A critical literature review: from searching and reading to a coherent argument
Twenty studies described one by one, in chronological order, do not yet form an argument. A literature review should compare evidence, evaluate disagreements, identify what remains unanswered and lead directly to your question.
A critical review develops a reasoned account of what the literature does and does not establish. Organise it by theme, narrow from the broader field to your question, distinguish findings from opinions and connect the gap with your contribution. Searching, reading, note-taking and writing are iterative, so plan the search and organise notes from the beginning.
1. What is the review for?
The difficult part is evaluating individual contributions and integrating them into an organised account, rather than simply finding and reading them. Literature informs the initial idea and proposal, the context and theoretical framework, and the discussion of your findings (Saunders et al., 2023, pp. 72–73).
The book describes a student review resembling a catalogue: equally long summaries arranged by author or year, without comparison. The space devoted to a study should reflect its relevance to your question (Saunders et al., 2023, p. 74).
A review can establish history, locate current debates, address inconsistent evidence, explain concepts and terminology, identify work to extend or challenge, and justify the question. Deductive work develops a framework for testing; inductive work identifies a phenomenon requiring further exploration (Saunders et al., 2023, p. 78).
2. What does “critical” mean?
Critical does not mean hostile. It means constructive evaluation, informed scepticism and an argument supported by evidence. Saunders et al. (2023, pp. 75–76), drawing on Mingers, discuss scrutiny of reasoning, conventional views, dominant perspectives and the values embedded in knowledge.
Evaluate established work fairly, consider supporting and contradictory studies, explain relevance, separate findings from author opinions and cite accurately (Saunders et al., 2023, p. 77). Galvan and Galvan (2024, pp. 59, 76) similarly distinguish claims from evidence and advise looking for limitations that materially affect interpretation.
Agreement may motivate new evidence, testing an assumption or examining another context. Disagreement may concern definitions, classifications or causal reasoning. Explain the issue rather than merely announcing agreement or disagreement (Booth et al., 2024, Section 4.3).
3. Forms of literature review
Saunders et al. (2023, p. 79) distinguish six forms that can be combined within a project.
| Form | Focus | Useful when |
|---|---|---|
| Integrative | Key ideas and relationships across selected literature | Developing an overview or conceptual framework; common in dissertations |
| Theoretical | Accumulated theories and their relationships | Explaining why existing theory does not fully address a problem |
| Historical | Development of a topic over time | Locating the issue historically and identifying future directions |
| Methodological | Approaches, strategies, collection and analysis | Evaluating a method or methodological choice |
| Argumentative | Selected evidence supporting or challenging a position | Developing an alternative perspective while acknowledging selection bias |
| Systematic | Prespecified, comprehensive identification, appraisal and synthesis | Answering a well-defined review question transparently |
4. Structure by theme and narrow the focus
Replace the sequence “author A, author B, author C” with themes in which studies can be compared. Saunders et al. (2023, p. 81) suggest a funnel:
Begin broadly
Introduce the field and narrow towards the question and objectives.
Introduce the main themes
Give readers a clear map of the discussion.
Compare key studies within themes
Explain agreement, disagreement and relevant differences.
Narrow to the most relevant evidence
Keep the question in view.
Explain key theories and findings
Show how the evidence fits together.
Identify your contribution
Make the connection between gap and question explicit.
Lead into the rest of the study
Explain how later chapters address the gap.
A thematic outline helps reorder notes around the argument rather than the order of reading. Record contrasts and unanswered issues within each theme (Galvan and Galvan, 2024, pp. 99–101).
Issues raised here should reappear when interpreting findings. In SEM work, relate supported and unsupported hypotheses to the studies used to develop them (Saunders et al., 2023, p. 81).
5. Plan the search
Write down scope, keywords, databases and selection criteria, then discuss the plan with your supervisor. Scope includes language, discipline, industry, geography, publication period and document type. An apparent lack of studies may reflect narrow limits or unsuitable terminology rather than a genuine gap (Saunders et al., 2023, pp. 91–92).
Recent reviews and influential studies help identify accepted keywords. Combine synonyms and concepts deliberately:
| Operator | Purpose | Example |
|---|---|---|
| AND | Narrow: both concepts must appear | “green HRM” AND “hotel” |
| OR | Broaden: at least one alternative appears | “green HRM” OR “environmental HRM” |
| NOT | Exclude a term, with care | “green HRM” NOT “manufacturing” |
Searches can fail through British–American spelling differences, outdated terms, informal vocabulary, narrow date limits or inappropriate source types (Saunders et al., 2023, p. 97). Begin with recent work, trace backwards to reviews and theory, and identify seminal studies (Galvan and Galvan, 2024, pp. 33–35).
6. Read and evaluate
Wallace and Wray's questions, as discussed by Saunders et al. (2023, p. 105), are: why am I reading this, what is the author trying to do, what relates to my question, how convincing is the reasoning, and how can I use it?
Separate relevance from quality. Relevance concerns your question and prespecified inclusion criteria. Quality concerns methods, theoretical grounding and reasoning. Business memoirs and trade publications may offer useful perspectives without the evidential strength of systematic research (Saunders et al., 2023, p. 106).
Rankings are broad signals. Articles within the same journal differ, and useful evidence can appear outside highly ranked outlets. Appraise each study (Saunders et al., 2023, p. 106).
