaluated at the NBDC/DBCLS BioHackathon 2018 in Matsue. We thank Jiwen Xin, Sébastien Lelong and Chunlei Wu from the Wu Lab at Scripps Research for their help.

The National Center for Advancing Translational Sciences Biomedical Data Translator program (OT3TR002019); National Institute of General Medical Sciences (R01GM089820).

Conflict of interest

The authors declare they have no conflict of interest.

How To Structure Your Literature Review

3 options to help structure your chapter.

By: Amy Rommelspacher (PhD) | Reviewer: Dr Eunice Rautenbach | November 2020 (Updated May 2023)

Writing the literature review chapter can seem pretty daunting when you’re piecing together your dissertation or thesis. As  we’ve discussed before , a good literature review needs to achieve a few very important objectives – it should:

• Demonstrate your knowledge of the research topic
• Identify the gaps in the literature and show how your research links to these
• Provide the foundation for your conceptual framework (if you have one)
• Inform your own  methodology and research design

To achieve this, your literature review needs a well-thought-out structure . Get the structure of your literature review chapter wrong and you’ll struggle to achieve these objectives. Don’t worry though – in this post, we’ll look at how to structure your literature review for maximum impact (and marks!).

But wait – is this the right time?

Deciding on the structure of your literature review should come towards the end of the literature review process – after you have collected and digested the literature, but before you start writing the chapter. 

In other words, you need to first develop a rich understanding of the literature before you even attempt to map out a structure. There’s no use trying to develop a structure before you’ve fully wrapped your head around the existing research.

Equally importantly, you need to have a structure in place before you start writing , or your literature review will most likely end up a rambling, disjointed mess. 

Importantly, don’t feel that once you’ve defined a structure you can’t iterate on it. It’s perfectly natural to adjust as you engage in the writing process. As we’ve discussed before , writing is a way of developing your thinking, so it’s quite common for your thinking to change – and therefore, for your chapter structure to change – as you write. 

Need a helping hand?

Like any other chapter in your thesis or dissertation, your literature review needs to have a clear, logical structure. At a minimum, it should have three essential components – an  introduction , a  body   and a  conclusion . 

Let’s take a closer look at each of these.

1: The Introduction Section

Just like any good introduction, the introduction section of your literature review should introduce the purpose and layout (organisation) of the chapter. In other words, your introduction needs to give the reader a taste of what’s to come, and how you’re going to lay that out. Essentially, you should provide the reader with a high-level roadmap of your chapter to give them a taste of the journey that lies ahead.

Here’s an example of the layout visualised in a literature review introduction:

Your introduction should also outline your topic (including any tricky terminology or jargon) and provide an explanation of the scope of your literature review – in other words, what you  will   and  won’t   be covering (the delimitations ). This helps ringfence your review and achieve a clear focus . The clearer and narrower your focus, the deeper you can dive into the topic (which is typically where the magic lies). 

Depending on the nature of your project, you could also present your stance or point of view at this stage. In other words, after grappling with the literature you’ll have an opinion about what the trends and concerns are in the field as well as what’s lacking. The introduction section can then present these ideas so that it is clear to examiners that you’re aware of how your research connects with existing knowledge .

2: The Body Section

The body of your literature review is the centre of your work. This is where you’ll present, analyse, evaluate and synthesise the existing research. In other words, this is where you’re going to earn (or lose) the most marks. Therefore, it’s important to carefully think about how you will organise your discussion to present it in a clear way. 

The body of your literature review should do just as the description of this chapter suggests. It should “review” the literature – in other words, identify, analyse, and synthesise it. So, when thinking about structuring your literature review, you need to think about which structural approach will provide the best “review” for your specific type of research and objectives (we’ll get to this shortly).

There are (broadly speaking)  three options  for organising your literature review.

Option 1: Chronological (according to date)

Organising the literature chronologically is one of the simplest ways to structure your literature review. You start with what was published first and work your way through the literature until you reach the work published most recently. Pretty straightforward.

The benefit of this option is that it makes it easy to discuss the developments and debates in the field as they emerged over time. Organising your literature chronologically also allows you to highlight how specific articles or pieces of work might have changed the course of the field – in other words, which research has had the most impact . Therefore, this approach is very useful when your research is aimed at understanding how the topic has unfolded over time and is often used by scholars in the field of history. That said, this approach can be utilised by anyone that wants to explore change over time .

For example , if a student of politics is investigating how the understanding of democracy has evolved over time, they could use the chronological approach to provide a narrative that demonstrates how this understanding has changed through the ages.

Here are some questions you can ask yourself to help you structure your literature review chronologically.

• What is the earliest literature published relating to this topic?
• How has the field changed over time? Why?
• What are the most recent discoveries/theories?

