Investigating latent decision constructs using computational modeling of behavioral and brain data

The current dissertation is aimed at the study of a variety of latent decision-making constructs, that is, constructs reasoned to underlie variability in decision-making behavior but which cannot be directly observed (e.g., decision strategies). This is achieved via computational modeling, the use of mathematical models to describe, understand, and test hypotheses concerning complex phenomena. The models applied in these studies take into account, or focusses specifically on, individual differences in decision making in either behavioral or combined behavioral and neuroimaging data.
The dissertation consists of five separate research articles, split into two parts. Part I focusses on the MIMIC model approach, a structural equation model allowing to empirically test whether individual differences in decision making are quantitative (i.e., or qualitative (i.e., categorical) in nature. In chapter 2, the method is introduced, tested, and illustrated in combined behavioral and neuroimaging data. In chapter 3, the method is applied to demographic and behavioral data of a gambling machine task. Part II focusses on the study of individual differences in decision strategies, qualitatively distinct latent mechanisms that describe how available information is used to reach decisions. Chapter 4 involved the study of individual differences and age effects in decision strategies in perceptual decision making including advice from others. In chapter 5, we investigate if individuals with ADHD utilize less complex decision strategies due to a reduced need for cognition. In chapter 6, we test if individual differences in the framing effect relate to differences in decision strategy and brain activity.