How to Plan the First Search for a New Thesis Topic

A new thesis topic rarely arrives fully formed. More often you have a rough direction, a supervisor's comment, or a folder of half-read papers, and you need to find out whether the topic is real, crowded, or still open. The first search is not a single query typed into a box. It is a short, deliberate project: define what counts as relevant, search several sources, record what you did, and stop when the evidence tells you the topic is workable.

This guide gives you a complete sequence for that first search, with checkpoints, a worked example, a reusable checklist, and clear rules for when to stop or escalate to your supervisor. It is written for students and early-career researchers finishing a course paper, thesis chapter, or first manuscript and preparing the next project.

Most people start searching too early, with a topic that is still a sentence fragment. Before you open a database, decide which of these states you are in:

• Vague interest. You have a field (for example, "urban heat and public health") but no question. You need exploratory reading first.
• Supervisor-shaped direction. You have a comment like "look at how cities measure heat exposure" but no boundaries. You need to turn that into eligibility criteria.
• Candidate topic with a hunch. You believe there is a gap, but you have not tested it. You need a structured search to confirm or reject the hunch.
• Leftover project. You are cleaning up after a finished paper or a winter break and want to convert loose notes into a new project. You need to consolidate what you already have before searching again.

Your starting state determines how much time you spend on exploration versus verification. A vague interest needs broad reading; a candidate topic needs a tight, documented search.

The Cochrane Handbook's guidance on searching for studies begins with eligibility criteria: decide in advance what kinds of studies, populations, interventions, and outcomes count as relevant. This applies to any first search, not only systematic reviews. Writing criteria forces you to say what you are looking for and what you will ignore.

A practical format is a short table with four rows:

Add two boundary rules: a date range and a language or publication-type limit, if any. Write these down. They become the backbone of every query you run and every screening decision you make.

Checkpoint 1: You can state your topic in one sentence and list your four criteria without looking at your notes. If not, keep refining before searching.

Cochrane recommends using multiple sources rather than relying on one database. For a first thesis search, a workable set is usually two or three subject databases plus one broad scholarly search engine. Google Scholar's help pages describe searching by author and title, sorting and limiting by date, using Cited by and Related articles, exporting citations, and setting email alerts. PubMed's user guide describes field tags, phrase searching, filters, Advanced Search, saved searches, Similar Articles, citation lookup, and a Search details view that shows how terms were translated and combined.

Those features suggest a query plan:

1. Build a term list. For each criterion, list free-text synonyms and, where the database supports them, subject headings. Cochrane recommends combining free-text terms with subject headings.
2. Test one database first. Run your query and open the Search details or equivalent view to see how the system interpreted your terms.
3. Adjust and rerun. Add or remove synonyms based on what the translation reveals.
4. Repeat in your other sources. Keep the same criteria but adapt syntax to each database.

If you want a scholarly search interface that keeps this step in one place, you can run your first queries through WisPaper search. For a public explanation of what scholarly search involves, the WisPaper Scholar Search page is a useful entry point.

Checkpoint 2: You have a written query for each source, a note of the date you ran it, and a record of how many records each returned.

Cochrane stresses documenting the search process, deduplicating records, and screening in stages. Documentation is not bureaucracy; it is what lets you explain your method in a thesis chapter and repeat the search later. Keep a simple search log:

Cochrane also recommends rerunning searches before publication. For a first search, note in your log that the search will be refreshed later, so you build the habit early.

Cochrane recommends screening in stages and linking multiple reports of one study. In practice, that means:

1. Deduplicate records across sources.
2. Title screen against your criteria, keeping anything uncertain.
3. Abstract screen the survivors.
4. Full-text screen the remaining candidates.
5. Group reports that describe the same study so you do not count one project twice.

At the title and abstract stages, keep a short reason for each exclusion. This makes your later write-up straightforward and helps your supervisor see your reasoning.

As your candidate list grows, save the papers you intend to keep. My Library lets you save or upload papers and organize them in one place, which keeps your screening decisions attached to the actual records.

Checkpoint 3: You have a deduplicated list, a screening log with reasons, and a set of saved papers you can return to.

