Hello everyone:

I am trying to compare the correlation among the same two numerical variables (x1 and x2) in two different populations (n1 and n2) in NHANES data (survey data). I have been using the corr_svy to calculate the correlations but I want to test whether the correlation between x1 and x2 differs significantly in n1 as opposed to n2. I have seen the user command cortesti but this does not take weights. Is it necessary to use weights when comparing these correlations? I also see the user command corrcii (which does take weights) and I could examine whether confidence intervals overlap but I am wondering if it is possible to get an actual p-value. Does anyone have any thoughts or suggestions?

I appreciate any assistance, thank you. I am using Stata 16.1.