Reliability and construct validation of the neuropsychological symptom inventory

The Neuropsychological Symptom Inventory (Dean, 1982) was developed to examine the neuropsychological symptomology presented by patients. It consists of 17 items that assess family/medical history that are responded to on a true/false format. In addition, there are 100 items which assess cognitive, emotional-social, sensory-perceptual, and behavioral symptomology, which are to be responded to on a Likert scale from 1 to 4.The present study investigated the validation of the Neuropsychological Symptom Inventory (NSI) (Dean, 1982). In a series of investigations the construct validity, internal consistency, and test-retest reliability of the NSI were examined. Such procedures were considered essential to investigate the utility of the NSI.The subjects numbered 826, for the factor analysis study and the internal consistency analysis, which consisted of college students from introductory psychology classes, and patients seen in both outpatient and inpatient settings. Additionally, there were 132 subjects who participated in the study for test-retest.Utilizing a varimax rotation, a four factor solution was obtained for the NSI, accounting for 47.9 % of the total variance. The results demonstrated high internal consistency for the NSI, with an overall Alpha coefficient of .967. In addition, moderate test-retest estimates were found, ranging from .3365 for Factor 3 to .6223 for Factor 4.These results showed strong psychometric properties for the NSI and suggest its use as a viable alternative to currently available measures. Results are discussed in terms of a need for further refinement of the NSI.Department of Educational PsychologyThesis (Ph. D.