For quantitative studies, record experimental status, random assignment, measurement reliability and validity, differences in operationalisation, sampling and effect sizes rather than significance alone (Galvan and Galvan, 2024, pp. 65–76). Two studies using the same construct label may employ non-equivalent scales, limiting direct comparison.
7. Notes and a synthesis matrix
Record complete bibliographic information before taking substantive notes, including DOI where available (Booth et al., 2024, Section 4.1; Saunders et al., 2023, pp. 108–109). A spreadsheet with one study per row and comparison dimensions as columns supports synthesis. Tables of definitions or methods can help, but must be discussed in the prose (Galvan and Galvan, 2024, pp. 47, 87–92).
| Author/year | Context/sample | Theory | Main measures | Analysis | Findings | Limitations/gap |
|---|---|---|---|---|---|---|
| [Study] | Country, industry, n | [Framework] | Scale source and items | CB-SEM, PLS-SEM, regression, etc. | Estimates and uncertainty | Sample, cross-sectional design, unresolved issue |
Ordering rows chronologically while keeping thematic columns can also help reveal developments over time (Saunders et al., 2023, p. 113).
8. Systematic reviews
A systematic review uses an explicit, reproducible strategy to identify, appraise, synthesise and report evidence. It can be demanding and is suited to defined questions, including unresolved issues or intervention effectiveness (Saunders et al., 2023, pp. 110–111).
The book outlines five stages: define scope and question, possibly using context–intervention–mechanism–outcome (CIMO); search comprehensively; screen titles, abstracts and full texts using prespecified criteria and extract data; analyse and synthesise; and report the strategy and decisions (pp. 111–112). PRISMA 2020 provides reporting guidance, including a flow diagram for record selection (Page et al., 2021). A diagram alone does not make a review systematic.
9. Drafting and avoiding plagiarism
Ask how far existing work answers your question. What remains should motivate the rest of the study. Check that the title conveys focus, the search and selection are explained, themes narrow logically, subheadings guide reading and the ending leads onward (Saunders et al., 2023, pp. 113, 116).
- Avoid unsupported global claims in the opening.
- Specify dates when calling a topic new or increasingly studied.
- Explain where and how you searched before claiming that no studies exist.
- Replace vague strings of citations with identifiable findings.
- Group conflicting evidence and consider measurement, sample or context as possible explanations.
- Make the need for your study visible throughout the review.
These writing principles are developed in Galvan and Galvan (2024, pp. 108–120).
Changing a few words or following another author's argument without attribution is not adequate paraphrasing. Do not cite an original as though you read it when you only encountered it through another source; identify the secondary citation (Saunders et al., 2023, pp. 116–117).
Label copied wording immediately and record the page. Mixed notes containing unmarked quotations and your own sentences are a common route to accidental plagiarism (Galvan and Galvan, 2024, p. 49).
10. Sample comparison, mistakes and checklist
Illustrative thematic paragraph
Evidence concerning [A] and [B] is inconsistent. [Study 1] and [Study 2] reported positive relationships in [context], with samples of [n1] and [n2]. [Study 3] did not find a statistically significant relationship in [other context]. Measurement differences may help explain the contrast: the first two used [scale], whereas the third used [alternative]. All three were cross-sectional and did not examine [mediator/moderator]. The present study evaluates the relationship in [context] and includes [construct] to investigate the proposed mechanism.
| Weak approach | Problem | Better approach |
|---|---|---|
| One paragraph per study in date order | Description without synthesis | Compare studies within themes |
| “No one has studied this” | Unverifiable scope | Specify databases, years, search terms and the precise gap |
| “Many studies show…” with six citations | Unclear findings and contexts | Identify representative studies and their results |
| Citing Ajzen (1991) after reading only a later article | Misrepresents source access | Read the original or identify the secondary source |
| A gap disconnected from the question | No rationale for the present study | Link gaps to questions or hypotheses |
Checklist
- Document scope, keywords, databases and selection criteria.
- Read recent reviews and seminal studies.
- Use a matrix with complete references and DOIs.
- Organise themes from broad to specific.
- Separate evidence from opinion.
- Discuss supporting and contradictory findings and plausible reasons for disagreement.
- Connect a specific gap with your question or hypotheses.
- Use accurate citations and genuine paraphrasing.
Conclusion
A strong review is measured by the reader's understanding of what is known, disputed and unanswered, and why your question is a reasonable next step. Much of that quality is established before drafting, through a clear search plan, purposeful reading and notes detailed enough for comparison.
References
Main source
- Saunders, M. N. K., Lewis, P., & Thornhill, A. (2023). Research methods for business students (9th ed.). Pearson. Pearson
Additional sources
- Booth, W. C., Colomb, G. G., Williams, J. M., Bizup, J., & FitzGerald, W. T. (2024). The craft of research (5th ed.). University of Chicago Press. https://doi.org/10.7208/chicago/9780226826660.001.0001
- Galvan, J. L., & Galvan, M. C. (2024). Writing literature reviews: A guide for students of the social and behavioral sciences (8th ed.). Routledge.
- Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L., Tetzlaff, J. M., Akl, E. A., Brennan, S. E., Chou, R., Glanville, J., Grimshaw, J. M., Hróbjartsson, A., Lalu, M. M., Li, T., Loder, E. W., Mayo-Wilson, E., McDonald, S., . . . Moher, D. (2021). The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ, 372, n71. https://doi.org/10.1136/bmj.n71
Sources are paraphrased with page or section references. Secondary citations are identified. Search examples, the synthesis matrix and sample paragraph were written for this article.
