Melvin's digital garden

effect size

Showing CI of effect size is more informative than binary statistically significant or not

Use relative risk for change in proportion of two groups

Use difference of mean/median of two groups

Single case effect sizes https://cran.r-project.org/web/packages/SingleCaseES/vignettes/Effect-size-definitions.html

  • same cohort with different phases, phase A followed by phase B

Links to this note