Your science literature review's basic different from literature reviews in humanities or social sciences. You're not summarising what researchers have written and discussed. You're conducting rigorous appraisal of empirical evidence. You're identifying gaps. You're evaluating methodology. You're synthesising findings to support your research.

Many science students approach literature reviews as narrative summaries. They read studies and write about them chronologically or by author. That produces work reading like an annotated bibliography, not critical appraisal. Examiners expect different. They expect you to demonstrate field mastery. That means understanding not just what studies found, but whether they found it using sound methods, and how different findings fit together into a coherent evidence picture.

Understanding Your Field's Evidence Hierarchy

Science research operates within a formal evidence hierarchy. This concept transforms how you evaluate and present sources. Understanding it's absolutely key.

At the top sit systematic reviews and meta-analyses. These synthesise multiple high-quality studies using standardised methods. They represent the strongest evidence because they pool findings systematically. If a systematic review exists in your field, it's your key starting point. You're not reading it to discover what studies exist. You're reading it to understand what careful synthesis of evidence reveals. You're identifying whether gaps remain that your research addresses.

Below systematic reviews sit randomised controlled trials. These eliminate selection bias and confounding variables through their design. But methodology matters profoundly. A large multicentre RCT with long follow-up, published in a top-tier journal with adequate blinding, constitutes stronger evidence than a small single-centre trial from a specialist journal. You need to understand study quality.

Observational studies come next, cohort studies, case-control studies, followed by qualitative research, case reports, and opinion pieces. This hierarchy matters because it shapes how you weigh findings. If one high-quality RCT contradicts findings from fifty observational studies, the RCT typically carries more epistemological weight. But you need to explain why. Simply citing hierarchy isn't good enough. You're explaining the reasoning.

Secondary sources play an important role in any dissertation, providing the theoretical and empirical context within which your own research is situated and helping to establish the significance of your research question. However, it is important not to rely too heavily on secondary sources at the expense of engaging directly with the primary sources, original texts, and raw data that form the foundation of your academic field. A dissertation that draws on a variety of high-quality sources and demonstrates the ability to synthesise those sources into a coherent argument will always be more favourably received than one that relies on a small number of introductory texts. As you gather sources for your dissertation, keep careful records of the bibliographic details of each source, since reconstructing this information at the end of the writing process is time-consuming and can introduce errors into your reference list.

Selecting Appropriate Databases

Your database choice shapes your entire literature review. Using only Google Scholar or PubMed alone produces incomplete results. Science students need multiple complementary databases.

PubMed remains important for biomedical literature. It covers medicine, nursing, life sciences particularly well. The advanced search features let you combine search terms systematically. You can filter by article type, date range, and other variables.

Web of Science and Scopus provide broader coverage across pure and applied sciences, chemistry, physics, engineering, earth sciences. Both allow sophisticated filtering by article type, date range, and journal quality indicators. Both track citation patterns, showing how frequently papers get cited.

Google Scholar shouldn't be your primary database, though it's useful for supplementary searching. Its inclusion of grey literature and simpler interface appeal, but it includes preprints, theses, and unrefereed sources requiring careful quality assessment.

Key Considerations and Best Practices

For specific fields, specialised databases matter. CINAHL dominates nursing literature. The Cochrane Library contains systematic reviews and controlled trials in healthcare. CAB Abstracts covers agriculture and veterinary science. BIOSIS Previews indexes biological research thoroughly. Your university library provides access to most of these. Your supervisor or librarian can recommend which databases matter for your specific discipline.

Using PICO to Structure Your Search and Appraisal

PICO (Population, Intervention, Comparison, Outcome) provides structured approach to formulating your literature search and appraising what you find.

Population refers to the specific group your research concerns. Age range, disease status, geographical location, other characteristics, these matter. A study of hypertension management in adults aged 18 to 65 differs basic from one examining the same intervention in children or elderly populations. Define your population precisely.

Intervention describes what you're investigating. In experimental research, this is the treatment or exposure. In observational studies, it might be an exposure, risk factor, or characteristic you're examining. Be specific. "Cognitive behavioural therapy delivered face-to-face by qualified therapists over eight weeks" differs substantially from "some form of psychological support."

