03 · What You Need to Know
The Problem Disappears Once Themes Stop Behaving Like Exclusive Folders
Literature-review synthesis requires you to examine relationships across sources. James Cook University describes synthesis as bringing multiple sources together to identify patterns, themes, common findings, contradictions, and gaps, while George Mason University's Writing Center emphasizes organizing literature around topics or themes rather than writing a sequence of source summaries.
That has an important organizational consequence. Your paper collection and your thematic structure perform different jobs.
Your collection answers:
Which sources do I have?
Your themes answer:
Which sources contribute evidence or ideas to this particular analytical issue?
Once those functions are separated, there is no reason a source must have only one thematic relationship.
Keep One Source Record, but Allow Many Thematic Connections
Suppose a study investigates students' use of generative AI for writing assignments and reports findings about writing performance, student attitudes, academic integrity concerns, and differences according to prior AI experience.
You could reasonably connect that paper to several themes:
- learning or writing outcomes;
- student perceptions;
- academic integrity; and
- learner characteristics.
This does not require four copies of the paper. It requires one source record with four analytical relationships.
Source storage
Maintain one identifiable record for the paper, its citation metadata, PDF, and source-level notes.
Thematic organization
Connect that source to as many themes, concepts, claims, or debates as its content genuinely informs.
This distinction becomes especially useful when you keep PDFs, notes, and citation records connected. You can preserve a stable source record while allowing multiple pathways back to it.
Think in Terms of Many-to-Many Relationships
A useful mental model comes from databases rather than filing cabinets.
One paper can connect to several themes. One theme can connect to many papers.
In other words:
Paper A → Theme 1, Theme 2, Theme 4
Theme 1 → Paper A, Paper C, Paper D, Paper F
This many-to-many structure more accurately reflects how scholarship behaves. Research studies are often multidimensional, and themes usually emerge from patterns distributed across several sources.
University of Sheffield guidance on synthesis matrices explicitly organizes sources by theme and encourages researchers to examine how sources build on, challenge, contradict, or complement one another. A matrix naturally accommodates the fact that a source may contain relevant material in several thematic areas.
Do Not Assign the Whole Paper When Only One Finding Fits the Theme
There is another important refinement. Sometimes saying that a paper belongs to a theme is too coarse.
Imagine that a study measures three outcomes:
- academic performance;
- student satisfaction; and
- continued intention to use the technology.
The performance result may belong in your discussion of learning outcomes. The satisfaction result may contribute to student perceptions. The intention measure may inform technology acceptance.
Instead of recording only:
Paper A = Themes 1, 2, and 3
record what specifically connects the source to each theme:
| Theme |
Contribution From the Paper |
Source Location |
| Learning outcomes |
Reports change in writing-performance scores |
Results section / relevant table or page |
| Student perceptions |
Reports satisfaction with feedback |
Results section / relevant page |
| Technology acceptance |
Reports intention to continue using the tool |
Results section / relevant page |
Now the thematic connection carries analytical information rather than merely a label.
Use Tags When You Need Flexible Retrieval
Tags can be useful when themes overlap extensively or are still evolving. A single source can receive several tags without being physically duplicated.
The University of Arizona Libraries recommends tagging articles with themes and keywords to identify emerging patterns and grouping sources according to themes, methods, theories, or subtopics. This kind of flexible categorization can be useful because literature-review structures often change during analysis.
Your tags might represent:
- substantive themes;
- methods;
- theories;
- populations;
- contexts;
- important findings; or
- roles in an emerging argument.
Be careful, however, not to turn every observation into a tag. A tagging system containing dozens of nearly synonymous categories can become harder to navigate than the papers themselves.
Use a Matrix When You Need to See Theme-by-Source Intersections
A synthesis matrix is particularly well suited to overlapping themes because one axis can contain sources and the other themes. At each intersection, you record what the source contributes to that theme.
The University of Sheffield describes a synthesis matrix as a spreadsheet organized by theme that helps researchers examine relationships among sources. Similar guidance from writing centers treats the matrix as a way to identify overlap and divergence rather than merely catalogue papers.
| Source |
Learning Outcomes |
Student Perceptions |
Academic Integrity |
AI Literacy |
| Study A |
Improved revision performance |
Generally positive |
Not examined |
Prior experience may matter |
| Study B |
No clear difference |
Mixed perceptions |
Concern about appropriate use |
Not examined |
| Study C |
Not measured |
High perceived usefulness |
Students uncertain about boundaries |
Literacy linked to confidence |
Notice that Study A does not have to be assigned to one column. Its contribution is represented wherever relevant.
