|
|
|
|
|
|
by rm999
4235 days ago
|
|
|
The sample size doesn't indicate how representative a sample is, the polling method does. In this case, it's almost certainly not representative - internet polls rarely are. But even a sample of 5000 wouldn't be representative. You may be thinking of statistical significance. 471 is actually often a pretty good sample size, you would e.g. need <100 responses to have a fairly significant estimate of the mean salary of HNers (assuming a representative sample). |
|
|