Pollsters implement "weights" for their data to test to make their effects mirror the make-up of your voting population. Therefore if a selected poll features a smaller proportion of feminine respondents than you'll find in the wider inhabitants, their responses might be presented further weight in the final effects. But different providers conside