Question

Based on inference, probability, sampling, standard error, and the CLT, how can polls be wrong, yet at the same time still be trusted/relied upon in the future?

Answer 1

There are few reasons why the polls are wrong. Some of them are

1) when sample is collected, one randomly selected likely voters and therefore at the time of results, we get another result as compare to polls.

2) Since polls are based on sample and a sample only can represent population. So people who are not included in sample may vote against those who are include in sample and hence the results can be changed.

3) sampling technique used in poll may be not completely random and it is skewed to some candidate and therefore we get the other result in poll as compare to real one.

4) Even if we collect a random sample, even though we do not know that individual's responses are correct or not. may be their mind is changed at the time of voting.