This talk will examine the use of machine learning models in personality research, how such models can be considered as measurement tools, and how they fare concerning standard criteria of test evaluation, such as validity. Uljana Feest will discuss an overview of the big-five model of personality, as well as the notion of construct validity. Following that, he will show that even though the big-five model of personality has long been regarded as having construct validity, there remain open questions concerning the theoretical interpretation of those factors. Feest will argue that while ML models add to the construct validation of the big five, they have similar shortcomings as traditional measures of the big five. Questions about possible positive contributions ML models might make to personality research, while also taking a closer look at potential problems of algorithmic biases that might be said to arise from the data that ML models are trained on, will be brought up and discussed.

Uljana Feest studied psychology, philosophy and history and philosophy of science (HPS) in Frankfurt, Bristol and Pittsburgh. After completing her Ph.D. at the University of Pittsburgh, she was a postdoctoral researcher at the Max Planck Institute for the History of Science (Berlin) and then held a position as assistant professor at the Technische Universität (TU) Berlin. Since March 2014, she is professor of philosophy at the Leibniz University of Hannover.

https://msu.zoom.us/j/99329403867 Passcode: Levels