Research Plan Soundness Review
Check that a study can answer its question.
The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.
Review my research plan before I start. Plan text: [RESEARCH_PLAN]. The decision the research should inform: [DECISION]. Participants available: [PARTICIPANTS]. Time and budget: [TIME_AND_BUDGET]. Check that the research question is specific and answerable, the method fits the question, the sample fits the method, tasks or questions are neutral, the analysis approach is stated in advance, ethics and consent are covered, and the outputs will arrive before the decision date. Rate each from 1 to 5 with a reason that quotes the plan. Return a scorecard, the three biggest weaknesses with fixes, a better version of the research question, and a one-page timeline. Do not invent details I did not give; list the questions you need answered. Add a worked scorecard on an invented plan: study of why new users abandon setup, ten interviews, three weeks. Research question: specific enough, 4 of 5. Method fits the question, interviews suit why questions, 4. Sample: ten recent drop-offs recruited from the user list, risk of recruiting only people who reply, 3. Neutral questions, 3, because two questions in the guide ask do you find setup too long. Analysis stated in advance, 2, only the sentence we will code the notes. Ethics and consent: 4. Outputs before the decision: the decision is in two weeks and the analysis ends in three, 1. Total 21 out of 35, which is 60 percent. Biggest weaknesses: the timeline, the analysis plan and two leading questions. Fixes: start the analysis on day four, agree codes in advance, rewrite the questions neutrally, add a way to reach people who do not reply. Propose a better question: What happened between creating an account and deciding to stop, in the words of the people who stopped? Add a timeline: week one recruit and pilot with two people, week two sessions and daily notes, week three synthesis and readout, with the decision date in the diagram. Include a short list of checks to run before every study: what decision will change, what would we do if the answer is X, and is there existing data that answers it?
Fill in before use
Optional: type your details and the prompt above updates. Nothing is saved or sent.
Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.
See an example result (fictional data)
Invented plan: ten interviews on why new users abandon setup, three weeks. Scores out of 5: research question 4; method fit 4; sample 3 (recruited from people who reply); neutral questions 3; analysis plan 2; ethics 4; outputs before the decision 1. Total = 4 + 4 + 3 + 3 + 2 + 4 + 1 = 21 out of 35 = 60 percent. Biggest weaknesses: timeline (decision in two weeks, analysis ends in three), analysis plan, two leading questions. Fixes: start analysis on day four; agree codes in advance; rewrite questions; add a way to reach non-responders.
Illustrative only, made-up data. Your results will differ and must be reviewed.