Th e long-term impact of studies of statistical power is investigated using Cohen's (1962) pioneering work as<br>an example. We argue that the impact is nil; the power of studies in the same journal that Cohen reviewed<br>(now the Journal of Abnormal Psychology) has not increased over the past 24 years. In 1960 the median power<br>(i.e., the probability that a signifi cant result will be obtained if there is a true eff ect) was .46 for a medium<br>size eff ect, whereas in 1984 it was only .37. Th e decline of power is a result of alpha-adjusted procedures.<br>Low power seems to go unnoticed: only 2 out of 64 experiments mentioned power, and it was never estimated.<br>Nonsignifi cance was generally interpreted as confi rmation of the null hypothesis (if this was the<br>research hypothesis), although the median power was as low as .25 in these cases. We discuss reasons for<br>the ongoing neglect of power.
The CAUSE Research Group is supported in part by a member initiative grant from the American Statistical Association’s Section on Statistics and Data Science Education