I suspect they take a very small sample of the population at random and multiply up the numbers. I would like to know how significant (in the statistical sense) the figures are.

That is exactly how it is done.

Say for example Hull has 100,000 people living in it, they ring 100 people in Hull and ask if they participate in RL, if 1 person

out of the 100 says they play RL, Sport England say that 1% of the Hull population play RL which in this simplistic case Hull would have a figure of 1000 participants.

It is by no means an ideal way to find out figures but it the one used and all sports have to live with it. I know the RFL do submit there own figures based on how many games are played each week and how many player registrations there are.

I wonder if anyone on here has been contacted and asked to take part in a survey?

Problem is a lot of people just refuse point blank to take part in phone surveys.