Methods notes for reviewers mid-extraction.

Converting a standard error to a standard deviation for meta-analysis

The Cochrane Handbook formula for recovering an SD from a reported SE or confidence interval, with a worked example and the two mistakes that flip results.

2026-07-10

RevMan or metafor for your first meta-analysis?

An honest comparison of RevMan and metafor by the data shape each wants at import and the review situation you're actually in, not tool tribalism.

2026-07-10

Getting a mean and SD from a median and IQR: when it holds, when it lies

The Wan/Luo estimators for converting a reported median and IQR into a mean and SD, the skew check that tells you when the conversion is defensible, and a worked example.

2026-07-10

How long does data extraction actually take for a systematic review?

The arithmetic nobody does before committing to a review timeline: per-paper hours, dual extraction, reconciliation, and where the time really goes.

2026-07-10

Hedges' g vs Cohen's d: the small-sample correction reviewers forget

Cohen's d overstates effects in small trials. The J correction that fixes it: worked example, when it matters, and how to report it.

2026-07-10

Extracting a crossover trial: why its conditions are not two arms

Per-condition means from a crossover trial aren't independent arms. The paired-difference route (Cochrane Handbook ch. 23), the within-person correlation, worked.

2026-07-10

What a data extraction form for a systematic review should contain

The extraction fields that actually prevent meta-analysis errors: arm-level N at the timepoint, measure plus unit plus direction, an SD-versus-SE flag, and a source location for every value.

2026-07-10

What data extraction errors actually look like

A working taxonomy of the errors that survive extraction (transposition, wrong arm, SE-as-SD, unit drift, change-vs-endpoint) and why dual extraction catches some but structurally misses others.

2026-07-10