6.17 Calculating Cohen’s d
Please note the calculations for pooled standard deviation differ depending on whether sample sizes in each group are equal.
Interpreting Cohen’s d
There are general guidelines for interpreting the result of Cohen’s d. A common guideline is this6:
- A value of 0.2 represents a small effect size.
- A value of 0.5 represents a medium effect size.
- A value of 0.8 represents a large effect size.
Another variation is this:
- < 0.1 = trivial effect
- 0.1 – 0.3 = small effect
- 0.3 – 0.5 = moderate effect
- > 0.5 = large difference effect7
Regardless of which threshold you subscribe to, the Cohen’s d guidelines are disputed by some who argue that researchers should exercise their domain knowledge and subjective judgment instead of relying on arbitrary cutoffs that may not be applicable to every business. But as we’ll see in the section below, Cohen’s d is used in the calculation of sample sizes.
6 https://www.statology.org/interpret-cohens-d/
7 MEERA (n.d.) ‘Power analysis, statistical significance, and effect size’. https://meera.snre.umich.edu/power-analysis-statistical-significance-effect-size