[Statistical Assumptions] Do elite samples violate assumptions?


No cake for spunky
I doubt there would be any way to know if the variances were equal and normal ahead of time. It would depend on your specific sample.

A different question is what population are you trying to model? And would a sample of CEO's reasonably represent this? I add in passing that the odds of CEO's responding to your sample are not real good and response rate is one of the most serious (if commonly ignored) problem in sampling.