Power: Using a post-hoc power analysis, the study is slightly underpowered using cutoff of .80. Effect size is small.
回應:
Response:
Cohen recommended to set the type I error α as 0.05 and type II error β as 0.2 (i.e. power=0.8) to present the acceptable possibility of type I and type II error. In our study, we used the data of the 157 patients who completed both the admission and discharge assessments and found effect size of the PDT was 0.4. Using the post-hoc power analysis with set value of type I error α = 0.05, we found that the power was 0.99 (i.e. type II error=0.01) in our study. So the possibilities of making an inferential mistake would be low.
