Maybe you’ve heard that you can get clues about whether someone is lying by what direction they look when they talk. The most common form of this idea is that if someone is looking up and to their left, they’re more likely to be accessing real visual memories (associated with truth), and if they’re looking up and to their right, they’re more likely to be constructing visual images (associated with lies). But there is no basis for this; in fact, many studies have found evidence against that claim. This idea and other more broad ideas about eye movement clues were popularized by NLP (Neuro-Linguistic Programming), a school of thought whose core ideas have been debunked time and time again. In this episode, I talk to Tim Levine, a respected deception detection researcher. We talk about: the eye direction idea; the huge amount of bullshit in the pop behavior analysis space (e.g., shows like the Behavior Panel); reasons why the spreaders of this bullshit are so popular and successful; what the science says about using behavior to detect deception; why it’s so difficult to use behavior to detect deception; the idea that you need to establish “baselines” for people to aid you in reading them; how behavioral patterns in games/sports can differ from more real-world non-game scenarios; confirmation bias in the behavior analysis space, and how even smart researchers can be unreasonably biased in favor of their own ideas; Paul Ekman’s work; and more. 

Learn more about the show and get transcripts at⁠ ⁠behavior-podcast.com⁠⁠.   

Learn more about your ad choices. Visit megaphone.fm/adchoices