← Back to Blog
Methodology
How to Write a Research Methodology for Q1 Journals
Key insight: The methodology section is where most peer review rejections happen — not because the research is bad, but because the reporting is incomplete. Q1 journals expect specific validity evidence, reliability coefficients, sample size justification, and statistical power — all explicitly reported, not assumed.
A strong methodology section does three things: it tells the reader exactly what you did, it justifies why you did it that way, and it provides enough detail for replication. Most researchers underreport — assuming reviewers will accept standard practices without documentation. This assumption causes major revision requests.
The 7 Components of a Complete Methodology Section
- Research design — quantitative/qualitative/mixed; cross-sectional/longitudinal; explanatory/exploratory
- Sample and sampling strategy — population, sampling method, sample size with justification
- Data collection instrument — survey, interview guide, archival data; source of each scale
- Operationalization of constructs — how each variable is measured, number of items, scale range
- Validity evidence — content validity, construct validity (convergent and discriminant)
- Reliability — Cronbach's alpha, composite reliability (CR)
- Data analysis method — software, version, analytical approach with citation
Minimum Sample Size by Method
| Method | Minimum Sample | Recommended | Rule of Thumb |
|---|---|---|---|
| Multiple regression | 50 | 100+ | 10–20 per predictor variable |
| PLS-SEM | 100 | 200+ | 10× largest construct indicator count |
| CB-SEM (AMOS/lavaan) | 200 | 400+ | 10–20 per free parameter |
| Exploratory Factor Analysis | 100 | 300+ | 5–10 per item |
| Qualitative (interviews) | 12–15 | 20–30 | Until theoretical saturation |
| Case study | 1–4 cases | 4–6 cases | Theoretical sampling |
Validity Reporting Requirements
Convergent validity:
- Average Variance Extracted (AVE) ≥ 0.50 for all constructs
- Outer loadings ≥ 0.70 (PLS-SEM) or standardized loadings ≥ 0.60 (CB-SEM)
- Composite Reliability (CR) ≥ 0.70
Discriminant validity:
- HTMT ratio < 0.85 (strict) or < 0.90 (lenient) — PLS-SEM
- Fornell-Larcker criterion: AVE > squared inter-construct correlations
- CB-SEM: model fit indices CFI ≥ 0.95, RMSEA ≤ 0.06, SRMR ≤ 0.08
Common Methodology Mistakes That Cause Major Revision
- Missing sample size justification — never just report n=X without explaining why that size is adequate
- No reliability coefficients — Cronbach's alpha must be reported for every scale (α ≥ 0.70 is the minimum)
- Copied instrument description — reusing methodology text from prior papers verbatim = self-plagiarism flagged by iThenticate
- Missing ethical approval statement — IRB or equivalent approval number must appear in this section
- No common method bias treatment — for survey-based studies, Harman's single factor test or procedural remedies must be documented
Get Your Methodology Section Reviewed
ScholarAI's structural analysis evaluates your methodology against Q1 journal standards and identifies specific gaps before submission.
Analyze My Manuscript Free →