I have stratified my results according to Ethnicity (south asians and Europeans), Gender (Male and female) and Age (4 different age groups). So I have a Beta value (from regression) for South Asian/European male/female in a particular age group. I want to carry out a meta analysis of these non overlapping strata. i have been reading around meta-analysis of non-overlapping strata The Mantel-Haenszel Formula, But I understand that I have multiple strata and the formula is not suitable here.
Can any one suggest any possible method for combining the results across different strata please?
Thank you
Saima
I can't really follow your question. What is the outcome? Maybe showing a bit of the data would help other work out what you are trying to do.
Dear Julia,
I am studying the association of Physical activity with glucose and incident Type two diabetes. I have stratified my results for Ethnicity (south Asians and europeans), gender (male and female) and 4 age groups (30-39,40-49,50-59 and 60 or above).
I stratified my results because of a strong correlation between the age and physical activity which was affecting the results. is there any other way to deal with a strong correlation between predictors?
After getting my results for all these strata, i want to use meta analysis to get one value for each ethnic group.
I hope I am making sense!
Please suggest what should I do.
Thank you
Saima
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression