Dear All,
I am keen to understand whether I can use mean of one of the variables used in my sample set of studies as moderator. This variable is not used to calculate effect size. For example, if I am meta analytically studying the impact of board size on firm performance, can I use mean board independence, which is one of the variables used in my set of studies, as continuous moderator to test the previously mentioned relationship.
Thanks
Saneesh.
Yes, you can use a continuous variable that varies within a study (e.g., subject age) as a moderator variable in meta-regression. For each study, you could use the mean or median of the continuous variable as the moderator variable level for that study. Just beware that if the within-study variability is greater than the between-study variability, the meta-regression analysis may not be very useful.
Absolutely, the descriptive statistics can serve as possible moderators in a meta-regression analysis. In addition, we have taken continuous moderator variables and categorized them for possible between-group subgroup comparisons. For example,
Thank you Gordon and Michael for your answers. It is really appreciated. It would be great if you could share references of some studies that have used this method.
Article Testing for Moderator Variables in Meta-Analysis: Issues and Methods
Article Testing moderation in network meta-analysis with individual ...
https://campbellcollaboration.org/images/pdf/plain-language/Moderator_Analysis_Williams.pdf
http://www.hermanaguinis.com/JOB2011.pdf
https://www.tandfonline.com/doi/full/10.1080/00220973.2018.1561404?src=recsys
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis