Stephen interviews psychologist Daryl Bem to discuss his theory of “extrasensory pornception.” The interview does have some educational value for an introductory statistics course because:
- It does a good job of describing the research methods used in the study.
- Colbert calls out Bem because the better-than-average, statistically significant predictions amounted to guessing correctly 53% of the time when chance would dictate 50% of the time.
This change is statistically significant, but not a very powerful finding.
The clip and description come from Jess Hartnett (Twitter):
I love this older example. Darryl Bem's interview on The Colbert Report illustrates the p-value vs. effect size debate without using any statistical jargon. AND some cite the Bem research as an important event in the replication crisis. https://t.co/ei63eNv7SX
— Jess Hartnett 📊 (@Notawful) February 18, 2019
Leave a Reply