Since when does the broadness of an observation or argument have any impact on its validity? Who made up that rule?

I'm confident that people tend to extrapolate anecdotal negative experiences about groups of people in bad faith, and that, as the group size increases, this tendency increases relative to the availability of more objective observations. If that's true, the broader the group subjected to a negative generalization, the more likely it is that the observation is as I described. Combined with intuition about the nature of the negativity, I can sometimes feel confident expressing an opinion about the observation one way or the other (e.g. "I think X people are less conversational" vs "I think X people are mostly violent thugs").