Abstract
Small effect sizes in even the most well-executed of longitudinal studies raise the question of how good social science will ever be at predicting important life outcomes. Here, we elaborate on Lewin (1936) and propose that life outcomes (L) are determined by the cumulative effects of the person (P) and their environment (E): Life Outcomes L = f(P, E) + e. Using this model, we propose that there are at least four reasons why life outcomes are inherently hard to predict: (1) the dynamic nature and the plurality of both personal characteristics and their environmental circumstances (i.e., P and E change over the life course and are themselves multiply determined); (2) complex interactions between P and E; (3) the substantialness of e due to different types of errors; and (4) the low base rates and range restriction in certain life outcomes. Understanding the theoretical limitations in predicting life outcomes is crucial for the interpretation and advancement of social science.
Chayce Baldwin
PhD Candidate
Chayce Baldwin is a social psychologist studying well-being and culture using diverse methods and populations.