Date of Submission

2011

Degree Type

Dissertation

Degree Name

Doctor of Psychology (PsyD)

Department

Psychology

Department Chair

Robert A. DiTomasso, Ph.D., ABPP

First Advisor

George McCloskey, Ph.D., Chairperson

Second Advisor

James Brad Hale, Ph.D.

Third Advisor

Dr. Candice Stefanou

Abstract

Subtest-level interpretation of intelligence tests is necessary for understanding the relationship between cognitive deficiencies and academic problems and for designing interventions based on assessment results. However, the practice of subtest interpretation continues to be discouraged by those who claim that subtests have poor reliability and thus minimal interpretative power. This perception of subtest instability may be the result of misguided conceptions of reliability and not actual properties of subtests. With this in mind, the present study sought to determine the extent to which a neuropsychologically based performance model fit WISC-IV subtest test-retest data and offered an alternate means of understanding and interpreting the concept of subtest reliability. Higher rates of score progression versus regression were demonstrated for all subtests regardless of whether or not time 1 scores were above or below the mean. Rates of score increases from time 1 to time 2 varied based on the neuropsychological basis of the task being assessed. Results suggest that a neuropsychologically based performance model is superior to a traditional psychometric model for representing WISC-IV subtest reliability and the manner in which individuals use their brains when they engage tasks.

Share

COinS