People you know are a random sample: if you saw Uncle Jim start trying Wellbutrin, and you observed the effects, he’s essentially a scientific study with a sample of one.
If he’s your blood-related uncle, he’s not just random, he’s better than random (when predicting the effects on you and not some arbitrary person).
If he’s your blood-related uncle, he’s not just random, he’s better than random (when predicting the effects on you and not some arbitrary person).