Why are the conclusions drawn about voter sociology sometimes inaccurate

There are many reasons why conclusions drawn about voter sociology might sometimes be inaccurate. One of the problems with studying voters is that of samples. Samples, even when random, are sometimes biased. Moreover, depending on how high voter turnout is, samples might not reflect the reality of all society. The problem of turnout also makes it difficult for the studies to be replicated, which is a problem considering that these conclusions are usually based upon a variety of studies conducted.

i would think becouse not all votes are counted toward whatever everyone is voteing for