mograph
Master of Arts
Sorry, I was speaking statistically, where stats defines the "population" as the group under study. If we define the population as being all LEOs, then generalizing from a random sample of LEOs with a large enough sample (at least 30, but 1000 would be nice) would not be unreasonable.Not necessarily. Sticking with the example used, sampling 100% of all LEO-involved altercations still doesn't give you any basis for extrapolating results to the general population.
The key is to define the population: sampling LEOs then generalizing to a population of LEOs makes sense. But generalizing to all humans, all men, or all American men (for example) does not. Yes?
(If you expected a nerdy discussion of stats on a martial arts board, raise your hand ...)