Conversation

ppl often tell me my twitter polls are bad cause sampling bias or lizardmen's constant. but uh, naw man. I often do twitter polls intended specifically to check the extent to which these problems impact my data - and it's different per type of question!
7
141
nearly all data is great, the question is how you interpret it - and I very rarely treat my polls as though they are a representative sample of an honest population.
8
159