This is one reason a literature matrix designed around analytical comparisons can be more useful than a spreadsheet containing one generic "Theme" column.
Separate Descriptive Tags From Analytical Themes
Not every category serves the same purpose.
You may tag a paper as:
higher education; undergraduate; mixed methods; Philippines; generative AI
Those descriptors help you retrieve and compare sources, but they are not necessarily themes in the written literature review.
An analytical theme might instead be:
Students' judgments about appropriate AI use depend on how academic integrity boundaries are communicated.
The distinction matters because a literature review is not simply a taxonomy of papers. Themes should help you interpret the scholarly conversation.
As James Cook University notes, thematic organization should emerge from patterns, trends, inconsistencies, and relationships across sources, with sections or paragraphs focused on key ideas rather than individual studies.
Let Themes Overlap When the Phenomenon Overlaps
Researchers sometimes try to make themes completely distinct because overlap feels untidy. But conceptual overlap is not automatically a flaw.
Consider three themes:
- student trust in AI feedback;
- AI literacy;
- uptake of AI feedback.
These may be analytically distinguishable while remaining related. A paper could plausibly inform all three because students' ability to evaluate AI output may influence trust, which may in turn affect whether feedback is used.
Forcing the study into only one category would remove a potentially important relationship.
The more useful question is whether your themes are distinguishable enough to perform different analytical work, not whether they are perfectly non-overlapping.
But Heavy Overlap Can Signal That Your Themes Need Revision
Overlap is legitimate, but it can also diagnose a structural problem.
If nearly every paper belongs to Themes A, B, and C for essentially the same reason, perhaps those themes are not genuinely distinct.
Ask:
- Do the themes answer different analytical questions?
- Would separating them produce meaningfully different synthesis?
- Could two themes be dimensions of a broader theme?
- Is one theme actually a cause, mechanism, outcome, or context of another?
- Are the labels simply different words for the same underlying idea?
Sometimes the solution is to merge themes. In other cases, a hierarchical structure works better, with broad themes containing narrower subthemes.
Use Parent Themes and Subthemes When the Literature Has Levels
Suppose your broad theme is Student Engagement With AI Feedback.
Within it, the literature may distinguish:
- trust in feedback;
- perceived usefulness;
- feedback uptake; and
- revision behavior.
A paper could contribute to the broad parent theme and several subthemes without creating a conceptual problem.
Hierarchical themes can reduce unnecessary duplication when several narrow categories are clearly part of a larger phenomenon. They also help preserve the difference between broad organizing sections and more specific ideas that may structure paragraphs within those sections.
Do Not Duplicate Full Notes Every Time a Paper Appears Under Another Theme
Copying the same notes into several folders or documents creates a version-control problem.
You may later correct your interpretation in one copy but not the others. A page number may be added to one version while another remains incomplete. Eventually, you may no longer know which note is authoritative.
Watch Out
Multiple thematic connections are useful; multiple independent copies of the same source record are usually not. Maintain one authoritative source-level record and use tags, links, matrix intersections, references, or theme-level notes to point back to it.
This also reduces the risk of losing the original source of an idea as information moves from paper-level notes into thematic synthesis.
Your Written Review Can Reuse a Source Without Repeating the Same Summary
If a paper contributes to several themes, it may legitimately appear in several sections of the literature review.
The key is that each appearance should perform different analytical work.
Suppose Study A reports both improved writing performance and concerns about students accepting inaccurate AI feedback.
In a section on learning outcomes, you might use the study as evidence about writing performance.
In a section on AI literacy, you might use a different result from the same study to discuss students' ability to evaluate generated feedback.
That is not redundant merely because the citation appears twice.
Redundancy occurs when you repeatedly summarize the same purpose, method, and findings simply because the paper has been assigned to several themes.
Themes Are Tools for Synthesis, Not Permanent Properties of Papers
A theme does not exist inside a paper in quite the same way that its title or DOI does. Themes are analytical categories that you use to make sense of relationships across the literature.
That means thematic assignments can change.
A source you initially tag under "student attitudes" may later become important to a more precise theme concerning trust. A broad "learning outcomes" category may eventually separate into immediate performance and retention. Two categories may merge when you discover that the literature does not meaningfully distinguish them.
This flexibility is consistent with the iterative nature of literature reviewing. The University of Arizona describes literature reviews as requiring repeated cycles of searching, refining, and analyzing rather than a strictly linear process.
Your organizational system should therefore make reclassification inexpensive. If changing a theme requires manually moving or duplicating dozens of files, the system is working against the analysis.