26: Interpreting effect sizes
Everything Hertz
00:00
Are You Trying to Get a Power of 80?
The mean median sample size in these 397 studies was 23 per cell. If you're testing populations where you would expect a large effect size, let's say heavily medicated person with schizophrenia versus healthy controls, that sample size is sufficient. That's enough to achieve a power of 80. I like the serious questions about whether that should be added to some kind of permanent record is being accurate. It can be, it can be that bad through to this is actively misleading and a huge pile of burning garbage.
Play episode from 22:39
Transcript
