I have the SMRs of different local areas. The same reference population was used for the standardization. The local areas were divided into 5 groups (group1, group2,...,group5). I want to know if the groups have the same mean SMR or not.
I performed an ANOVA to test if the means of the SMRs between the groups are significantly different. The question is: is it correct/reasonable to calculate the mean of SMRs?

  • $\begingroup$ You should include the size of the groups, this is important when for ex. combining means. Better yet, do not average anything and use the whole data (with subtracted reference mean) and run an ANOVA. $\endgroup$ Feb 28 at 14:08
  • $\begingroup$ How do I run an ANOVA if I have only one value per group? $\endgroup$
    – Luke
    Feb 29 at 7:45
  • $\begingroup$ That's what I said, do not aggregate anything to run an ANOVA. $\endgroup$ Feb 29 at 7:45


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.