Then wouldn't your initial statement also be conjecture?
![]()
No. In a small sample, 10% of the respondents can easily be within a three sigma standard deviation from being split down the middle, i.e., a coin flip. For a very simple example, flip a coin three times and get two heads and one tails - 66% and 33%, roughly. Does that mean that the odds will always be 2:1 that you'll get heads? Of course not, it just means that the margin of error in such a tiny sample size is very large.