In some ways, chronology plays a part whichever way you decide to structure your literature review, because you will always, to a certain extent, be analysing how the literature has developed. However, with the chronological approach, the emphasis is very firmly on how the discussion has evolved over time , as opposed to how all the literature links together (which we’ll discuss next ).

Option 2: Thematic (grouped by theme)

The thematic approach to structuring a literature review means organising your literature by theme or category – for example, by independent variables (i.e. factors that have an impact on a specific outcome).

As you’ve been collecting and synthesising literature , you’ll likely have started seeing some themes or patterns emerging. You can then use these themes or patterns as a structure for your body discussion. The thematic approach is the most common approach and is useful for structuring literature reviews in most fields.

For example, if you were researching which factors contributed towards people trusting an organisation, you might find themes such as consumers’ perceptions of an organisation’s competence, benevolence and integrity. Structuring your literature review thematically would mean structuring your literature review’s body section to discuss each of these themes, one section at a time.

Here are some questions to ask yourself when structuring your literature review by themes:

• Are there any patterns that have come to light in the literature?
• What are the central themes and categories used by the researchers?
• Do I have enough evidence of these themes?

PS – you can see an example of a thematically structured literature review in our literature review sample walkthrough video here.

Option 3: Methodological

The methodological option is a way of structuring your literature review by the research methodologies used . In other words, organising your discussion based on the angle from which each piece of research was approached – for example, qualitative , quantitative or mixed  methodologies.

Structuring your literature review by methodology can be useful if you are drawing research from a variety of disciplines and are critiquing different methodologies. The point of this approach is to question  how  existing research has been conducted, as opposed to  what  the conclusions and/or findings the research were.

For example, a sociologist might centre their research around critiquing specific fieldwork practices. Their literature review will then be a summary of the fieldwork methodologies used by different studies.

Here are some questions you can ask yourself when structuring your literature review according to methodology:

• Which methodologies have been utilised in this field?
• Which methodology is the most popular (and why)?
• What are the strengths and weaknesses of the various methodologies?
• How can the existing methodologies inform my own methodology?

3: The Conclusion Section

Once you’ve completed the body section of your literature review using one of the structural approaches we discussed above, you’ll need to “wrap up” your literature review and pull all the pieces together to set the direction for the rest of your dissertation or thesis.

The conclusion is where you’ll present the key findings of your literature review. In this section, you should emphasise the research that is especially important to your research questions and highlight the gaps that exist in the literature. Based on this, you need to make it clear what you will add to the literature – in other words, justify your own research by showing how it will help fill one or more of the gaps you just identified.

Last but not least, if it’s your intention to develop a conceptual framework for your dissertation or thesis, the conclusion section is a good place to present this.

Example: Thematically Structured Review

In the video below, we unpack a literature review chapter so that you can see an example of a thematically structure review in practice.

Let’s Recap

In this article, we’ve  discussed how to structure your literature review for maximum impact. Here’s a quick recap of what  you need to keep in mind when deciding on your literature review structure:

• Just like other chapters, your literature review needs a clear introduction , body and conclusion .
• The introduction section should provide an overview of what you will discuss in your literature review.
• The body section of your literature review can be organised by chronology , theme or methodology . The right structural approach depends on what you’re trying to achieve with your research.
• The conclusion section should draw together the key findings of your literature review and link them to your research questions.

If you’re ready to get started, be sure to download our free literature review template to fast-track your chapter outline.

Psst… there’s more!

This post is an extract from our bestselling short course, Literature Review Bootcamp . If you want to work smart, you don't want to miss this .

You Might Also Like:

27 Comments

Great work. This is exactly what I was looking for and helps a lot together with your previous post on literature review. One last thing is missing: a link to a great literature chapter of an journal article (maybe with comments of the different sections in this review chapter). Do you know any great literature review chapters?

I agree with you Marin… A great piece

I agree with Marin. This would be quite helpful if you annotate a nicely structured literature from previously published research articles.

Awesome article for my research.

I thank you immensely for this wonderful guide

It is indeed thought and supportive work for the futurist researcher and students

Very educative and good time to get guide. Thank you

Great work, very insightful. Thank you.

Thanks for this wonderful presentation. My question is that do I put all the variables into a single conceptual framework or each hypothesis will have it own conceptual framework?

Thank you very much, very helpful

This is very educative and precise . Thank you very much for dropping this kind of write up .

Pheeww, so damn helpful, thank you for this informative piece.

I’m doing a research project topic ; stool analysis for parasitic worm (enteric) worm, how do I structure it, thanks.

comprehensive explanation. Help us by pasting the URL of some good “literature review” for better understanding.

great piece. thanks for the awesome explanation. it is really worth sharing. I have a little question, if anyone can help me out, which of the options in the body of literature can be best fit if you are writing an architectural thesis that deals with design?

