transparentstats / guidelines

Transparent Statistics in HCI guidelines, FAQs, and exemplar analyses
https://transparentstats.github.io/guidelines/
Other
20 stars 7 forks source link

Add something about alternative effect sizes to cohen's d #102

Open mjskay opened 6 years ago

mjskay commented 6 years ago

E.g. probability of superiority as an alternative that makes fewer assumptions and is easier for people to interpret, but incorporates variance (unlike mean differences). Plus cite that paper on the topic (see slack channel).