What was the sample size, 100, 1000, 10000, 100000, 1000000?
I would guarantee that it was not 26M
I see that it was 8,140. Not a lot in the scheme of things.
Actually 8,000+ is a huge sample for a survey. For polling on simple questions, a sample of just 2,401 can return results within a 2 percent margin of error at a 95 percent confidence level for a population of 26 million ...
so long as the sample is properly representative.
That properly representative sample was what was missing with the Western Sydney University flag survey, at least with regards to the question of whether Australians want to retain the current Australian flag. The survey, titled the '
Alternative Australian Flag Survey', was web-based and open to any and all comers who knew about it. Despite being sponsored by the Australian Research Council, the survey was not all that well publicised and by its name alone, it almost certainly attracted participants of a particular mindset, specifically those looking for an 'Alternative Australian Flag'.
With regards to what the survey set out to do, namely determining which of six alternative flag designs was most popular, the survey generated reasonable results. With regards to gauging the general population's appetite for actually changing the flag, that survey was generally, if not completely, useless.