A Tri-Factor Approach to ADHD Assessment: Integrating Multi-Rater Consensus, Situational Specificity, and Symptom-Level Prognostics of Impairment and Diagnostic Status
Restricted (Penn State Only)
Author:
Feldman, Jason Scot
Graduate Program:
Psychology
Degree:
Doctor of Philosophy
Document Type:
Dissertation
Date of Defense:
May 10, 2023
Committee Members:
Lisa Kopp, Outside Unit & Field Member Bradley Wyble, Major Field Member Cynthia Huang-Pollock, Chair & Dissertation Advisor Peter Arnett, Major Field Member Kristin Buss (she/her), Program Head/Chair
Keywords:
ADHD Trifactor Model Informant Discrepancies Assessment ADHD Trifactor Model Informant Discrepancies Assessment
Abstract:
OBJECTIVE: Discrepancies between parent and teacher ratings of ADHD symptomology
reflect legitimate situational variability in symptom expression, with additional evidence that
some symptoms and informants may be differentially predictive of the ADHD syndrome. The
present study utilizes a trifactor model of ADHD to determine whether weighing multi-informant
ratings based on shared and unique discriminative strength enhances prediction of ADHD
diagnoses and related impairment over modern integration approaches. METHOD: In separate
samples of 155 early elementary and 410 middle elementary school aged children, structural
equation modeling (SEM) was used to examine whether rater-specific reports of ADHD
symptoms incrementally predicted cognitive and social impairment over a “rater consensus”
general ADHD factor. Factor loadings from these models were used to calculate informant and
symptom-weighted variants of the “or” and “average” integration approaches, which were
assessed for clinical utility. RESULTS: Teacher-specific (both samples) and parent-specific
(early elementary sample only) ratings predicted ADHD-related impairment over and above the
consensus factor. For 12 (early elementary) and 17 (middle elementary) individual symptoms,
teacher ratings were more strongly associated with broader symptomology than were parent
ratings. Unweighted algorithms showed higher sensitivity to gold-standard ADHD diagnoses than
did weighted ones, while weighted variants showed better specificity. CONCLUSION:
Collecting behavioral ratings from multiple informants is a critical component of the ADHD
diagnostic process, though determinations regarding the optimal approach to integrating this data
should consider the context of the assessment. Implications for clinical practice are discussed.
