Prompt · Research

Thematic analysis of interview transcripts

Code user or research interview transcripts into themes with supporting quotes, counts across participants and outliers, so you can review and refine the analysis instead of starting from scratch.

The prompt

You are a qualitative researcher. Help me with a first-pass thematic analysis of the interview transcripts below.

Research question: [WHAT YOU WANT TO UNDERSTAND]
Context: [e.g. "8 interviews with nurses about shift handovers"]
Approach: [e.g. "inductive: let themes emerge" or "deductive: use these codes: ..."]

Transcripts (participants labelled P1, P2, ... with personal details removed):
"""
[PASTE TRANSCRIPTS]
"""

Steps:
1. Initial codes: list short codes (2-5 words) with a one-line definition each.
2. Themes: group codes into 4-7 themes. For each theme give:
   - Name and a 2-sentence description.
   - Which participants mention it (e.g. "P1, P3, P4 - 3 of 8").
   - 2-3 verbatim quotes with participant labels. Quote exactly; do not paraphrase inside quotation marks.
3. Divergent views: where participants disagree or a single participant says something important that does not fit.
4. Relationships: how themes might connect (e.g. one appears to cause another), labelled "TENTATIVE".
5. Limitations of this first pass: what a human analyst should check.

Rules:
- Do not report a theme supported by only one participant as a pattern.
- Do not infer feelings or motives that participants did not express.
- Keep my research question in focus; put off-topic material under "Other observations".

Example output

Example (excerpt):
Theme 2: Information lost at night handovers
Description: Participants describe details being dropped when handovers happen in a rush at the end of night shifts.
Mentioned by: P1, P2, P5, P7 (4 of 8)
Quotes:
- P2: "By 7 a.m. you just say 'bed four is fine' and go."
- P5: "The allergy was on the chart but nobody said it out loud."

Divergent view: P6 says the printed handover sheet works well on her ward.
Relationship (TENTATIVE): staffing shortages (Theme 4) may explain the rushed handovers.

When to use it

Use this prompt after a round of user interviews, customer discovery calls or qualitative research interviews, when you have hours of transcripts and need a structured first view. It suits product teams, UX researchers, students and policy analysts. Treat it as a fast first pass that a person then checks and refines, not as the finished analysis. It also helps teams share findings quickly while a person completes the careful coding.

How to customize it

[RESEARCH QUESTION] keeps the coding focused. Without it, themes drift towards whatever participants talked about most.
[CONTEXT] helps the model understand jargon and roles.
[APPROACH] chooses between inductive coding (themes emerge from the data) and deductive coding (you provide a codebook). Supply your codes if you have them.
[PASTE TRANSCRIPTS] should be anonymised. Replace names, employers and other identifying details with labels before pasting, and make sure your consent forms allow this kind of processing.

Why it works

Thematic analysis is a widely used qualitative method: code the data, group codes into themes, and support each theme with evidence. The prompt follows that sequence, so you can review each step instead of receiving an opaque summary.

Participant counts and labelled quotes keep the analysis grounded. Requiring verbatim quotes matters because models may smooth or paraphrase speech; check a sample against the transcript. The rule against treating single mentions as patterns guards against over-generalising, while the “Divergent views” section ensures that important outliers are not lost. Labelling causal links as “TENTATIVE” reflects that interviews rarely prove cause and effect.

Tips and variations

Ask the model to list quotes that contradict each theme as well as those that support it; this makes it easier to judge how strong a theme really is.
For many long transcripts, analyse in batches and then ask the model to merge the codebooks; see our guide on summarizing long documents.
Ask for a codebook table (code, definition, example quote) you can reuse for the next round of interviews.
Have a second person code two transcripts independently and compare results.
If you used an automatic transcription tool, our article on AI meeting note-takers and privacy is worth a read.

#Qualitative Research #Thematic Analysis #User Research #Interviews

More Research prompts

View all

Researchintermediate

Fact-check list for every claim in a draft

You are a fact-checker at a careful publication. Do not judge whether claims are true from memory. Your job is to build a verification checklist. Type of text: [e.g. "blog post", "company report", "speech", "social media thread"] Audience and stakes: [e.g. "public article, legal risk if wrong"] Sources the author used (if known): [LIST, or "unknown"] Steps: 1. Extract every checkable claim: numbers, dates, names and titles, quotes, causal statements ("X led to Y"), comparisons ("the largest", "the first") and descriptions of studies or laws. 2. For each claim give, in a table: # | Claim (exact words) | Type | Risk if wrong (high/medium/low) | What would verify it (primary source to look for) | Notes 3. Mark as high risk: statistics, quotes, health, legal or financial statements, and anything about a named person or company. 4. Flag vague claims that cannot be checked as written ("experts say", "studies show") and suggest how to make them specific. 5. List claims where the wording goes beyond what a typical source would support (e.g. "proves" instead of "suggests"). If you can browse the web: for high-risk claims only, look for a primary source, give the link and quote the relevant sentence. Mark each as "SUPPORTED", "CONTRADICTED" or "NOT FOUND". If you cannot browse, say so and do not assign these labels. Draft: """ [PASTE TEXT] """

Works with: ChatGPT, Claude, Gemini, Perplexity

Researchadvanced

Literature review matrix from your sources

You are a research assistant experienced in systematic reading. Build a literature review matrix from the sources I provide. Work only from the text I paste. My research question: [QUESTION] Field: [e.g. "public health", "education", "HCI"] What I need the review for: [e.g. "thesis chapter", "grant proposal background", "internal report"] Sources (each labelled, with abstract, methods and results sections where possible): """ [P1] [AUTHORS, YEAR, TITLE] [PASTED TEXT] [P2] ... """ Step 1 - Matrix. A table with columns: ID | Authors & year | Research question | Method / design | Sample & setting | Key findings | Limitations stated by the authors | Relevance to my question (high/medium/low) Write "NOT REPORTED" when the pasted text does not include the information. Step 2 - Synthesis: - Themes: where do 2 or more sources agree? Cite IDs. - Conflicts: where do findings disagree, and could differences in method or sample explain it? - Gaps: what my question needs that no source addresses. - Quality notes: which findings rest on small samples, short time frames or self-reported data. Rules: do not add studies, authors or statistics I did not provide. Quote exact numbers only as they appear in the text. If a claim in the synthesis is your interpretation, label it "INTERPRETATION".

Works with: ChatGPT, Claude, Gemini, Any LLM

Researchintermediate

Research brief with source checking

You are a careful research analyst. Prepare a research brief on this question: [RESEARCH QUESTION] Purpose of the brief: [WHAT DECISION OR PIECE OF WORK IT SUPPORTS] Time frame that matters: [e.g. "studies since 2022"] Sources to use: [PASTE SOURCES, OR WRITE "search the web" IF YOUR ASSISTANT CAN BROWSE] Work in this order: 1. Restate the question and list 3-5 sub-questions that together answer it. 2. For each sub-question, give the key findings. After every factual claim, cite the source in brackets. 3. Build a source table: Source | Type (peer-reviewed, official, news, opinion, vendor) | Date | Reliability (high/medium/low) | Why. 4. List points where sources disagree and explain the likely reason. 5. List what you could not find or verify. Do not fill gaps with assumptions. 6. End with a 5-sentence answer to the original question and a confidence level (high/medium/low). Rules: - Use only the sources provided or found. If you cannot access a source, say so. - Never invent titles, authors, URLs or statistics. - Mark any claim that relies on a single source as "single source".

Works with: ChatGPT, Claude, Gemini, Perplexity