I am doing a research on nanofluids how can l structure it?

Beautifully clear.nThank you!

Lucid! Thankyou!

Brilliant work, well understood, many thanks

I like how this was so clear with simple language 😊😊 thank you so much 😊 for these information 😊

Insightful. I was struggling to come up with a sensible literature review but this has been really helpful. Thank you!

You have given thought-provoking information about the review of the literature.

Thank you. It has made my own research better and to impart your work to students I teach

I learnt a lot from this teaching. It’s a great piece.

I am doing research on EFL teacher motivation for his/her job. How Can I structure it? Is there any detailed template, additional to this?

You are so cool! I do not think I’ve read through something like this before. So nice to find somebody with some genuine thoughts on this issue. Seriously.. thank you for starting this up. This site is one thing that is required on the internet, someone with a little originality!

I’m asked to do conceptual, theoretical and empirical literature, and i just don’t know how to structure it

Submit a Comment Cancel reply

Your email address will not be published. Required fields are marked *

Save my name, email, and website in this browser for the next time I comment.

• Print Friendly

To read this content please select one of the options below:

Please note you do not have access to teaching notes, on the shoulders of giants: undertaking a structured literature review in accounting.

Accounting, Auditing & Accountability Journal

ISSN : 0951-3574

Article publication date: 20 June 2016

The purpose of this paper is to present a method for a structured literature review (SLR). An SLR is a method for examining a corpus of scholarly literature, to develop insights, critical reflections, future research paths and research questions. SLRs are common in scientific disciplines dominated by quantitative approaches, but they can be adapted in accounting studies since quantitative and qualitative approaches are commonly accepted.

Design/methodology/approach

A literature review, as a piece of academic writing, must have a logical, planned structure. The authors also argue it requires tests based on qualitative and quantitative methods. Therefore, the authors describe ten steps for developing an SLR.

The SLR method is a way that scholars can stand “on the shoulders of giants” and provide insightful and impactful research that is different to the traditional authorship approaches to literature reviews.

Research limitations/implications

Traditional literature reviews can have varied results because of a lack of rigour. SLRs use a process that, through a set of rules, potentially offers less bias and more transparency of the execution and measures and techniques of validation and reliability.

Practical implications

SLRs provide an approach that can help academics to discover under-investigated topics and methods, nurturing, therefore, the development of new knowledge areas and research approaches.

Originality/value

The paper presents accounting researchers with an opportunity to develop insightful and publishable studies, and also serves as a basis for developing future research agendas in the accounting field. The authors advocate the SLR method especially to higher degree research students and emerging scholars as a way of potentially developing robust and defensible research agendas and questions.

• Citation analysis
• Reliability
• Structured literature review
• Traditional authorship literature review

Acknowledgements

The authors are grateful to participants of PhD and research workshops at Macquarie University, Sydney Australia, the University of Cassino and Southern Lazio, Italy, the University of Udine, Italy, National Institute of Education, Singapore, and University of Waikato, New Zealand and this paper ' s reviewers for their most valuable input into refining the SLR method. Additionally, thanks to our PhD students for working with the authors and giving feedback on the process. Thanks also to Fiona Crawford from the Editorial Collective for her sterling editorial advice.

Massaro, M. , Dumay, J. and Guthrie, J. (2016), "On the shoulders of giants: undertaking a structured literature review in accounting", Accounting, Auditing & Accountability Journal , Vol. 29 No. 5, pp. 767-801. https://doi.org/10.1108/AAAJ-01-2015-1939

Emerald Group Publishing Limited

Copyright © 2016, Emerald Group Publishing Limited

Related articles

We’re listening — tell us what you think, something didn’t work….

Report bugs here

All feedback is valuable

Please share your general feedback

Join us on our journey

Platform update page.

Visit emeraldpublishing.com/platformupdate to discover the latest news and updates

Questions & More Information

Answers to the most commonly asked questions here

• Franklin University |
• Help & Support |
• Locations & Maps |

• | Research Guides

To access Safari eBooks,

• Select not listed in the Select Your Institution drop down menu.
• Enter your Franklin email address and click Go
• click "Already a user? Click here" link
• Enter your Franklin email and the password you used to create your Safari account.

