Honors Theses

Date of Award

5-6-2024

Document Type

Undergraduate Thesis

Degree Name

BA

Department

Psychology

Faculty Mentor

Benjamin D. Hill

Advisor(s)

Lisa Turner, Dakota Lindsey

Abstract

Attention-Deficit/Hyperactivity Disorder (ADHD) is the most common psychiatric diagnosis in children and a frequent psychiatric diagnosis in adults. ADHD is a clinical diagnosis based on self-reported symptoms which makes accurate diagnosis challenging. Actuarial prediction has been demonstrated to be more accurate than clinical judgment. The current study explored an actuarial approach to predicting ADHD based on a comprehensive battery of neuropsychological tests. This project utilized test data from Meyers Neuropsychological Battery (MNB) to develop a logistic regression model to accurately predict diagnosed ADHD cases (n=65) from normal functioning cases (n=79) with a history of Mild Traumatic Brain Injury (mTBI). Binary logistic regression analysis was used to refine a predictive ADHD model based on all available test data. The final model correctly identified 73.4% of total cases. The model had 53.8% accuracy in identifying cases diagnosed with ADHD (sensitivity) and 90.5% accuracy in identifying the normal functioning mTBI control cases (specificity). This model was then compared to a theory-driven model of tests commonly associated with impaired performance in ADHD. The theory-driven model did not perform better than the data driven model. Further, as the data driven model was composed of predictor tests that are not typically associated with cognitive deficits in ADHD, the presented algorithm may be more robust to malingering than other diagnostic methods. v

Comments

© 2024 Kimberlyn Williams ALL RIGHTS RESERVED

Download

Available for download on Friday, November 29, 2024

Share

COinS