I have evaluated the prevalence (proportion) of a disease in men and women, in different studies available in the literature. So I have sample sizes for men and women (in different studies) and also know the number of the affected people in each gender. So I can calculate standard errors for each prevalence.

However, the more interesting finding is the ratio of the prevalence (proportion) of disease in males divided by its prevalence (proportion) in females (male:female ratio). I have calculated these too. However I need the SE for those "ratios of proportions" in order to draw a funnel plot.

I couldn't find a proper way to calculate SE for those ratios. Do you think it is possible? If not, how am I supposed to draw funnel plots for my male:female ratios?

Many thanks.

  • 3
    $\begingroup$ I think you're talking about relative risk. That should help with searching. $\endgroup$ – Thomas Aug 4 '13 at 13:13
  • 2
    $\begingroup$ stats.stackexchange.com/questions/19576/… $\endgroup$ – Glen_b Aug 4 '13 at 23:07
  • $\begingroup$ I was familiar with RR but since the groups were not control and treatment, and since I am new to stats, I couldn't figure the similarity out. By Thomas' nice suggestion, I could draw funnel plots of RRs and SEs very fast. Thanks again. $\endgroup$ – Vic Aug 5 '13 at 1:12

Your Answer

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

Browse other questions tagged or ask your own question.