Continue Close

Literature Review

• Getting Started
• Framing the Literature Review

Literature Review Process

• Mistakes to Avoid & Additional Help

The structure of a literature review should include the following :

• An overview of the subject, issue or theory under consideration, along with the objectives of the literature review,
• Division of works under review into themes or categories (e.g. works that support of a particular position, those against, and those offering alternative approaches entirely),
• An explanation of how each work is similar to and how it varies from the others,
• Conclusions as to which pieces are best considered in their argument, are most convincing of their opinions, and make the greatest contribution to the understanding and development of their area of research

The critical evaluation of each work should consider :

• Provenance  -- what are the author's credentials? Are the author's arguments supported by evidence (e.g. primary historical material, case studies, narratives, statistics, recent scientific findings)?
• Objectivity  -- is the author's perspective even-handed or prejudicial? Is contrary data considered or is certain pertinent information ignored to prove the author's point?
• Persuasiveness  -- which of the author's theses are most/least convincing?
• Value  -- are the author's arguments and conclusions convincing? Does the work ultimately contribute in any significant way to an understanding of the subject?

Development of the Literature Review

Four stages:.

• Introduce the reader to the importance of the topic being studied . The reader is oriented to the significance of the study and the research questions or hypotheses to follow.
• Places the problem into a particular context  that defines the parameters of what is to be investigated.
• Provides the framework for reporting the results  and indicates what is probably necessary to conduct the study and explain how the findings will present this information.
• Literature search -- finding materials relevant to the subject being explored.
• Evaluation of resources  -- determining which literature makes a significant contribution to the understanding of the topic.
• Analysis and interpretation -- discussing the findings and conclusions of pertinent literature.

Consider the following issues before writing the literature review:

Sources and expectations.  if your assignment is not very specific about what form your literature review should take, seek clarification from your professor by asking these questions:.

• Roughly how many sources should I include?
• What types of sources should I review (books, journal articles, websites)?
• Should I summarize, synthesize, or critique your sources by discussing a common theme or issue?
• Should I evaluate the sources?
• Should I provide subheadings and other background information, such as definitions and/or a history?

Find Models.   When reviewing the current literature, examine how authors in your discipline or area of interest have organized their literature reviews. Read not only for information, but also to get a sense of the types of themes you might want to look for in your own research review.

Narrow the topic.  the narrower your topic, the easier it will be to limit the number of sources you need to read in order to obtain a good survey of relevant resources., consider whether your sources are current and applicable.  s ome disciplines require that you use information that is as current as possible. this is very common in the sciences where research conducted only two years ago could be obsolete. however, when writing a review in the social sciences, a survey of the history of the literature may be what is needed because what is important is how perspectives have changed over the years or within a certain time period. try sorting through some other current bibliographies or literature reviews in the field to get a sense of what your discipline expects. you can also use this method to consider what is consider by scholars to be a "hot topic" and what is not., follow the bread crumb trail.  the bibliography or reference section of sources you read are excellent entry points for further exploration. you might find resourced listed in a bibliography that points you in the direction you wish to take your own research., ways to organize your literature review, chronologically:  .

If your review follows the chronological method, you could write about the materials according to when they were published or the time period they cover.

By Publication:  

Order your sources chronologically by publication date, only if the order demonstrates a more important trend. For instance, you could order a review of literature on environmental studies of brown fields if the progression revealed, for example, a change in the soil collection practices of the researchers who wrote and/or conducted the studies.

Conceptual Categories:

The literature review is organized around a topic or issue, rather than the progression of time. However, progression of time may still be an important factor in a thematic review. For example, a review of the Internet’s impact on American presidential politics could focus on the development of online political satire. While the study focuses on one topic, the Internet’s impact on American presidential politics, it will still be organized chronologically reflecting technological developments in media. The only difference here between a "chronological" and a "thematic" approach is what is emphasized the most.

Methodological:  

A methodological approach focuses on the methods utilized by the researcher.  A methodological scope will influence either the types of documents in the review or the way in which these documents are discussed.

Sections of Your Literature Review:  

Once you've decided on the organizational method for your literature review, the sections you need to include should be easy to figure out because they arise from your organizational strategy.

Here are examples of other sections you may need to include depending on the type of review you write:

• Current Situation : information necessary to understand the topic or focus of the literature review.
• History : the chronological progression of the field, the literature, or an idea that is necessary to understand the literature review, if the body of the literature review is not already a chronology.
• Selection Methods : the criteria you used to select (and perhaps exclude) sources in your literature review. For instance, you might explain that your review includes only peer-reviewed articles and journals.
• Standards : the way in which you present your information.
• Questions for Further Research : What questions about the field has the review sparked? How will you further your research as a result of the review?

Writing Your Literature Review

Once you've settled on how to organize your literature review, you're ready to write each section. When writing your review, keep in mind these issues.

Use Evidence:

A literature review in this sense is just like any other academic research paper. Your interpretation of the available sources must be backed up with evidence to show that what you are saying is valid.

Be Selective:  

Select only the most important points in each source to highlight in the review. The type of information you choose to mention should relate directly to the research problem, whether it is thematic, methodological, or chronological.

