v3.2.0

Minor release: a new NMA template, mentor-reviewed methods text, and build-time versioning. See CHANGELOG.md for details.

• Concept DOI (always latest): 10.5281/zenodo.20586625
• Companion paper (MetaArXiv): 10.31222/osf.io/wj52n

Added

• New network meta-analysis template (templates/NMA/, English only; PRISMA-NMA, CINeMA).
• Mentor comments link to SRWS-PSG Slack training lessons instead of the Facebook group.

Changed

• Methods text (intervention review, NMA), following the template mentor's review: cross-over trials, clinical comparability before pooling, and between-study variance / confidence interval / zero-event settings stated explicitly.
• "Use of artificial intelligence" section in all four templates is organised by purpose (writing vs. conducting the review).
• The version and date shown in rendered templates now come from the git tag at build time; per-template version numbers are retired.

Fixed

• Embase (Dialog) RCT filter, line S23: missing parentheses that removed genuine RCTs.
• Missing-participant sensitivity analysis, cluster-RCT handling (Cochrane Handbook 23.1.4), and funnel-plot tests by effect measure (Cochrane Handbook 13.3.4.4).
• Bibliographic details of the AI position statement (Flemyng 2025) and Gartlehner 2025.

Licensed CC BY 4.0.