A Big Data Analysis of the Associations Between Cognitive Parameters and Socioeconomic Outcomes

Evidence accumulation models allow researchers to estimate a set of cognitive parameters fromempirical response times and accuracy data. These parameters can quantify individual differencesand index, for example, a person’s mental speed or decision caution in a cognitive task. Recently,researchers have become interested in how these parameters might be related to other measuresof individual differences, particularly cognitive abilities. Higher estimates of mental speed, asindexed by the drift rate parameter, are linked to higher cognitive ability scores which, in turn,can predict key socioeconomic outcomes (e.g., educational attainment). Yet, we know very lit-tle about how evidence accumulation model parameters relate to real-life outcomes. To addressthis gap, we explore the associations between evidence accumulation model parameters and threesocioeconomic outcomes—educational attainment, income, and job prestige—in a large onlineimplicit association test sample. Our analysis allowed us to extract small, yet stable associationsbetween cognitive parameters and socioeconomic outcomes from a task with relatively few exper-imental trials. Surprisingly, drift rate variability indexing trial-by-trial stability of mental speedemerged as the strongest predictor of socioeconomic outcomes. Variability parameters have beenlargely overlooked in studies of individual differences, and our exploratory results suggest promis-ing directions for research on higher moments of cognitive variables. Our study also highlightsthe utility of model-based big data approaches that have only recently become practically feasiblethrough advances at the intersection of deep learning and statistics.