A diffusion model analysis of transitivity and lexicographic semiorder

The primary aim of the present dissertation is to examine the underlying cognitive processes of transitivity and lexicographic semiorders. To this end, I apply the diffusion model to preferential choice data, where transitivity or lexicographic semiorders are typically considered to model the choice data. In literature, transitivity is often associated with rationality, whereas lexicographic semiorders are usually considered an alternative way to make decisions specifically when the given task seems daunting. Despite their clearly different decision-making processes, little empirical evidence of such different cognitive processes has been reported, so I decide to run the diffusion model analysis to provide empirical evidence of the underlying cognitive processes behind these two models. To do that, I reparameterize drift rate of the diffusion model in terms of subjective values (or utility) of the alternatives. And I conduct a simulation study to test the new diffusion model's ability to recover the data-generating parameter values. Then, I apply the diffusion model to three sets of real data, one from Cavagnaro and Davis-Stober's (2014) experiment, and two from my own experiment. The results imply that people classified to transitivity tend to integrate more information than those classified to lexicographic semiorders to make a decision. More details about the results and implications are discussed in Chapters 4 and 5.