I want to make a sensitivity analysis to test whether there is a statistically significant difference between the observations included for the study (n = 100,000) and the excluded observations (n = 20,000).

I generally have binary categories, for which I am using the two-sample test of proportions. Then, I have one continuous variable, for which I use a t-test. My question is: I have one categorical variable with 4 categories. Which test should I use to test if this variable differs significantly between the included/excluded groups?

  • I was thinking a chi2-test, but I fear that the very unequal group sizes will not provide a valuable results.

Thank you for your time.

  • 1
    $\begingroup$ Also look at other threads in our equivalence tag. $\endgroup$
    – mkt
    Aug 9 at 10:04