Use Quotes Sparingly:  

Some short quotes are okay if you want to emphasize a point, or if what the author said just cannot be rewritten in your own words. Sometimes you may need to quote certain terms that were coined by the author, not common knowledge, or taken directly from the study. Do not use extensive quotes as a substitute your own summary and interpretation of the literature.

Summarize and Synthesize:  

Remember to summarize and synthesize your sources within each paragraph as well as throughout the review. Recapitulate important features of a research study, but then synthesize it by rephrasing the study's significance and relating it to their own work.

Keep Your Own Voice:  

While the literature review presents others' ideas, your voice (the writer's) should remain front and center. For example, weave references to other sources into what you are writing but maintain your own voice by starting and ending the paragraph with your own ideas and wording.

Use Caution When Paraphrasing:  

When paraphrasing a source that is not your own, be sure to represent the author's information or opinions accurately and in your own words. Even when paraphrasing an author’s work, you still must provide a citation to that work.

• << Previous: Getting Started
• Next: Mistakes to Avoid & Additional Help >>
• Last Updated: Oct 3, 2023 2:44 PM
• URL: https://guides.franklin.edu/LITREVIEW

Writing the Literature Review: Common Mistakes and Best Practices

• First Online: 21 November 2023

Cite this chapter

• Kelly Heider 3  

Part of the book series: Springer Texts in Education ((SPTE))

409 Accesses

1 Altmetric

The literature review is an essential component of academic research writing, providing a comprehensive overview of existing research and informing the development of new studies. However, writing an effective literature review can be a challenging task for many authors, particularly those new to academic writing. This chapter aims to guide authors through the process of writing a literature review by highlighting common mistakes and best practices. The chapter begins with three short narratives that describe difficulties both novice and prolific authors encounter when writing the literature review. A chapter activity follows with steps that guide authors through the process of developing a research question to frame the literature review. Authors are then prompted to complete a self-assessment activity which includes a series of questions designed to build their skills as academic research writers. The body of the chapter recommends strategies and techniques to help authors locate and evaluate sources that will serve as the building blocks for a literature review that is thorough, current, and well-written. The chapter concludes with a discussion of the threats and benefits of artificial intelligence-based text production in relationship to academic research writing. Overall, this chapter provides practical guidance for authors looking to improve their literature review writing skills and enhance the quality of their research output.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

• Available as PDF
• Read on any device
• Instant download
• Own it forever
• Available as EPUB and PDF
• Compact, lightweight edition
• Dispatched in 3 to 5 business days
• Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Anson, C. M. (2022). AI-based text generation and the social construction of “fraudulent authorship”: A revisitation. Composition Studies, 50 (1), 37–46.

Google Scholar  

Aylward, K., Sbaffi, L., & Weist, A. (2020). Peer-led information literacy training: A qualitative study of students’ experiences of the NICE evidence search student champion scheme. Health Information and Libraries Journal, 37 (3), 216–227. https://doi.org/10.1111/hir.12301

Article   Google Scholar  

Bohannon, J. (2013, October 4). Who’s afraid of peer review? Science, 342 (6154), 60–65. https://www.science.org/doi/10.1126/science.342.6154.60

Bouchrika, I. (2023, March 17). Top 10 qualities of good academic research . Research. https://research.com/research/top-10-qualities-of-good-academic-research

Bowler, M., & Street, K. (2008). Investigating the efficacy of embedment: Experiments in information literacy integration. Reference Services Review, 36 , 438–449. https://doi.org/10.1108/00907320810920397

De La Torre, M. (2018, August 29). Academic racism: The repression of marginalized voices in academia . The Activist History Review. https://activisthistory.com/2018/08/29/academic-racism-the-repression-of-marginalized-voices-in-academia/

Drewes, K., & Hoffman, N. (2010). Academic embedded librarianship: An introduction. Public Services Quarterly, 6 (2–3), 75–82. https://doi.org/10.1080/15228959.2010.498773

Elsevier Author Services. (n.d.a). Journal acceptance rates: Everything you need to know . Publication process. https://scientific-publishing.webshop.elsevier.com/publication-process/journal-acceptance-rates/#:~:text=your%20paper%20to%3F-,What%20Our%20Research%20Shows,over%201%25%20to%2093.2%25 .

