The one-line version: a literature review is not a summary of what you read. It is an argument that a specific gap exists and that your question is the right one to close it. Everything below is in service of that sentence.
Most rejected chapters fail in the same way. The student read widely, wrote a paragraph about each source in the order they were read, and produced an annotated bibliography with the headings removed. The supervisor writes “where is the synthesis?” and the student, reasonably, does not know what is being asked for. The nine steps below end at an answer to that.
Step 1: Write the scoping question before you read anything else
Turn the topic into a single answerable question naming three things: a population, a phenomenon, and a context. “Teacher burnout” is a topic. “How do workload-reduction interventions affect burnout among secondary-school teachers in Canadian public boards?” is a scoping question, and it tells you which studies are in and which are out.
Write it down and date it. You will revise it three times, and the revisions are evidence of thinking, not of failure.
Step 2: Turn the question into search blocks
- Break the question into two to four concepts.
- Under each concept, list every word the literature uses for it — including the ones you dislike. Burnout, emotional exhaustion, occupational stress, compassion fatigue.
- Combine within a block with OR, and across blocks with AND.
- Add the truncation your database uses so that teach* catches teacher, teachers and teaching.
The synonym list is the step that separates a thorough review from a thin one, and it is almost never done properly on the first attempt. A subject librarian will improve it in fifteen minutes; book the appointment.
Step 3: Record the search as you run it
For every database: the database name, the exact string, the date you ran it, any limits applied, and the number of results. Keep it in one table.
This is not bureaucracy. Your committee may ask how you know you did not miss something, and a reproducible search is the only answer that holds. It also means that when you re-run the search eight months later — and you will — you can report what changed instead of starting over.
Step 4: Search the Canadian collections most students skip
Your discipline’s databases are obvious. These three are not, and they are where Canadian reviews gain their edge.
Theses Canada
Prior Canadian theses are the most under-used source in a Canadian literature review, and they are the closest thing you will find to a worked example of your own genre. The Theses Canada portal indexes more than 425,000 Canadian theses and dissertations from 1965 and provides free access to over 200,000 full-text electronic Canadian theses and dissertations, drawn from around 70 institutions. Since April 2014 only digital theses are accepted, and electronic theses are harvested monthly from Canadian universities.
What to take from them: the literature-review chapters themselves. A thesis in your field completed three years ago has already done a version of your search, and its reference list is a free head start — provided you read it as a lead rather than as a source to cite second-hand.
Érudit
Érudit is the main platform for francophone Canadian scholarship, and it is not a journals-only database: its publication types are journals, theses and dissertations, books and proceedings, and research reports, in both official languages. If your topic touches education, health services, law, social policy or anything else where Québec’s system differs from the rest of Canada, the francophone literature is not optional context — it is a distinct evidence base, and omitting it is a finding your examiner may make on your behalf.
Your provincial consortium’s discovery layer
Canadian university libraries buy collectively, and the consortium’s search tool reaches far past your own campus holdings. In Ontario, Omni is the shared library system built through the Ontario Council of University Libraries’ Collaborative Futures initiative, powered by 19 Ontario universities, giving access to a province-wide collection with materials delivered to a pick-up location of your choice. Other provinces run their own equivalents. Find yours before you pay for a single article.
Step 5: Screen in two passes, and count what you exclude
Write your inclusion and exclusion criteria down first — study design, population, date range, language, setting. Then:
- Pass one: titles and abstracts. Fast, generous, and only against the written criteria.
- Pass two: full texts of everything that survived, against the same criteria.
Record how many you excluded at each pass and the main reasons. Even a narrative review benefits from these numbers, because they convert “I read a lot” into “I screened 412 records and included 38.”
Step 6: Extract into a matrix, not into prose
Build one table with a row per study and columns for: citation, question, design, sample, setting, main finding, limitation, and your one-line note on relevance. Fill it as you read.
The matrix is what makes synthesis possible, because themes become visible as patterns down a column rather than as memories across forty PDFs. It is also the artefact that survives if you change supervisors, change questions, or come back to the project after a term away.
Step 7: Group by claim, not by author
This is the move. Stop asking “what did Chen say?” and start asking “who says X, who says not-X, and why do they disagree?”
