Question

You conducted a t-test in your 691 capstone research. You did not find a statistically significant

difference between your groups even though your sample size was relatively small. Upon

conducting a post hoc power analysis you found that your actual power was 0.57. What should

you do? Is there a problem with having a power less than 0.80 when you do not have significant

differences?

Answer 1

Answer: No, there is not a problem with having power less than .80. Since, the sample size is small it is harder to detect the significant difference, as the probability is low, but detectable. Redesigning the experiment is unnecessary. Since, the power of a statistical test gives the likelihood of rejecting the null hypothesis when the null hypothesis is false. Power quantifies the chance that the null hypothesis will be rejected when it is false or a ‘correct decision’. Thus, power is the ability of a test to correctly with confidence reject the null hypothesis or not. Less power means there are more chances of accepting the False null hypothesis until my head hurts.