statistical power
statistical power
the probability of supporting a research hypothesis when it is true. Fisher's (1925) approach to inferential statistics was that a Null hypothesis should be tested. If the results were statistically significant then the Null hypothesis could be rejected. Neyman and Pearson (1928) pointed out that there is a second hypothesis involved in significance testing: the research hypothesis. If the results of a study achieve statistical significance then researchers may be committing a Type I error by rejecting the Null Hypothesis when it is in fact true. On the other hand, Neyman and Pearson noted that when a result fails to reach statistical significance a Type II error (which they called an error of the second type) may be committed: the failure to reject the Null hypothesis when the research hypothesis is true. Greater statistical power lessens the likelihood of committing a Type II error.Many writers (e.g. Cohen, 1988) argue that failure to consider statistical power renders statistical significance testing relatively meaningless.