Read down the findings column and cluster the studies into three to six themes. A theme is a claim, a tension, or a methodological divide — not a topic. Compare these two openings:
Summary (weak): “Chen (2021) studied workload interventions in three boards and found reduced burnout. Okonkwo (2022) examined a similar programme and found no effect.”
Synthesis (strong): “Studies reporting reduced burnout after workload interventions have measured outcomes within one term (Chen, 2021), while those following participants across a full academic year report no lasting effect (Okonkwo, 2022). The disagreement is therefore about measurement window rather than about the intervention.”
The second version is not longer. It contains a claim the reader could not have made themselves, and that is the whole difference.
Step 8: Write the chapter in a fixed shape
A structure that works across disciplines:
- Opening: what the chapter will establish, in three sentences.
- Scope and method: the question, the databases, the criteria, the counts.
- Theme sections: one per theme, each opening with the claim and closing with what remains unresolved.
- Theoretical framing: the concepts your study will use, taken from the literature you have just reviewed rather than introduced from nowhere.
- Gap statement: one paragraph naming precisely what is not known.
- Bridge: one paragraph in which your research question answers that gap.
Write the gap statement and the bridge first, in draft, then write the themes towards them. A chapter written forwards tends to arrive somewhere; a chapter written backwards from the gap arrives where it must.
Step 9: Keep the apparatus consistent while the chapter grows
A literature review chapter carries most of a thesis’s citations, and it is the chapter most likely to be rewritten three times. Two mechanical things save days:
- Keep every reference in one managed library, so a corrected entry propagates instead of being fixed in four places. Which manager depends on your library’s licences — see our comparison of Zotero, Mendeley and EndNote for a Canadian thesis.
- Format as you go, using the rules rather than a generator’s guess. Our complete APA 7 guide for Canadian theses covers the four elements and the Canadian source types the manual does not include.
Draft your literature review in Tesify — the chapter structure and referencing stay consistent through every rewrite while every argument remains 100% written by you.
Two warnings specific to Canadian graduate work
If your study involves human participants, the literature review is not the gate. Ethics clearance is, and it has to be in place before data collection begins — see REB approval and TCPS 2 in Canada. Reviewing literature is not research on participants, so the review itself can proceed while the application is pending.
If you plan to use AI anywhere in the screening or drafting, get the approval before you start rather than describing it afterwards. What the major Canadian graduate schools require is set out in our guide to AI use in a Canadian thesis.
Frequently asked questions
How long should a thesis literature review be?
Long enough to establish the gap and no longer. Length is set by your programme, not by a general rule, and a tight 25-page chapter beats a diffuse 60-page one at every defence.
How many sources do I need?
There is no number. What you need is coverage of the themes, including the studies that disagree with your position. Report your screening counts rather than a target.
What is the difference between a literature review and an annotated bibliography?
An annotated bibliography describes sources one at a time. A literature review groups them by claim and argues that something is missing.
Should I search French-language sources?
In Canada, usually yes. Érudit carries francophone journals, theses, books and research reports, and for topics where Québec’s system differs the francophone literature is a separate evidence base.
Are other people’s theses citable?
Yes, and Theses Canada indexes more than 425,000 Canadian theses with over 200,000 available in full text. Use them for their reference lists and their method chapters as well as for their findings.
Do I have to do a systematic review?
Only if your programme requires one. But borrowing its habits — written criteria, recorded searches, counted exclusions — improves a narrative review at no cost.
How do I show synthesis rather than summary?
Open each paragraph with a claim about the literature and cite several studies in support or in tension, instead of opening with an author’s name.
How do I find grey literature?
Government and agency reports, plus research reports on platforms such as Érudit. Record where you searched, because grey literature searches are the least reproducible part of any review.
When should I stop searching?
When new searches return sources you have already screened. Note the date you reached that point and re-run the search before submission.
Can I reuse my coursework literature review in my thesis?
Ask first. Reusing your own submitted work without permission is a recognised offence at many Canadian universities.
Where does the theoretical framework go?
Usually at the end of the review, built from the literature you have just discussed, so it reads as a conclusion rather than an import.
What do I do when two respected studies contradict each other?
Explain the contradiction rather than choosing a side. Differences in sample, setting, measurement window or outcome definition are the material a good review is made of.