Date of Award

May 2014

Degree Type

Thesis

Degree Name

Master of Science

Department

Psychology

First Advisor

David C. Osmon

Committee Members

Krista M. Lisdahl, Bonnie Klein-Tasman

Keywords

Elementary Cognitive Tasks, Executive Functions, Information Theory, Reaction Time

Abstract

Executive functions (EF) are an umbrella construct in neuropsychology that have received significant attention from both clinicians and researchers in recent years. Despite the wide array of definitions of EF and lack of agreement about such constructs, there seems to be a commonality underlying their theoretical frameworks that has to do with the ability to internally regulate one's behavior. In an attempt to overcome inherent limitations to the construct of EF, the present study used elementary cognitive tasks (ECTs), based on information theory (IT) and a reaction time (RT) paradigm, to establish preliminary feasibility of ECTs to assess behavior regulated by internal rules as a measurement of EF and distinguish EF from non-EF cognitive abilities. Therefore, four ECTs, two putative non-executive direct response tasks (0- and 1-bit non-EF tasks) and two putative executive internal rule tasks (1- and 2-bit EF tasks), were developed and administered in college students. These tasks were given to 30 intact undergraduate students. It was hypothesized that the non-EF tasks would show a linear increase in RT as task complexity increases that follows the Hick's law. Additionally, it was hypothesized that the EF tasks would show an exponential increase in RT as task complexity increases. Results supported the hypothesis showing a linear increase in RT on the 0- and 1-bit non-EF tasks, consistent with past literature, and a nonlinear slope associated with the 1- and 2-bit EF tasks; the dramatic nature of this nonlinear relationship was even better demonstrated when increasing EF complexity (1- to 2-bit EF tasks). This nonlinear increase from direct response to internal rule response was demonstrated by the increased variance explained by the quartic curve fit compared to a simple linear fit. These results strongly support the thesis that the EF bit tasks are qualitatively different from direct response tasks, and puts EF assessment on a firm measurement basis that not only precisely defines the construct, but also measures it at the ratio level of quantification.

Share

COinS