former

bchadwick Wrote:
-------------------------------------------------------
> Just be sure they didn't remove all unemployed
> charterholders from the sample. There is likely
> self-reporting and selection bias on that anyway.

Actually, you bring up a good point about unemployed people. Maybe it's more accurate to report compensation averages with a zero entry for the percentage of people with no jobs. I don't think most surveys do this. This would give a better expected value of compensation.