3 ms·
Any number of things may be true. But one cannot sensibly infer those things from this sample of people that is very far from being a random sample, but is actu
by calderknight 3y ago
Any number of things may be true. But one cannot sensibly infer those things from this sample of people that is very far from being a random sample, but is actually a heavily selected sample. If you're really interested in the state of things "back home", don't seek out a group of people who are very likely to be slanted toward having a particular view (whether that view be positive or negative), and when they express that view conclude that all must have that view. All may have that view, but to conclude that it is actually the case, based on such terrible evidence, is totally absurd!
- MadSudaca 3y agoI'll grant you that it's not a random sample and that there will be some skewness in opinions. But ignoring the wider context of the journey that these people must have made, and considering the nature of totalitarian governments, it seems to me that you're taking an obtuse position.
- calderknight 3y agoIt seems that you have come to agree with me, but haven't realised it yet! You're saying that we must look at more than just the skewed sample, and that's exactly my point.
- robertlagrant 3y agoI don't think that's correct. The point is you can't just dismiss based on selection bias. You could say that the survivors of the Titanic didn't want to drown, sure, but you're selecting for the people who survived. The people who all drowned might have wanted to! I.e. there's a balance. People basically always migrating in one direction must tell you something other than "those exact people just wanted to".
- calderknight 3y agoCan you please point out for me specifically where you and I disagree?
- robertlagrant 3y agoI'm a different person. But the previous person said "yes it is a skewed sample, but the sample is still useful information" and you said "exactly - we should look outside the sample for information".