Bachelor's degree holders encompass a much larger set of population, with a lot more variance than the set of PhD holders. I'd think that if you can get a set o

Bachelor's degree holders encompass a much larger set of population, with a lot more variance than the set of PhD holders. I'd think that if you can get a set of "people without PhD that could have been capable of getting one" and compare it with PhD holders, the difference would be far smaller, if any.