Keynote Address | Psychonomic Society 59th Annual Meeting (2018) | New Orleans, Louisiana, USA
Building a More Replicable Experimental Psychology: Key Challenges
Hal Pashler, University of California, San Diego, USA
Over the past seven years or so, the “Replication Crisis” has revealed that a disturbing number of well-known findings in many areas of psychology are difficult or impossible to reproduce. Public confidence in the veracity of behavioral research has been seriously undermined. I will discuss some of the key intellectual challenges that the Replicability Crisis poses for behavioral science in general and “psychonomics researchers” in particular. I will contend that the lack of reproducibility that plagues certain areas of psychology is not likely to be attributable to “ubiquitous moderator effects” as some have argued. As to the core fields of cognition, attention, and perception, I will argue that due to the statistical power of within-subject designs, relatively few well-known effects in these fields are likely to be illusory. On the other hand, I will suggest that more critical attention to the discriminating power of data for choosing amongst theories, and greater focus on measurement issues (especially with respect to interactions) could promote not only empirical solidity but also deeper theoretical understanding.