(PS13-C58) Ecological Momentary Assessment of Youth Anxiety: A Psychometric Evaluation

Sunday, November 19, 2023

9:00 AM – 10:00 AM PST

Location: Regency Ballroom, Level 7

Author(s)

MB

Meghan E. Byrne, Ph.D.

Postdoctoral Fellow
National Institute of Mental Health
Washington, District of Columbia

Co-Author(s)

RB

Rachel Bernstein, B.A.

Postbaccalaureate Intramural Research Training Awardee
National Institute of Mental Health
Bethesda, Maryland
Daniel S. Pine, M.D.

Chief, Section on Development and Affective Neuroscience
National Institute of Mental Health
Bethesda, Maryland
Katharina Kircanski, Ph.D.

Staff Scientist
National Institute of Mental Health
Bethesda, Maryland

Ecological momentary assessment (EMA) captures naturalistic experience in real time and holds promise to improve our understanding and treatment of youth psychopathology. While psychometric evaluation of EMA methods is crucial, there is a lack of consensus on validation standards for EMA measures, which threatens internal validity and poses barriers for interpretation and direct comparisons across studies. The present study involves a comprehensive psychometric evaluation of EMA items probing anxiety symptoms, examining the within- and between-person variability, internal consistency, test-retest reliability, and convergent and discriminant validity. One additional positive affect item served as the comparison for discriminant validity. We hypothesized that the items would demonstrate satisfactory psychometric properties on all parameters. EMA responses (three prompts per day) from 204 child and adolescent participants (M age=12.54, 60.8% female) were evaluated across two seven-day time intervals, including 131 participants with an anxiety disorder and 73 participants with no psychiatric diagnosis. Three well-validated anxiety questionnaire measures were completed by multiple informants, including child self-report as well as parent and clinician-rated anxiety, to assess for convergent validity. Results demonstrated at least satisfactory psychometric properties of the anxiety items on all parameters in youth with anxiety disorders, indicating the items are suitable for use in clinical youth anxiety populations. However, there was poor variability and test-retest reliability in youth with no psychiatric diagnosis, suggesting use of these items in non-clinical samples may jeopardize validity. Accounting for the multilevel nature of EMA data, results also demonstrated lower internal consistency at the within-person prompt level than at the between-person level among both groups, underscoring that EMA methodology is robust to capturing contextually dynamic constructs. Further, the items converged with self-report measures of anxiety, but not with questionnaire measures completed by alternate informants (i.e., parents and clinicians), suggesting parents and clinicians may offer discrete additional information regarding a child’s anxiety symptoms. These results facilitate future studies using EMA to investigate pediatric anxiety and highlight that items as currently structured may not be suited for participants at the lowest end of the clinical severity spectrum. Utilizing appropriate, well-validated methods of assessment is critical to researchers’ abilities to test clinical theories and develop personalized interventions for youth. Future work should continue to examine the psychometric properties of EMA protocols.