Advancing Methods and Mathematical Models of Perceptual Decision Making

Abstract

In this thesis I argue that cognitive psychologists can use the combination of sequential sampling models, Bayesian estimation methods, and model comparison via predictive accuracy to investigate underlying cognitive processes of perceptual decision-making. I show that sequential sampling models of simple and choice response time allow for researchers to analyze behavioral data and translate them into the constitute components of processing, such as speed of processing, response caution, and the time needed for perceptual encoding and overt motor responses. I use these methods and models to investigate underlying mental processes related to cognitive load, speech perception, and lexical decision-making. I also show that using different sequential sampling models to analyze the same data can lead researchers to draw different conclusions about cognitive processes, which serves as a caution for carelessly using these models. I also present a novel method that researchers can use to observe cognitive processes unfold online during perceptual decision-making tasks. I then discuss a promising collaboration emerging between researchers in the field of mathematical modeling and neuroscience.