Why null hypotheses?
If we hope to demonstrate the research hypothesis that children under stress are more likely than normal children to exhibit behavior problems, we immediately set up the null hypothesis, population mean = 50. We use the null hypothesis for several reasons. First, the philosophical argument, put forth by Fisher, the father of statistical hypothesis testing, is that we can never prove something to be true, but we can prove something to be false. Observing 300 dogs with 4 legs does not prove that the statement “Every dog has four legs.” However, finding one dog with three legs does disprove the original statement without any shadow of doubt. Second, the null hypothesis provides us with the starting point for any statistical test. Consider the case in which you want to show that the mean behavior problem score is greater than 50. Suppose further that you were granted the privilege of proving the truth of some hypothesis. What hypothesis are you going to test? Should you test the hypothesis
