Category Archives: statistical description

Vox lies with statistics

Okay, maybe lies is too strong a word, but Vox‘s German Lopez sure does a good job illustrating numerical dishonesty in this piece on marijuana use in Colorado and Washington since recreational legalization. The headline is “Marijuana use rises in states with legalization,” and … Continue reading

Posted in SCIENCE!, statistical description | Comments Off on Vox lies with statistics

Yglesias makes an innumerate (non-)funny

Yesterday, Matt Yglesias tried to point out the absurdity of a Joe Klein column by turning the tables with the numbers Klein invoked, thereby mimicking and mocking Klein’s argument. Yglesias quotes Klein: Blacks represent 13% of the population but commit 50% … Continue reading

Posted in statistical description | Comments Off on Yglesias makes an innumerate (non-)funny

For when you have a t but you want a d

Suppose you’re doing a meta-analysis and you have a t statistic which you want to use to estimate a Cohen’s  measure of effect size. Appendix D in the 2010 meta-analysis by Spada and Tomita gives the following formula (citing Rosenthal, 1991, Meta-analytic … Continue reading

Posted in statistical description | Comments Off on For when you have a t but you want a d