[R-meta] Calculate correlation coefficient from (un)standardized regression coefficient

Wiertsema, M. m@w|ert@em@ @end|ng |rom rug@n|
Wed May 12 11:02:26 CEST 2021

Hi everyone,

I am currently working on a meta-analysis on the longitudinal
association between bullying (predictor) and social status (outcome).
Because most studies used continuous variables for the outcome, we
want to convert all reported effect sizes into a correlation

Here is where the trouble starts:

I want to calculate r from a path estimate (i.e., regression
coefficient), but the only code I can find for this type of converting
uses an argument for sample sizes of two groups (treatment and
control; r package "esc"), whereas 'my' study has no groups and only a
total n. Do you know whether there is  another way to calculate the
correlation coefficient from an (un)standardized regression
coefficient? Also, if yes: is it enough to calculate r with only a
regression coefficient and a total n or is more information needed and
what type of information is then needed to calculate r ?

For example: I have a standardized path estimate between T1 bullying
and T2 popularity of 0.06 and total n of model = 432. Is it at all
possible to convert this into a r coefficient or do I need more

Please let me know if anything is unclear.
Many thanks in advance!


PhD student
University of Groningen, Faculty of Behavioural and Social Sciences
Grote Rozenstraat 38, 9712 TJ Groningen
email: m.wiertsema using rug.nl | phone: 050 - 363 3182

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