Date of Award

August 2019

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Psychology

First Advisor

David Osmon

Committee Members

Bonita Klein-Tasman, Susan Lima, Hanjoo Lee, Christine Larson

Keywords

diffusion model, elementary cognitive tasks, executive function, neuropsychology, non-Gaussian statistics, reaction time

Abstract

The present study seeks to further investigate and refine the three-factor model of executive function (EF; Inhibition, Shifting, and Monitoring/Updating) known as the unity/diversity framework (Miyake et al., 2000). Past work in this area utilized “power” tasks that prioritize accuracy and difficulty, but real-world problem-solving incentivizes quick and efficient solutions. Ten computerized reaction time (RT) tasks: four elementary cognitive tasks (ECTs; Jensen, 1987; Santos, 2016) with progressively increasing task demands and six EF tasks. The ratio scale of RT necessitated the use of non-Gaussian statistics to better describe distribution shape, while diffusion modeling (DM; Ratcliff, 1978) was used to interpret task complexity and performance. Generalized Regressions used ECT parameters to predict EF task parameters. DM analyses indicated Shifting was the most complex factor, followed by Monitoring/Updating, and Inhibition. Shifting and Monitoring/Updating were predicted by internal rule ECT parameters and non-executive-parameters, although the specific internal rule parameters were not unexpected. Inhibition was solely predicted by non-executive parameters, and almost exclusively choice RT. Within-task correlations between DM parameters were either positive or non-significant, save for the STOP-IT task, and not negative as expected. Overall, the present study demonstrated the utility of computerized RT tasks in evaluating EF, non-Gaussian parameters in better describing RT data, and DM in interpreting task complexity. Investigating the efficiency aspect of EF offers an important complement to tradition “power” approaches to psychological measurement and represents an element of ecological validity that current widely-used measures lack.

Share

COinS