ABSTRACT
A number of general issues must be considered when people use inferential statistics. A key differentiation to be made in experimental designs and statistical inference is that concerning between- and within-groups designs and analyses. In the one-tailed t-test, the coach is really only interested in determining whether the new training regime produces significantly better results than the old regime. The condition of robustness for a one-tailed t-test requires that the smallest sample size be larger than 20 subjects. In a two-tailed z-test, one-half of the critical region occurs at each end of the normal distribution. The researcher doesn't know which mean will be larger than the other one. The level of confidence that people choose is the exact probability of committing a First Type error. A second Type error can potentially be made only when the researcher accepts the null hypothesis. Power is the probability of rejecting the null hypothesis when it is false.
