0

I've calculated an effect size for the difference between two standardized pre-post differences of two independent groups of subjects.

That is: $[Mean_{dif_{group~1}}/SD_{dif_{group~1}}]-[Mean_{dif_{group~2}}/SD_{dif_{group~2}}]$

where $Mean_{dif}$ is the difference between the means of pre- and post-tests of each group. And $SD_{dif}$ is the standard deviation of the difference scores for each of the two independent groups.

I was wondering what is the standard error of this effect size?

P.S1: This effect size also appears in Table 1 (row 5), p. 114 of this article. The same article, however, doesn't seem to provide the sampling variance of this effect size in Table 2, on p. 117.

P.S2: This answer provides an SEM solution to obtain SE for pre-post differences in only one group of subjects.

rnorouzian
  • 3,056
  • 2
  • 16
  • 40
  • Does this help https://stats.stackexchange.com/questions/256053/variance-of-cohens-d-for-within-subjects-designs – mdewey May 05 '19 at 11:02
  • @mdewey, unfortunately that article doesn't provide any direct insights. But Table 1 and Table 2 of [**this article**](https://pdfs.semanticscholar.org/d88e/7804b7df89d229fd5db7a1c567b94ff18f33.pdf) does provide some better insights. – rnorouzian May 05 '19 at 13:30
  • It should be straightforward to extend it to multiple-group analysis in SEM if you have the required means and covariance matrices. – Mike Cheung Jun 08 '19 at 06:56

0 Answers0