What are some terms related to hypothesis testing that you are already familiar with? Why does a null and alternative hypothesis have to be mutually exclusive?
I actually did not know any specific terminology related to hypothesis testing. I can remember from high school that a hypothesis has to be testable. It is not just a prediction, but a prediction that can be empirically tested. That is about it though. After reading about hypothesis testing, however, I think I must understand even less, or that is just the way it seems I suppose. I understand the necessity for a null hypothesis because in my life it has sometimes been easier to explain why I do not want to do something, rather than specifically why I want to do something. As to why they must be mutually exclusive, if both could be true, then the chances of statistical significance are greatly diminished. Seeing as how we are looking for an extreme change from the comparison distribution, the overlap would fundamentally undermine the purpose of null-hypothesis testing (NHT). Besides, the p-value is supposed to be the last whatever percentile (.05 or .01) away from the mean of the comparison distribution. So I don’t see how they could overlap. If the results do not indicate enough contrasts between the sample distribution and the comparison distribution, then it is not statistically significant anyway. So for hypotheses to overlap statistical insignificance is almost a must. I hope I got that right. It makes sense in my head anyway.
Aron, A., Aron, E., & Coups, E. (2006). Statistics for psychology (4th ed.). Upper Saddle River, NJ: Pearson/Allyn Bacon.