Regular contact with your supervisor throughout the dissertation period helps you stay on track, receive timely feedback, and avoid the isolation that can make a long research project feel more difficult than it needs to be.

Comparison identifies what you're comparing against. Standard treatment, placebo, no treatment, alternative intervention. Studies without appropriate comparisons produce ambiguous results.

Outcomes specify what you're measuring. Many studies measure multiple outcomes. Your literature review identifies which outcomes matter most to your question and which outcomes studies actually measured. A study demonstrating intervention improves biomarker levels but not patient-reported symptoms tells a different story than one showing the reverse.

Using PICO, you construct search strings combining terms from each element. This produces focused results than simple keyword searching and ensures you're systematically evaluating relevant literature rather than cherry-picking studies supporting your hypothesis.

The process of receiving and responding to feedback from your supervisor is one of the most valuable parts of the dissertation journey, yet many students find it difficult to translate written comments into concrete improvements in their work. When you receive feedback, try to approach it as an opportunity to develop your academic skills rather than as a judgement of your intelligence or your worth as a student, since supervisors give feedback because they want you to succeed. If you receive a comment that you do not understand or disagree with, it is entirely appropriate to ask your supervisor to clarify their feedback or to discuss your response with them in a meeting or by email. Keeping a record of the feedback you receive throughout the dissertation process and revisiting it regularly will help you to identify patterns in the areas where you most need to improve and to track your progress over time.

The gap between what you intended to research and what you actually discovered is often where the most interesting parts of your discussion chapter can be found, so do not shy away from examining unexpected results.

Critical Appraisal Rather Than Just Reporting

The distinction between describing what studies found and appraising whether findings are trustworthy separates strong reviews from weak ones.

Expert Guidance for Academic Success

Weak reviews report conclusions: "Smith et al. found that intervention X improved outcome Y." Strong reviews appraise: "Smith et al.'s randomised controlled trial (N=150, 12-week follow-up, adequate randomisation and blinding) found that intervention X improved outcome Y (p<0.05, effect size d=0.6). However, high attrition (28 per cent) and lack of long-term follow-up limit confidence in sustained effects."

Critical appraisal requires understanding study design. What methodology did researchers use? Did design match their research question? A qualitative study examining patients' experiences of treatment represents appropriate methodology. A randomised trial would be unnecessary and arguably unethical. A labouratory experiment testing chemical toxicity represents appropriate methodology. A survey asking people about their toxicity experiences wouldn't.

Assessment tools help structure your appraisal. CASP (Critical Appraisal Skills Programme) tools provide checklists for different study types. Cochrane's risk of bias tool guides RCT assessment. The GRADE approach evaluates overall evidence quality for specific outcomes. Using these tools ensures your appraisal's rigorous rather than subjective.

The relationship between theory and practice is one of the most productive tensions in academic research, and dissertations that engage seriously with both theoretical and empirical dimensions of their topic tend to produce the most interesting and well-rounded analyses. Purely descriptive dissertations that report findings without engaging with theoretical frameworks often lack the analytical depth required for the higher grade bands, since they do not demonstrate the capacity for independent critical thought that distinguishes undergraduate and postgraduate research. Dissertations that are strong on theoretical sophistication but weak on empirical grounding can feel abstract and disconnected from the real-world problems that motivated the research in the first place. The most successful dissertations find a productive balance between theoretical rigour and empirical substance, using theory to illuminate the data and using the data to test, refine, or challenge the theoretical assumptions that frame the study.

Synthesis Rather Than Description

The discussion chapter of your dissertation is where you bring everything together, showing how your findings relate to the literature you reviewed and what they mean for the broader questions in your field.

Organising your literature review matters profoundly. Many students organise chronologically (what researchers found in 2015, then 2016, then 2017) or alphabetically by author. This produces description, not synthesis.

Thematic organisation groups studies around key concepts or findings. You might organise around different hypotheses explaining your phenomenon, different populations studied, different methodological approaches, or different outcome measures. This structure forces you to compare and contrast rather than simply summarise.

Synthesis involves identifying patterns. Do most studies show consistent findings, or is evidence mixed? When evidence contradicts, why might this be? Did studies examine different populations? Use different methods? Measure different outcomes? Have different quality levels? Answering these questions requires careful reading and critical thinking about study relationships.

A strong synthesis paragraph might read: "while three high-quality randomised trials demonstrated considerable walking distance improvements with intervention X (Smith et al., 2020; Jones et al., 2021; Brown et al., 2022), five observational studies found modest benefits only in subgroups with mild disease severity (Anderson et al., 2019; Lee et al., 2020; others). This discrepancy likely reflects strict inclusion criteria in RCTs, which typically enrolled highly motivated participants in specialist centres. Observational studies conducted in routine practise settings probably better reflect real-world effectiveness."

Academic integrity is a principle of higher education that your university will take seriously, regardless of whether any breach was intentional or the result of careless academic practice. Plagiarism is not limited to copying passages from other sources without attribution; it also includes paraphrasing someone else's ideas without proper citation, submitting work that has been completed by another person, or submitting work you have previously submitted for a different module. Developing good habits of academic integrity from the beginning of your studies will protect you from the anxiety of submitting work when you are unsure whether your referencing and attribution practices meet the required standard. If you are ever in doubt about whether a particular practice constitutes plagiarism or another form of academic misconduct, the most sensible course of action is to consult your university's academic integrity guidelines or speak to your module tutor.

Identifying Contradictions and Gaps

Strong literature reviews don't ignore contradictory findings. They explain them.

Practical Steps You Should Follow

When studies reach different conclusions, investigate why. Look at population differences. A medication helping younger adults might harm elderly people. Look at study quality. Does the contradicting study have important limitations? Look at outcome differences. Studies might agree on one outcome but disagree on another. Your task involves explaining these patterns, not pretending contradictions don't exist.

The structure of your dissertation should make it easy for the reader to follow your argument without having to work too hard to understand how different sections relate to each other and contribute to the whole.

Identifying gaps justifies your research. What questions remain unanswered? What populations remain understudied? What methodologies remain absent? A literature review concluding "all evidence is clear and no further research is needed" fails to justify why your dissertation should exist. A strong conclusion identifies specific, justified gaps that your research will address.

Frequently Asked Questions

Q: How many sources should a science literature review include? A: There's no magic number. Quality exceeds quantity. For an undergraduate science dissertation, 40 to 60 high-quality sources typically suffice. For a postgraduate dissertation, 80 to 150 sources might be appropriate depending on your field. The criterion is whether your review covers the field thoroughly. If you keep finding new relevant studies, you might be missing areas. If you're seeing the same studies cited repeatedly, you've likely reached saturation.

Q: Should I include the results section of studies or just the abstract? A: Read the full paper. Abstracts often overstate findings and omit important methodological details or limitations. The results section shows precisely what was measured and how effect sizes matter. A result that's "statistically considerable" (p<0.05) might be clinically trivial if effect size is negligible. Full-text reading's non-negotiable for critical appraisal.

Q: How do I handle grey literature like conference presentations and theses? A: Use grey literature selectively. Conference abstracts lack peer review and often report preliminary findings that might not replicate. Search whether authors later published full results in peer-reviewed journals. Theses from prestigious institutions can be valuable, particularly for methodology descriptions, but they lack external peer review. Include grey literature only when it fills genuine gaps and when you've assessed quality carefully.

---

The quality of your dissertation conclusion will often determine the final impression your work makes on your marker, as it is the last thing they read before forming their overall assessment of your academic achievement. A strong conclusion does more than simply repeat the main points of your dissertation; it synthesises your findings in a way that demonstrates the overall contribution your research has made to knowledge in your field. You should also take the opportunity in your conclusion to reflect on what you would do differently if you were conducting the research again, as this kind of reflexivity demonstrates intellectual maturity and an honest assessment of your work. Ending with a clear statement of the implications of your research and the questions it leaves open for future investigation gives your dissertation a sense of intellectual momentum and leaves your reader with a positive final impression.