Elsevier Author Services. (n.d.b). What is a good H-index? Publication recognition. https://scientific-publishing.webshop.elsevier.com/publication-recognition/what-good-h-index/

Emerald Publishing. (2023). How to…conduct empirical research . https://www.emeraldgrouppublishing.com/how-to/research-methods/conduct-empirical-research

Enago Academy. (2022, May 3). How much research is enough for a good literature review? https://www.enago.com/academy/research-for-a-good-literature-review/

Fitzgibbons, M. (2021). Literature review synthesis matrix . Concordia University Library. [Adapted from original table by the WI+RE Team at UCLA Library.] https://www.concordia.ca/content/dam/library/docs/research-guides/gradproskills/Lit-review-synthesis-matrix-Word.docx

Garrett, W. (n.d.) Marginalized populations . Minnesota Psychological Association. https://www.mnpsych.org/index.php?option=com_dailyplanetblog&view=entry&category=division%20news&id=71:marginalized-populations

George Mason University Libraries. (2021, August 20). Find authors . Finding diverse voices in academic research. https://infoguides.gmu.edu/c.php?g=1080259&p=7871669

Gyles, C. (2014, February). Can we trust peer-reviewed science? The Canadian Veterinary Journal, 55 (2), 109–111. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3894865/#b2-cvj_02_109

Hern, A. (2022, December 4). AI bot ChatGPT stuns academics with essay-writing skills and usability . The Guardian. https://www.theguardian.com/technology/2022/dec/04/ai-bot-chatgpt-stuns-academics-with-essay-writing-skills-and-usability

Hoffman, N., Beatty, S., Feng, P., & Lee, J. (2017). Teaching research skills through embedded librarianship. Reference Services Review, 45 , 211–226. https://doi.org/10.1108/RSR-07-2016-0045

Indeed Editorial Team. (2022, June 28). Create a theoretical framework for your research in 4 steps . Indeed. https://www.indeed.com/career-advice/career-development/theoretical-framework

Jansen, D. (2021, June). How to choose your research methodology . GradCoach. https://gradcoach.com/choose-research-methodology/

Jansen, D., & Warren, K. (2020, June). What (exactly) is a literature review? GradCoach. https://gradcoach.com/what-is-a-literature-review/

Jin, Y. Y., Noh, H., Shin, H., & Lee, S. M. (2015). A typology of burnout among Korean teachers. Asia-Pacific Education Researcher, 24 (2), 309–318. https://doi.org/10.1007/s40299-014-0181-6

Koufogiannakis, D., Buckingham, J., Alibhai, A., & Rayner, D. (2005). Impact of librarians in first-year medical and dental student problem-based learning (PBL) groups: A controlled study. Health Information and Libraries Journal, 22 , 189–195. https://doi.org/10.1111/j.1471-1842.2005.00559.x

Larsen, C. M., Terkelsen, A. S., Carlsen, A. F., & Kristensen, H. K. (2019). Methods for teaching evidence-based practice: A scoping review. BMC Medical Education, 19 , 1–33. https://doi.org/10.1186/s12909-019-1681-0

Liberties EU. (2021, October 5). What is marginalization? Definition and coping strategies . https://www.liberties.eu/en/stories/marginalization-and-being-marginalized/43767

Lucey, B., & Dowling, M. (2023). ChatGPT: Our study shows AI can produce academic papers good enough for journals—just as some ban it . The Conversation. https://theconversation.com/chatgpt-our-study-shows-ai-can-produce-academic-papers-good-enough-for-journals-just-as-some-ban-it-197762

Maior, E., Dobrean, A., & Păsărelu, C. (2020). Teacher rationality, social-emotional competencies, and basic needs satisfaction: Direct and indirect effects on teacher burnout. Journal of Evidence—Based Psychotherapies, 20 (1), 135–152. https://doi.org/10.24193/jebp.2020.1.8

Mertens, D. M. (2019). Research and evaluation in education and psychology: Integrating diversity with quantitative, qualitative, and mixed methods (5th ed.). SAGE Publications.

Monash University. (2023). Structuring a literature review . Learn HQ. https://www.monash.edu/learnhq/excel-at-writing/how-to-write.../literature-review/structuring-a-literature-review

Nova, A. (2017, December 21). Learn how to write a literature review in simple steps . MyPerfectWords. https://myperfectwords.com/blog/research-paper-guide/how-to-write-a-literature-review

O’Byrne, I. (2018, February 9). Eight steps to write a literature review . https://wiobyrne.com/literature-review/

Online Campus Writing Center. (2023). Synthesis . The Chicago School of Professional Psychology. https://community.thechicagoschool.edu/writingresources/online/Pages/Synthesis.aspx

Phair, D. (2021, June). Writing a literature review: 7 common (and costly) mistakes to avoid . GradCoach. https://gradcoach.com/literature-review-mistakes/

Robinson, K. A., Akinyede, O., Dutta, T., Sawin, V. I., Li, T., Spencer, M. R., Turkelson, C. M., & Weston, C. (2013, February). Framework for determining research gaps during systematic review: Evaluation . Agency for Healthcare Research and Quality (U.S.). https://www.ncbi.nlm.nih.gov/books/NBK126702/

Rommelspacher, A. (2020, November). How to structure your literature review: Three options to help structure your chapter . GradCoach. https://gradcoach.com/literature-review-structure/

Royal Literary Fund. (2023). The structure of a literature review . https://www.rlf.org.uk/resources/the-structure-of-a-literature-review/

Rudestam, K. E., & Newton, R. R. (1992). Surviving your dissertation: A comprehensive guide to content and process . SAGE Publications.

