3 ms·
I fully agree with the critics about the p-values, but what are the best alternatives to analyze and compare data? Most of the time, scientists have to compare
by mandor 13y ago
I fully agree with the critics about the p-values, but what are the best alternatives to analyze and compare data? Most of the time, scientists have to compare the outcome of treatment 1 versus treatment 2; how should they do it "properly"?
What is the HN recommandation?
- Fomite 13y agoEffect measures. Don't just report your p-value, report the actual effect measure, and a measure of uncertainty around it, be it a frequentist confidence interval, Bayesian posterior distribution, etc. More information is better.
- capnrefsmmat 13y agoAgreed. A good reference is Geoff Cumming's "Understanding the new statistics," although its focus on analysis in Excel may put off the typical HN audience. Effect sizes and confidence intervals are much more useful than a p value or two.
- dewarrn1 13y agoCumming had a short treatment of the subject published in Psych Science recently: http://pss.sagepub.com/content/25/1/7.short http://pss.sagepub.com/content/25/1/7.short ; doi: 10.1177/0956797613504966.