A literature review is not a long list of studies. A good review helps readers understand how a topic has developed, what the main lines of argument are, where findings converge or diverge, and how the current study contributes to an ongoing academic conversation.
When supervising theses, dissertations and manuscripts, I often see the same problem. The writer has read many papers, but the review reads like a sequence of summaries. One paragraph begins with one author, reports a finding and then moves to another author. This proves that the writer has read, but it does not prove that the writer understands the logic of the field. The reader still cannot see how the studies connect, why findings differ and where the new study should stand.
A useful literature review performs three tasks at the same time. First, it organizes existing knowledge into meaningful lines of argument. Second, it identifies agreement, difference or tension across studies. Third, it leads logically to the research question, model or hypotheses of the current study. Without these functions, the review becomes a record of “what has been read” rather than an argument about “what is known and what should be examined next.”
Start from the problem, not from the folder of papers
Many writers open their reference folder and begin arranging papers by year, author or journal. That may help with file management, but it is not enough for argument development. A better starting point is to define the research problem clearly: what phenomenon requires explanation, in which context, with which group of actors and why the question matters. Once the problem is clear, the literature becomes evidence for understanding the problem rather than a pile of disconnected papers.
For example, if a study examines customer experience in omnichannel retailing, the writer should not simply list articles that contain the word omnichannel. It is more useful to ask what type of experience is being studied, which touchpoints matter, how customers evaluate seamlessness and whether loyalty means repeat behavior, attachment or advocacy. With these questions in mind, the literature can be grouped around customer journey, channel integration, service quality, customer data and behavioral outcomes.
Example
Instead of writing, “A studied customer experience, B studied loyalty, and C studied omnichannel retailing,” the writer can write: “Studies of omnichannel retailing suggest that customer experience is shaped not only by individual touchpoints but also by the perceived continuity across channels. One stream emphasizes information consistency, while another focuses on channel switching and customer control. This distinction implies that omnichannel experience should be treated as a construct with both operational and perceptual dimensions.” The second version shows synthesis and argument, not only citation.
Group the literature by ideas
A strong review is usually organized by ideas rather than by author names. These ideas may include foundational theories, central concepts, antecedents, outcomes, research contexts or measurement approaches. Once the groups are formed, each group should be written as a line of reasoning. The opening sentence states the main point. The following sentences use studies to support, compare, challenge or refine that point.
For a topic such as applied AI at work, the literature may be organized into four groups: technological capability, workflow design, user acceptance and risk control. If each study is summarized separately, readers may not see the structure of the field. If the four groups are used, readers can see that AI is not only a tool issue. It is a managerial issue involving technology, work design, people and governance.
| Review logic | When it is useful | Common weakness |
|---|---|---|
| By concept | When a topic has several definitions or measures. | Definitions are listed without explaining their differences. |
| By theory | When the study needs to build a model or hypotheses. | Theories are named but not used to explain relationships. |
| By context | When the phenomenon differs across industries, countries or user groups. | Context is described without drawing research implications. |
| By method | When findings vary because of measurement, sample or analysis technique. | The method is labelled as quantitative or qualitative without evaluating design quality. |
A research gap must be demonstrated
A research gap should not be written as a loose statement such as “there are few studies.” That sentence is weak because it does not specify what is missing, in what sense it is missing and why the gap matters. A convincing gap must emerge from synthesis. It may be a contextual gap, when most studies focus on developed markets but little is known about Vietnam. It may be a conceptual gap, when a variable is widely used but its measurement remains inconsistent. It may be a mechanism gap, when prior studies test direct effects but do not explain the conditions or processes through which those effects occur.
The writer should also distinguish an academic gap from personal interest. Being interested in a topic is not enough to make it a research gap. A gap must appear from the literature and must matter to knowledge development. If a study only repeats an old model in a new sample without explaining why the new context may change the mechanism, the academic contribution will be weak.
From review to research model
The literature review should not end abruptly and then be followed by a model as if the two parts were unrelated. The model should be the natural outcome of the review. After presenting theories and evidence, the writer needs to explain why each variable is selected, why the proposed relationships are reasonable and why the present context makes the test necessary. This is the difference between a model that is merely drawn and a model that is argued for.
A simple test is to read every hypothesis and ask: which part of the literature review supports this hypothesis? If the answer is unclear, the review has not done its job. Conversely, if the review contains long passages that do not support any question, variable or hypothesis, those passages should be shortened or removed.
Write the review as an academic conversation
A good review does more than report what studies have said. It places them in conversation. How does one study extend another? Are different results caused by context, measurement or method? Which theory explains the phenomenon more convincingly? Which assumption is widely used but not yet tested? When a writer asks these questions, the review becomes more analytical and shows the author’s academic judgement.
In practice, writers should prepare a reading matrix that records the research question, theory, variables, method, sample, main findings and limitations of each study. But the matrix is a working tool, not the review itself. In the manuscript, the information in the matrix must be transformed into paragraphs with argument, comparison and intermediate conclusions. This is the process of turning reading notes into academic reasoning.
Conclusion
The literature review is the foundation of a study. It does not only prove that the writer has read enough. More importantly, it shows that the writer understands the field, knows how to locate the study within a scholarly conversation and can develop a meaningful research question. To do this well, begin with the problem, group the literature by ideas, demonstrate the research gap and connect the review to the model. Written in this way, the literature review is no longer a compulsory section. It becomes the logic that supports the entire study.
References
Booth, A., Sutton, A., & Papaioannou, D. (2016). Systematic approaches to a successful literature review (2nd ed.). SAGE.
Hart, C. (2018). Doing a literature review: Releasing the research imagination (2nd ed.). SAGE.
Snyder, H. (2019). Literature review as a research methodology: An overview and guidelines. Journal of Business Research, 104, 333–339. https://doi.org/10.1016/j.jbusres.2019.07.039
Webster, J., & Watson, R. T. (2002). Analyzing the past to prepare for the future: Writing a literature review. MIS Quarterly, 26(2), xiii–xxiii.
