Statistical analysis

• 1:14Ask in plain EnglishA tour of the statistics workspace: type a question, watch the analysis run beside the results.
• 1:18The pooled estimateA forest plot read row by row — each study interval, its weight, and the pooled diamond below.
• 1:28Heterogeneity and influenceI² and τ² explained, then a leave-one-out check of whether one study carries the result.
• 1:15Publication biasAsking whether small negative studies are missing — the funnel plot and its caveats, stated plainly.
• 1:28Back to the rowsFrom a pooled number back to the study rows behind it, with each value traced to its source.
• 1:02Build missing variablesAsk for a variable nobody extracted: studies are sorted into categories with each placement recorded.
• 0:57Ask for a pictureDescribe a figure in words and it is drawn from queried data — rows first, picture second.
• 1:16Your own analysis codeRun your own analysis code on the frozen dataset, with the code and method kept on record.
• 1:09Repeatable analysesEvery run is kept — its question, method, and result — so past analyses can be revisited and rerun.

Prefer one continuous video? Watch the full 11:34 cut · Watch on YouTube