Scimago Lab. (2022a). About us . Scimago Journal and Country Rank. https://www.scimagojr.com/aboutus.php

Scimago Lab. (2022b). Journal rankings . Scimago Journal and Country Rank. https://www.scimagojr.com/journalrank.php

Shumaker, D. (2009). Who let the librarians out? Embedded librarianship and the library manager. Reference & User Services Quarterly, 48 (3), 239–257.

Statistics Solutions. (2023). Tips for the literature review: Synthesis and analysis . Complete dissertation. https://www.statisticssolutions.com/tips-for-the-literature-review-synthesis-and-analysis/

TechTarget. (2023). What is OpenAI? Open AI. https://www.techtarget.com/searchenterpriseai/definition/OpenAI

Texas A&M University Writing Center. (2023). Self-assessment . https://writingcenter.tamu.edu/Faculty-Advisors/Resources-for-Teaching/Feedback/Self-Assessment

University of Colorado at Colorado Springs. (2022, May 13). Evaluate sources. Public administration research. https://libguides.uccs.edu/c.php?g=617840&p=4299162

University of Maryland Global Campus. (n.d.). Discipline-specific research methods . https://coursecontent.umgc.edu/umgc/shareable-content/toolkits/BEHS000/1402/ResearchMethods/Discipline-SpecificResearchMethods.html

University of Melbourne. (2017, September 21). Academic writing: Writing the literature review [Video]. YouTube. https://www.youtube.com/watch?v=70n2-gAp7J0

Walker, R., & Solvason, C. (2014). Methodology: Choosing your research approach . SAGE Publications.

Wardle, E., & Downs, D. (2019). Writing about writing (4th ed.). Bedford/St. Martin’s.

Wiggers, K. (2023, January 31). OpenAI releases tool to detect AI-generated text, including from ChatGPT . TechCrunch. https://techcrunch.com/2023/01/31/openai-releases-tool-to-detect-ai-generated-text-including-from-chatgpt/

Yilmaz, K., Altinkurt, Y., Guner, M., & Sen, B. (2015). The relationship between teachers’ emotional labor and burnout level. Eurasian Journal of Educational Research, 59 , 75–90.

Download references

Author information

Authors and affiliations.

Acting Assistant Vice Provost for University Libraries, Indiana University of PA, Indiana, PA, USA

Kelly Heider

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Kelly Heider .

Editor information

Editors and affiliations.

Indiana University of Pennsylvania, Emerita, Indiana, PA, USA

Mary Renck Jalongo

Teaching & Learning, Policy & Leadership, University of Maryland, College Park, College Park, MD, USA

Olivia N. Saracho

Issue: AI-Based Text Production in Academic Research Writing

 Artificial Intelligence (AI) is a relatively new technological development that can be used to support academic research writing, particularly reviews of research. Although it holds promise as a research tool, there are also concerns that AI-based scholarly writing could threaten not only the integrity but also the creativity of academic research writing. ChatGPT, an advanced chatbot developed by OpenAI (a nonprofit research company founded by Elon Musk and Sam Altman in 2015) has been at the center of much controversy as of late because of its ability to produce academic papers using artificial intelligence-based text production that have fooled academics and academic journal publishers (Hern, 2022 ; Lucey & Dowling, 2023 ; TechTarget, 2023 ). To demonstrate ChatGPT’s capabilities, consider its response to the question, “What is artificial intelligence-based text production?”.

Artificial Intelligence-Based Text Production is a process where machine learning algorithms are used to generate written content such as articles, product descriptions, social media posts, and more. It involves training machine learning models with vast amounts of data to learn how to generate human-like text automatically. The process can involve different techniques such as Natural Language Processing (NLP), deep learning, and neural networks to analyze and understand language patterns, grammar, context, and semantics. The resulting AI-generated text can be used for a wide range of applications, including content marketing, chatbots, language translation, and more. One of the main benefits of AI-based text production is that it can help businesses and organizations automate their content creation process, saving time and resources while producing high-quality content. However, it is important to note that AI-generated text may not always be perfect and may require human editing and review to ensure accuracy and coherence.

This response was produced by ChatGPT in about ten seconds. According to Anson ( 2022 ), “cutting and pasting all or part of this auto-generated text into a new text box can continue to prompt further content. The material is not plagiarized from existing sources but freshly composed. For this reason, it usually evades plagiarism-detection programs like Turnitin” (p. 40).

How Might AI-Based Text Production Threaten Academic Research Writing?

