3 ms·
>If you don't have the causal explanations then you cannot generalize beyond the population you sampled. At the individual level. You can still know that you
by coldtea 7y ago
>If you don't have the causal explanations then you cannot generalize beyond the population you sampled.
At the individual level.
You can still know that you have statistically applicable results to the whole population, without having to sample everybody -- just enough to be representative of all kinds.
If you do a random enough pick, then if some people fall outside your sample ("you didn't test for people that don't like water, and the casual explanation requires being OK with water to work, so excludes them from showing the same results") they'll still be outliers statistically.