So at what point have you worked with enough women to be allowed to make inferences? When the difference has an over 95% confidence level you usually have enou

So at what point have you worked with enough women to be allowed to make inferences? When the difference has an over 95% confidence level you usually have enough ground to start making inferences on new individuals. Just don't cherry pick the data, i.e. "good male programmers" versus "all female programmers".