[R-meta] effect size similarities

Viechtbauer, Wolfgang (SP) wo||g@ng@v|echtb@uer @end|ng |rom m@@@tr|chtun|ver@|ty@n|
Wed Jun 23 09:20:24 CEST 2021

The equation for turning a two independent samples t-test statistic into a d-value is:

d = t * sqrt(1/n1 + 1/n2)

not t/sqrt(df). But this only applies if the t-statistic comes from a two independent samples t-test, not just any t-statistic.


>-----Original Message-----
>From: Patrizio E Tressoldi [mailto:patrizio.tressoldi using unipd.it]
>Sent: Wednesday, 23 June, 2021 8:40
>To: Viechtbauer, Wolfgang (SP); r-sig-meta-analysis using r-project.org
>Subject: Re: [R-meta] effect size similarities
>Il 22/06/2021 17:23, Viechtbauer, Wolfgang (SP) wrote:
>That is just a simple z-test of a single binomial proportion. This cannot be
>turned into a standardized mean difference (Cohen's d / Hedges' g value) or vice-
>However, for one sample, we can transform a t value in d with the formula
>t/sqr(df). Why is not correct to transform a z value in the same way z/sqr(N-1) ?

More information about the R-sig-meta-analysis mailing list