Screening tells you what exists. Reading tells you whether your topic is still worth pursuing. Use the tools each source offers: Google Scholar's Cited by and Related articles help you move forward and sideways through a literature; PubMed's Similar Articles does something similar within its index.

As you read, answer three questions in writing:

• What is already well established? Note the findings that appear repeatedly.
• Where do studies disagree or leave a gap? Record specific unresolved questions, not general impressions.
• What would your project add? State it in one sentence.

If you have uploaded your key papers to My Library, you can ask questions grounded in those papers using Library QA. This is useful when you need to compare how several saved papers define a concept or measure an outcome before you commit to a framing.

Checkpoint 4: You can write a short paragraph naming the established findings, the open question, and your proposed contribution.

Suppose your supervisor says, "Look at how cities measure heat exposure," and you have a winter break to turn this into a project.

Step 1 — Starting state. You have a supervisor-shaped direction. You do not yet have a question.

Step 2 — Criteria. Population: urban residents. Exposure: measured heat exposure. Outcome: health outcomes. Study type: observational studies with measured exposure. Date range: the last two decades.

Step 3 — Query plan. Free-text terms for heat, exposure, measurement, urban, and health, combined with subject headings where available. Run in two subject databases and one broad engine.

Step 4 — Document. Log each query, date, source, and count. Note how each database translated your terms.

Step 5 — Screen. Deduplicate, then screen titles, abstracts, and full texts in stages. Group reports from the same study.

Step 6 — Read. Use Cited by and Related articles to find the debates. Write down the established findings and the open question.

Step 7 — Test the topic. You discover that measurement approaches vary widely across studies. Your contribution becomes a comparison of measurement choices and their consequences for reported health associations.

Step 8 — Save and organize. Keep your screened papers in My Library so the evidence for your framing stays together.

Step 9 — Check citations. Before you draft, verify that the references you plan to cite are real. TrueCite checks BibTeX against real academic databases to flag hallucinated references, which is useful when your reading notes include citations gathered quickly.

Step 10 — Set up the project. Create a workspace for the new topic in Research Projects so your search log, saved papers, and notes live in one place.

The search is ready when you can answer all of these without hedging:

• Your topic is stated in one sentence, with eligibility criteria attached.
• You have run the same criteria across at least two sources and logged each query.
• You have deduplicated and screened in stages, with reasons for exclusions.
• You can name the established findings and the specific open question.
• You can state your proposed contribution in one sentence.
• Your key papers are saved and your citations have been checked.

If any item is missing, the search is not finished. If all are present, you have enough to write a short topic proposal or a methods paragraph for your thesis chapter.

Set rules in advance so you do not search indefinitely or stop too early.

Stop when:

• New queries return mostly records you have already seen.
• Your criteria have not changed after two rounds of screening.
• You can write the readiness checklist above without gaps.

Escalate to your supervisor when:

• Two sources give conflicting pictures of whether the gap exists.
• Your criteria keep changing because the field uses inconsistent definitions.
• Your proposed contribution overlaps with a study you found late in screening.
• You cannot access full texts needed to complete screening.

Pause and re-scope when:

• The topic turns out to be saturated and your contribution is unclear.
• The data or methods you would need are outside your available resources.

If you need help generating a sharper framing from your topic, background, references, and constraints, Idea Discovery is built for that step. For a longer research task such as a literature review, Scholar Agent fits the workflow.

• Starting state identified (vague, supervisor-shaped, candidate, leftover)
• Topic written in one sentence
• Four eligibility criteria written down
• Date range and limits set
• Two or more sources chosen
• Term list built with synonyms and subject headings
• Query tested and translation checked
• Search log started with date, source, query, count
• Records deduplicated
• Screening done in stages with exclusion reasons
• Reports of the same study grouped
• Key papers saved in one library
• Established findings, open question, and contribution written
• Citations checked
• Project workspace created
• Readiness checklist complete

Pick the smallest unfinished item on the checklist and do it today. If you have only a vague interest, write your four eligibility criteria. If you have criteria, run one query and open the search details to see how your terms were translated. If you have screened records, write the paragraph naming the established findings and the open question. Then create the project workspace and log your first search so the next session starts where this one ended.