well, let's say that there are 100 people in your class in dental school. the 50 worst of them ultimately decide that they aren't cut out for it and drop out. you, who were previously in the top half ranked 26th, now graduate in the bottom half.

Most dental schools are fraudulent institutions.

Truism. Useful idea: one should consider the effect of range restriction when thinking through statistical artifacts applied across groups. This becomes more crucial as you apply it to more g-loaded professions and groups.