I have been watching a series of lectures by Robert Sapolsky for his Stanford University course, Human Behavioral Biology. Stanford undergraduates are getting their money’s worth. The material itself is deeply compelling and Dr. Sapolsky’s up to the minute knowledge and storytelling ability is top-notch. Each lecture is around an hour and one-half long – a big commitment of time but Sapolsky is well-organized, humorous, summarizes well, and keeps the narrative moving. Some might be put off by his style of humor and/or his nerdy, slightly affected use of old Hippie argot.
Enough of the review. Watch seven and one-half minutes of lecture 8 starting here. Spoiler alert: The punchline is given away in the next sentence. The research result he concocted is an example of the importance of thinking about effect size when assessing research studies. I mentioned this idea in this post. Sapolsky continues with a “real-life” example on the relation between birth order and IQ here.
By the way I use the same teaching technique in my introductory statistics class. I pick from a list of possible critiques of research studies that I got from the book, Rival Hypotheses, create a bogus study, ask for student questions, construct far less clever responses than Sapolsky, and explain the critique after the discussion. On the tests and final the students have a briefly stated study to critique. I have a lot of fun with this and my students seem to respond as they did in Dr. Sapolsky’s class.