Date Available


Year of Publication


Degree Name

Doctor of Philosophy (PhD)

Document Type

Doctoral Dissertation


Arts and Sciences



First Advisor

Dr. Charles R. Carlson


Researchers have demonstrated that college students with strong religious beliefs unsupported by religious behaviors report greater involvement in underage drinking, drug use, and risky sex than students with concordant high or concordant low religious beliefs and behaviors. Recent research also suggests personality traits, belief systems, and environments may be influencing this group’s risky behaviors; however, further research is needed to identify factors contributing to these students’ life choices (including the decision to not support their religious beliefs with specific religious behaviors). This study reports on tests of a psychosocial mediational model, connecting personality traits, religious beliefs, religious behaviors, and underage drinking. Using Structural Equation Modeling and a sample of 411 underage college students, we tested whether the association between five impulsive personality traits and underage drinking was mediated by the discordance of religious beliefs and behaviors. We also tested whether the same predictive effects could be observed using three broader personality trait domains. Although students with discordant religious beliefs and behaviors drank more than their concordant religious peers, we did not find support for the proposed mediational models. Exploratory follow-up analyses offered support for an alternative direction; underage drinking mediated the relationship between eight out of nine personality variables and the discordance of religious beliefs and behaviors. Findings indicated students with strong religious beliefs unsupported by religious behaviors reported higher levels of impulsive traits and perceived invincibility and lower levels of Conscientiousness and Agreeableness than their peers high in both religious beliefs and behaviors; this effect operated indirectly through underage drinking. Implications for directional risk models and points of intervention are discussed.

Digital Object Identifier (DOI)