Medical Hypocrisy and Effectism in Cognitive Science
Wednesday, October 30, 2019, 11:00am - 12:30pm
Tillet Hall, Room 226, Livingston Campus
Rutgers Psychology [for Psychology Students (Undergraduate/Graduate)] presents:
J.P. de Ruiter [Tufts University]
In the cognitive sciences, our most common way to empirically support theories is to collect empirical data and then demonstrate the presence of a statistically significant "effect" that would be predicted by our theory.
Most psychologists believe that by using this general methodology, we are following (a variant of) Popper's falsificationism, because in our statistical standard paradigm, we try to falsify the null hypothesis that there is no effect. But in fact, we are doing the *opposite* of what Popper proposed.
Arguably even more dangerous is our assumption that a significant effect suggest its own explanation, which leads to weak, hand-wavy, and badly motivated theories. I will discuss these two methodological problems, provide illustrative examples from cognitive psychology, and suggest some ways to avoid them.