Date of Award

Summer 2016

Document Type


Degree Name

Doctor of Philosophy in Clinical Psychology (PhD)


Clinical Psychology

First Advisor/Committee Member

Amy H. Mezulis, Ph.D.

Second Advisor/Committee Member

John Thoburn, Ph.D.

Third Advisor/Committee Member

David G. Stewart, Ph.D.


The current study examines the relationship between temperament and physiological models of externalizing behavior to externalizing behavior in a community sample of early adolescents. The psychophysiological component of vagal tone, as measured by respirator sinus arrhythmia (RSA), is used with an emphasis placed on distinguishing between its function as a state versus trait measure of parasympathetic nervous system activity. Negative affectivity (NA) is proposed as a general risk factor for psychopathology. Emotion regulation, as indexed by effortful control (EC) and basal RSA, is hypothesized to function as a mediator between NA and externalizing behavior. A moderated mediation model is then proposed with physiological reactivity, as measured by mean stress RSA and RSA reactivity, moderating the β path. A total of 98 youth, 52% female, with a mean age of 12.89 (SD = .82) completed a laboratory task to measure their RSA at rest and while completing an unsolvable anagram stressor task. Temperament measures and externalizing behavior were assessed through self-report questionnaires. The direct effect of NA on externalizing behavior (B = 2.75, t = 3.41, p < .001), was significantly mediated by EC and decreased to (B = 1.28, t = 1.40, p = .17), while basal RSA did not (B = 2.81, t = 3.43, p < .001). Thus, individuals with high trait NA had lower EC and higher externalizing behaviors. Stress RSA did not function as moderator between EC or basal RSA to externalizing behavior. RSA reactivity did significantly act as moderator between EC and externalizing behavior (B = 3.31, t = 2.65, p = .009), and functioned in the full moderated mediation model with a significant conditional indirect effect (-1.65) through EC. The pattern of results suggests that at low levels of RSA reactivity, the effect of NA and EC on externalizing behavior is the strongest. However, with greater RSA reactivity the effect NA and EC on externalizing behavior is disrupted. Results support the utility of differentiating between state versus trait and emotion regulation versus reactivity processes in theoretical and statistical models. Analyzing RSA as a moderator provides a framework for conceptualizing conflicting findings in the literature.

Copyright Status

Additional Rights Information

Copyright held by author.


Copyright Status