An empirical review explains what relevant studies found, how they produced that evidence and what their findings mean together. Start with a question you need the literature to answer. Then group studies around that question, not around a list of authors.
“Author A found this; Author B found that” is useful as reading notes. It is rarely the finished review. Your reader needs to know whether the studies agree, whether their methods make them comparable and what remains uncertain.
Your reading notes need to become a connected argument. UniDraft helps organise the studies, source notes and project instructions you confirm into an editable Word document. Start your project document with UniDraft, then check every comparison and citation against the publications you read. See the project workflow for the inputs and output; you remain responsible for evaluating the evidence.
Build a source matrix before writing paragraphs
Use one row per verified study:
| Column | Question to answer while reading |
|---|---|
| Full reference and source location | Can I reopen the original paper and find the relevant passage? |
| Research question | What did this study investigate? |
| Context and participants or material | Who or what supplied the evidence? |
| Method | What was compared, measured or analysed? |
| Relevant finding | What does the evidence actually support? |
| Limits | Which claims would go beyond this study? |
| Link to my question | Why does this source belong in my review? |
Record page or section locations in your notes. An abstract helps you screen a paper, but it may not provide enough detail for a critique of sampling or analysis. Read the full study before making those claims. Leave a field unresolved if you cannot verify it.
Worked example using two real studies
The following example is about retrieval practice and memory. These are published studies, not Nigerian fieldwork or results from a UniDraft student. They are used to teach comparison, not to supply a ready-made literature review for an unrelated project.
| Study | Evidence relevant to this comparison | Boundary to preserve |
|---|---|---|
| Roediger and Karpicke (2006), Test-enhanced learning | Experiments compared further study with memory testing after reading prose. Testing benefited retention at delayed assessments, although immediate performance did not show the same pattern. | The material and assessment conditions matter; this does not establish an effect for every subject or classroom. |
| Karpicke and Roediger (2008), The critical importance of retrieval for learning | A vocabulary-learning experiment varied continued study and testing after an item had been recalled correctly. Continued retrieval supported delayed recall. | Vocabulary recall and reading prose are different learning tasks; the studies should not be treated as identical replications. |
Original records: 2006 paper and 2008 paper. These summaries deliberately omit sample sizes and numerical effect estimates rather than supplying unverified details. Consult the full papers before using them in your own assessed work.
Turn the comparison into a synthesis paragraph
Here is an original teaching paragraph based on that limited comparison:
Retrieval practice appears relevant to delayed remembering across more than one learning task. Roediger and Karpicke (2006) examined prose learning, whereas Karpicke and Roediger (2008) examined vocabulary learning and continued practice after initial success. Their findings support examining what learners can retrieve later, rather than judging learning solely by performance during study. However, the different materials and procedures limit a simple comparison of effect size. These papers alone do not establish how the approach would work in a particular Nigerian undergraduate course; that question would require a broader literature search and an appropriate local study.
The paragraph makes one point, connects both studies and ends at the boundary of the evidence. It does not claim “no Nigerian researcher has studied this.” Not finding a paper in two records is not proof that the research does not exist.
Organise the rest around useful comparisons
Possible paragraph themes include agreement about a relationship, conflicting findings, different measurement choices or differences between settings. Choose themes that serve your own research questions.
For each paragraph, ask:
- What is the point of this paragraph?
- Which studies support, qualify or challenge it?
- Are the findings comparable, or are the studies asking different questions?
- What can I reasonably conclude?
- How does this connect to my project?
A different finding is not automatically a contradiction. One study may measure intention while another measures actual behaviour; one may examine association and another a causal intervention. Explain the difference before declaring that the literature disagrees.
End with a justified research need
Use the review to narrow what your study can contribute. A new location alone does not automatically establish an important gap. Explain why the population, measurement, method or context could change the answer, and check recent reviews before claiming novelty.
The Chapter 2 guide covers the wider structure, source search and gap workflow. The conceptual versus theoretical framework guide helps keep different parts of the chapter distinct.
Before submission, run a citation and reference check. Never invent a study to fill an empty matrix row. UniDraft can help organise sources and facts you confirm, but you must check that each source exists and supports its claim before using the resulting document. Explore UniDraft.