Hello,
I am conducting my first meta-analysis examining neural differences in older and younger item-item memory. I recently read this article (Article Ten simple rules for neuroimaging meta-analysis
), that suggested finding 17-20 studies for your meta-analysis. This is an ALE study.I have 22 total studies, 16 of which include older adults and 13 of which include younger adults. Must these numbers be equal? Or would it be most important for the total sample size of each age group be equal? Or is neither necessary?
Thank you!
Hello Andrew
Several links are available. Examples are given below:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3514119/
https://www.meta-analysis.com/downloads/Meta%20Analysis%20Fixed%20vs%20Random%20effects.pdf
https://www.meta-analysis.com/downloads/when_does_it_make_sense_to_perform_a_meta-analysis.pdf
http://www.creative-wisdom.com/teaching/WBI/es.shtml
Regards,
Zuhair
First, a rigorous systematic review always precedes a meta-analysis. When you do a systematic review/meta-analysis, you are supposed to find (and use) ALL studies that meet the eligibility criteria for your research question. This includes published and unpublished studies. If you do not find all studies, your overall effect size will likely be inflated because of publication bias. Short of finding all studies, it would be appropriate to use a random sample of all studies (but how do you know when you have that?).
- Badges
- Science topic
- Similar topics
- Medicine
- Epidemiology
- Age Groups