PhD research guide

How to Write a Literature Review for a PhD Thesis

A step-by-step method for a PhD literature review. Search strategy, screening, a literature matrix, thematic synthesis and a clear gap statement.

Illustration of an open book, a magnifying glass and a lightbulb representing a literature review

A strong PhD literature review does more than summarise research: it maps what is known, shows where studies agree and disagree, evaluates their strengths and weaknesses, and leads logically to the gap your study fills. The method below takes you from searching to a finished chapter in six steps.

What a literature review should achieve

  • Map existing knowledge: what is known, what is debated and what is missing.
  • Situate your study within a theoretical and methodological landscape.
  • Justify your gap: population, context, variable, method or theory.
  • Build an argument for why your research is needed.
  • Lay the foundation for your conceptual framework and hypotheses.

Step 1: Define keywords, scope and search strategy

Use three layers of keywords:

  • Core concepts: for example work engagement, machine learning adoption, groundwater modelling.
  • Context: India, public sector, women entrepreneurs, Madhya Pradesh.
  • Method terms: structural equation modelling, panel data, thematic analysis.

Combine them with Boolean operators: ("work engagement" OR "employee engagement") AND (nurses OR "health workers") AND India.

Where to search:

  • Scopus and Web of Science, where your university provides access, for systematic searching
  • Google Scholar, broad but noisy
  • Discipline databases: PubMed for life sciences, IEEE Xplore for engineering, ERIC for education, and others
  • Shodhganga for Indian PhD theses (see using Shodhganga for your review)
  • Your library's e-resources, including subscriptions available through national consortia

Keep a search log of each database, the exact search string, filters, date and the number of results. You may need it for your methodology and examiners appreciate it.

Step 2: Screen efficiently

Do not try to read 500 papers in full. Screen in three passes:

  1. Titles: remove clearly irrelevant items.
  2. Abstracts: check scope, method, population and relevance to your questions.
  3. Full text: read closely only what is clearly connected to your variables or model.

Note why you excluded items at the full-text stage; short notes are enough.

Step 3: Build a literature matrix

A spreadsheet with one row per source turns reading into analysis:

Column What to record
Reference Author, year, title, journal, DOI or handle
Aim / questions What the study set out to do
Theory / framework e.g. Theory of Planned Behaviour
Context Country, region, sector, population
Design and sample Method, sample size, sampling
Variables / measures Key constructs and instruments
Findings Main results, in your own words
Limitations Stated and your own assessment
Relevance How it relates to your study

Use a reference manager (Zotero, Mendeley or EndNote) alongside the matrix to store PDFs and generate citations.

Step 4: Synthesise by theme

Sort and group your matrix to find patterns:

  • Themes that correspond to parts of your research question
  • Agreements and contradictions between studies
  • Methodological patterns, such as everyone using cross-sectional surveys
  • Contextual differences between Indian and international findings
  • Changes over time

Plan your chapter around these themes, not around individual studies.

Summary versus synthesis

Summary: "Sharma (2019) found that training increased retention. Gupta (2021) found no effect." Synthesis: "Evidence on training and retention is mixed: studies in private-sector firms report positive effects (Sharma, 2019; Rao, 2020), whereas studies in public organisations find none (Gupta, 2021), suggesting that organisational context may moderate the relationship."

Step 5: Structure the chapter

A common structure:

  1. Introduction: scope of the review, how literature was searched and organised.
  2. Theoretical background: key theories and models.
  3. Thematic sections: one per major theme or variable relationship, each ending with a short synthesis.
  4. Indian context: what Indian studies show and how they differ.
  5. Research gap: an explicit statement of what is missing, supported by citations.
  6. Conceptual framework: a diagram and explanation of the relationships you will study, leading to your hypotheses.

Step 6: Write, cite and check

  • Write in your own words from your matrix notes, not from the papers.
  • Evaluate as you go: comment on sample sizes, methods, contexts and limitations.
  • Cite every idea, not only direct quotes.
  • Use quotations sparingly, only where the exact wording matters.
  • Check similarity early through your library, and fix close paraphrasing. See plagiarism and referencing.

Quality checklist

  • Organised by themes, not by author
  • Includes both foundational and recent work
  • Covers Indian and international research where relevant
  • Critically evaluates, not just describes
  • Ends with a clear gap that your objectives address
  • Every citation is in the reference list, in one consistent style
  • No text copied or closely paraphrased without attribution

For one-to-one help organising and reviewing your chapter, see our literature review guidance.

Frequently asked questions

How many references should a PhD literature review have?

There is no fixed number; it depends on the field and topic. Many PhD theses cite somewhere between 100 and 300 sources overall. Relevance and quality matter more than count.

How recent should the sources be?

Cover the foundational works that defined the field, plus the most recent research, typically the last five to ten years in fast-moving fields. Update your review before submission.

Should I use a systematic review method?

A full systematic review (for example following PRISMA) is a research method in its own right and is required in some fields. For most PhD theses, a structured narrative review with a documented search strategy is sufficient. Ask your supervisor what is expected.

Sources and further reading

WhatsApp