So to test the alternative hypothesis, we set up a ``straw man'' null hypothesis that we then try to knock down by using the data - this is a fairground analogy where we imagine throwing balls (data) at a straw man (hypothesis) in order to knock it over. If the data are found to be inconsistent with the null hypothesis, we can reject the null hypothesis in favour of the alternative hypothesis - something signficantly different from sampling has happened. If the data are found to be consistent with the null hypothesis, this does not imply that the null hypothesis is necessarily true but only that ``data are not inconsistent with the null hypothesis''. In other words, we are not allowed in this case to make a big fuss about how exciting the result is - it could easily have happened by chance. This may seem a depressing kind of result but in fact non-rejections can often be as informative and as useful as rejections and so deserve to be published. For example, Sir Gilbert Walker successfully isolated the Southern Oscillation as a leading global climate pattern by using statistical testing to eliminate all the other confusing world-wide correlations that were due to chance.