I like that this podcast attempts to make behavioral science accessible, but this podcast obscures important details and sometimes even misleads its audience as a result. There is a way to e accessible but technically rigorous at the same time. Alas, this doesn’t achieve that.
For example, the rapid ode about gender stereotypes recommends taking the IAT to learn about one’s own individual biases, which is a widely known misuse of the test (Nosek et al., 2015). The test measures group-level phenomena and not stable individual traits.
Another example is the Cialdini episodes featuring the claim that elections could be “predicted”. This is misleading. The behaviors Cialdini studied were not an independent variable that predicted the dependent variable. The behaviors were just another indicator of the dependent variable (voting). The behaviors studies mapped neatly into how people voted. So to say that that behavior predicted the election outcome is as insightful as saying that votes predicted the election outcome—i.e., not insightful.
If anyone has recommendations for behavioral science podcasts that care about the details of open science, experimental design, statistical analysts, and logically supported conclusions, please feel free to recommend them. byrdnick.com/contact