[R-sig-ME] Effect sizes for mixed-effects models

Schäfer, L. (Lena) |@@ch@|er @end|ng |rom @tudent@ru@n|
Fri Oct 11 18:01:30 CEST 2019

Hi David,

Thank you for your response! We have the raw data for some studies but only info on the sample/cluster size, the lme4 output and the respective ICC for other studies. Since we are conducting a meta-analysis, we cannot get access to all data-sets (eg., problems with data protection, unpublished data-sets). However, using the info on the sample and cluster size and the ICC, we can derive the effective sample size (Aarts, Verhage, Veenvliet, Dolan, & van der Sluis, 2014) related to the study.

I am not sure what you are referring to with *lower bound for the effective sample size*. Is this value calculated using another formula or is it essentially equal to the number of participants (as suggested by Phillip Alday)?

Thank you for a clarification!


Am 10.10.2019 um 19:34 schrieb David Duffy <David.Duffy using qimrberghofer.edu.au<mailto:David.Duffy using qimrberghofer.edu.au>>:

we want to derive the variance related to Cohen’s d for a mixed-subjects design with some participant conducting a task
only in the control condition and other participants conducting the task in the control and in the experimental condition
(within-subjects design).

[...] we only have access to parts of the raw data-sets

Do you have the raw data or appropriate summary statistics from this mixed-subjects study? Could you get it? I think the "usual" approach would be to use a conservative lower bound for the effective sample size.

Cheers, David Duffy.

	[[alternative HTML version deleted]]

More information about the R-sig-mixed-models mailing list