Obviously, computer-generated text that evades plagiarism-detection programs threatens the integrity of academic research writing. Some academic publishers have already banned or limited the use of AI-generated text in papers submitted to their journals (Lucey & Dowling, 2023 ). However, that is easier said than done. OpenAI recently developed a tool that attempts to distinguish between human-written and AI-generated text to prevent chatbots like ChatGPT from being abused, but it is only 26% effective (Wiggers, 2023 ).

Lucey and Dowling ( 2023 ) tested the credibility of ChatGPT by having expert reviewers examine papers produced by the chatbot. First, they asked ChatGPT to generate four parts of a research study: (1) research idea, (2) literature review, (3) dataset, and (4) suggestions for testing and examination. They chose a broad subject and instructed the chatbot to create a paper that could be published in “a good finance journal” (para. 6). Second, they pasted 200 relevant abstracts into the ChatGPT search box and asked the chatbot to consider the abstracts when generating the four-part research study. Finally, they asked academic researchers to read both versions of the AI-generated text and make suggestions for improvement. A panel of thirty-two reviewers read all versions of the four-part research study and rated them. In all cases, the papers were considered acceptable by the reviewers, although the chatbot-created papers that also included input from academic researchers were rated higher. However, “a chatbot was deemed capable of generating quality academic research ideas. This raises fundamental questions around the meaning of creativity and ownership of creative ideas—questions to which nobody yet has solid answers” (Lucey & Dowling, 2023 , para. 10).

How Might AI-Based Text Production Benefit Academic Research Writing?

Despite several publishers deciding to ban the inclusion of AI-based text production in submissions, some researchers have already listed ChatGPT as a co-author on their papers (Lucey & Dowling, 2023 ). There are many who believe there is no difference between the way ChatGPT produces text and the way authors synthesize studies in their literature reviews. In fact, the chatbot’s review is much more exhaustive because it can analyze “billions of existing, human-produced texts and, through a process akin to the creation of neural networks, generate new text based on highly complex predictive machine analysis” (Anson, 2022 , p. 39).

There are other advantages to using AI-based text production. It has the potential to aid groups of researchers who lack funding to hire human research assistants such as emerging economy researchers, graduate students, and early career researchers. According to Lucey and Dowling ( 2023 ), AI-based text production “could help democratize the research process” (para. 18). Anson ( 2022 ) also sees the potential in AI-based text production to “spark some new human-generated ideas” (p. 42), extract keywords, and create abstracts. The development of AI-based text production might also force instructors to change the way they teach academic writing. Instead of trying to detect or prevent the use of chatbots like ChatGPT, “a more sensible approach could involve embracing the technology, showing students what it can and can’t do, and asking them to experiment with it” (Anson, 2022 , p. 44). In other words, students could be asked to write about writing which leads to a deeper understanding of the writing process and the ability to transfer that understanding to any writing project (Wardle & Downs, 2019 ).

The Responsible Use of AI-Based Text Production in Academic Research Writing

The responsible use of AI-based text production in academic research writing involves understanding the technology's capabilities and limitations, as well as considering its potential impact on the research process. Researchers must carefully evaluate the intended purpose and context of using AI-generated text and make certain they are not compromising the authenticity and integrity of their research work. To ensure responsible use, it is essential to balance the benefits of increased efficiency and new insights with the need for originality and critical thinking in academic research writing. Researchers must also be transparent in disclosing the use of AI-generated text when submitting their work for publication. By adopting a responsible and thoughtful approach to the use of AI-based text production, researchers can maximize the benefits of the technology while maintaining the quality and authenticity of their research.

Applications of Technology

How to Write a Paper in a Weekend : https://youtu.be/UY7sVKJPTMA

Note : University of Minnesota Chemistry Professor, Peter Carr is not advocating for procrastination. This video outlines a strategy for generating a first draft after you have all your reading and notes assembled.

Research Gap 101: What Is a Research Gap & How to Find One : https://youtu.be/Kabj0u8YQ4Y

Using Google Scholar for Academic Research : https://youtu.be/t8_CW6FV8Ac .

Rights and permissions

Reprints and permissions

Copyright information

© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG

About this chapter

Heider, K. (2023). Writing the Literature Review: Common Mistakes and Best Practices. In: Renck Jalongo, M., Saracho, O.N. (eds) Scholarly Writing. Springer Texts in Education. Springer, Cham. https://doi.org/10.1007/978-3-031-39516-1_3

Download citation

DOI : https://doi.org/10.1007/978-3-031-39516-1_3

Published : 21 November 2023

Publisher Name : Springer, Cham

Print ISBN : 978-3-031-39515-4

Online ISBN : 978-3-031-39516-1

eBook Packages : Education Education (R0)

Share this chapter

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

• Publish with us

Policies and ethics

• Find a journal
• Track your research