Research & Analysis / Source Summaries
Extract just the methods section from any study in a standardized, comparable format.
Why This Prompt Exists
A study’s conclusions are only as good as its methods — but methods sections are dense, jargon-filled, and easy to skim past.
You get:
- citing findings without knowing if the methods support them
- missing fatal flaws that invalidate conclusions
- comparing studies with different methods as if they were equivalent
- spending hours parsing methods sections to find basic information
- trusting peer review instead of verifying yourself
But methods can be standardized:
- design: experiment, quasi-experiment, correlational, qualitative
- sample: size, source, representativeness, attrition
- measures: what, how, validity, reliability
- analysis: statistical tests, controls, assumptions
- limitations: what the authors admit (and what they miss)
Without methods extraction, you trust blindly.
This prompt creates a standardized methods snapshot for any study.
The Prompt
Assume the role of a methods reviewer who extracts study methodology. Your task is to create a standardized methods snapshot from a study. Generate: 1. DESIGN - Research design: [Experimental / Quasi-experimental / Correlational / Qualitative / Mixed-methods] - Temporal: [Cross-sectional / Longitudinal / Retrospective] - Randomization? [Yes / No / Not applicable] - Control group? [Yes / No / Not applicable] 2. SAMPLE - Sample size: [N] - Population: [description] - Recruitment method: [Random / Convenience / Purposive / Snowball] - Attrition rate: [% if reported] - Power analysis? [Yes / No — if yes, effect size assumed] 3. MEASURES - Primary outcome: [what, how measured, reliability if reported] - Key predictors: [what, how measured] - Control variables: [list] 4. PROCEDURE - Setting: [Lab / Field / Online / Mixed] - Duration: [How long participants were involved] - Key steps: [brief timeline] 5. ANALYSIS - Primary statistical tests: [t-test / ANOVA / Regression / etc.] - Software: [SPSS / R / Stata / etc.] - Assumptions checked? [Yes / No / Not reported] - Multiple comparison correction? [Yes / No / Not applicable] 6. METHODS-RELATED LIMITATIONS - What the authors acknowledge - What they missed (based on methods summary) 7. METHODS TRUST SCORE - 5 (Exemplary) to 1 (Fatally flawed) INPUTS: Source content (methods section required): [PASTE OR UPLOAD] Study type: [ACADEMIC PAPER / INDUSTRY REPORT / GOVERNMENT STUDY] Your field expertise: [HIGH / MEDIUM / LOW] RULES: - Only report what the source explicitly states (don't assume) - Flag "not reported" explicitly — missing information is information - Distinguish between what authors claim (e.g., "random") and what they actually did - Note when methods are described vaguely (e.g., "standard procedures") - If sample size is small, flag as potential power concern
How To Use It
- Run this before citing any study — methods determine whether you can trust the findings.
- Use the standardized format to compare methods across multiple studies quickly.
- Pay closest attention to sample (is it representative?) and randomization (can they claim causation?).
- Flag “not reported” items — these are red flags for low-quality research.
- A low methods trust score doesn’t mean ignore the study — it means cite with heavy caveats.
Example Input
Source content (methods section):
“We recruited 150 undergraduate psychology students (mean age 19.2, 65% female) from participant pool. Participants completed a 10-minute online survey measuring grit (12-item Grit-S scale, α = .85) and academic performance (self-reported GPA). No power analysis reported. Data analyzed using Pearson correlations in SPSS.”
Study type:
Academic paper
Your field expertise:
Medium
Why It Works
Most people skip to the conclusions and trust that the methods were sound — a dangerous assumption.
This framework improves outcomes by forcing:
- design specification (what kind of evidence is this?)
- sample evaluation (who was studied — and who wasn’t)
- measurement review (how was the key outcome measured?)
- analysis audit (were the right tests used?)
- limitation identification (what the authors admit — and what they don’t)
- trust score (explicit evaluation, not implicit)
Great methodology snapshots don’t just describe methods — they tell you whether to trust the findings.
Build Better AI Systems
Subscribe for advanced prompt engineering, AI coding tools, debugging frameworks, and practical strategies for developers and engineers.