0
$\begingroup$

I am doing a meta-analysis. How can I change the reference category such as in a pooled analysis?

Most studies have reported male as the reference group (Female : Male). However in one study the reference group is Female (Male : Female).

Can someone advise me how to change the reference category in this situation? Is there is any reference available?

$\endgroup$
3
$\begingroup$

One way which will work for any software is to simply reverse the direction of your effect size before you enter it. So if you have a mean difference just reverse its sign. The standard error is unaffected but if you are entering confidence intervals you need to reverse them too. If you have odds ratios or some other ratio you need to take their reciprocal but really you should be modelling log odds ratios anyway for which again you just change the sign.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.