4 ms·

Sample size is a little anemic on this one

• closed 12y ago
  I disagree with this in terms of their main claim. Statistical power depends not only on sample size, but effect size as well. The effect sizes are strong (often Cohen's d is greater than or near 1). This would mean that on average, participants in one group are 1 standard deviation above the other. Pretty easy to detect. I'm on my phone and can't run the analyses, but I'd imagine the type 1 and type 2 power would come out quite high, especially for the second study, which had over 125 participants. (other effects they may have looked at may be underpowered, though)