MAS.S73: Moving Beyond The Replication Crisis: How to Spot Misleading Social Science and Design Better Experiments
Thursday, January 27, 2022 at 4:00pm
Building E15, E15-359
20 AMES ST, Cambridge, MA 02142
Offered for 3 credits P/D/F and open to listeners
Meets T/Th 4-6p in E15-359 (1/11, 1/13, 1/20, 1/25, and 1/27)
The replication and generalizability crisis in social psychology has dramatic implications for designers and engineers. Poor research methodology has damaged our understanding of the relationship between human behavior or experience with intervention design; moreover, the statistical mistakes that underlie the crisis continue to threaten modern HCI research. Luckily, there are many useful resources available to help us navigate and understand the existing literature and prevent our work from falling into similar empirical traps.
In this class we will separate useful, empirically grounded psychology from common misconceptions; we will review conceptual details and common errors for the most widespread statistical analyses, alongside basic meta-statistics, so we can identify and avoid bias in the literature and in our own research; and we’ll look at future directions that integrate strong psychological science with empirical design.
This class has no prerequisites and is tailored towards an intelligent beginner, but we expect even advanced practitioners to find something useful and enjoyable in the material. We hope that the class can serve as a gathering place for those interested in cross-disciplinary work with a psychological basis.