the problem is i get some data, im impatient and im like 'ok 6k people is good enough', i start analyzing it, then inevitably find some niche question where not enough ppl fell into the bins i'm measuring, and then im like 'fuck need more data' and then rinse and repeat 8 times
Conversation
Replying to
Have you considered reaching out to any university sociological researchers for help capturing a "more representative"/broader audience?
2
Replying to
I don't know - but they're experienced in conducting these types of surveys and are probably plugged into networks and such.
2
Replying to
from what i understand after talking to my academic friends in research, they usually do not have broader networks to pull from. there's various sampling methods that can be more (but not totally) random, but the samples u get from those are usually much lower cause expensive
1
6
and also if they involve paying ppl to do them, which is often the case, you introduce risk of noise from ppl not caring about the survey and just completing it to get the money. e.g. Amazon turk, a common "random" source, suffers from this