I am continuously amazed at some of the statistics and associated graphs I see in published research. One recent example I saw caught my interest first, when I saw the legend for the stars indicating significance levels.Traditionally, one star is for 0.10; (IF you still use it, see the dance of the p-values and others to see why this is no longer really used, ESPECIALLY when many variables are tested, as was the case here) two stars for 0,05, and three stars for 0,01. Alternatively,: 1 star: 0,05; 2 stars: 0,01 and 3 stars: 0,001.This article chose a different way, where two is best, one is medium and three is worst.

Looking at some of their results, with one finding at three star significance.. well..examine the graph… do you think this is a real finding, and that the difference between the two lines is a real and significant difference?

I saw a cartoon today, that summarizes this well.

 

Share This