Date
4-29-2026
Department
School of Education
Degree
Doctor of Philosophy in Education (PhD)
Chair
Maryna Svirska-Otero
Keywords
autism spectrum disorder, cognitive load, facial action units, heart rate variability, sustained attention
Disciplines
Education | Special Education and Teaching
Recommended Citation
Harmer, Carrie Michelle, "A Predictive Correlational Study of the Accuracy of Facial Action Units and Heart Rate Variability to Predict Sustained Attention in Kindergarten, First-, and Second-Grade Students with Autism Spectrum Disorder" (2026). Doctoral Dissertations and Projects. 8250.
https://digitalcommons.liberty.edu/doctoral/8250
Abstract
The purpose of this quantitative, predictive correlational study was to determine to what extent facial action units brow raise, lip press, lip suck, and mouth open and heart rate variability predict sustained attention and how accurately heart rate variability predicts the same facial action units in children ages 5 to 7 in kindergarten, first, and second grade with autism in a northeast Florida public school district. Results add to the existing body of research on how accurately facial expression analysis and heart rate variability predict sustained attention in a larger, younger sample of children with autism. They also clarify the usefulness of these biomarkers for informing instructional planning. Participants included 107 northeast Florida public school students in kindergarten through second grade, ages 5 to 7, with an autism diagnosis or classification on their Individual Education Plans. Affdex 5.2 coded facial action units from webcam footage, Polar H10 and Elite HRV measured heart rate variability, and the Conners Kiddie Continuous Performance Test, Second Edition provided sustained attention t-scores from hit reaction time block change. A multiple linear regression was conducted to examine the extent to which facial action units and heart rate variability predict sustained attention, and a bivariate linear regression was conducted to examine how accurately heart rate variability predicts facial action units. The researcher failed to reject Null Hypotheses 1 and 2. Heterogeneous sensory processing, language, and self‑regulation profiles among participants, along with the naturalistic study setting, appeared to influence results. Therefore, further research is recommended to examine how these autism characteristics shape heart rate variability, facial actions, and sustained attention across different contexts.
