摘要：Objective: The purpose of this study was to apply the health belief model to predict the text messaging (SMS) driving behaviors of undergraduate college students. Methods: Data for instrumentation were collected from a cross-sectional, convenience sample (n=150). Instrumentation included face and content validity by a panel of six experts, stability through test -retest, internal consistency through Cronbach's alpha, and confirmatory factor analysis through the maximum likelihood method. Multiple regression modeling was employed to develop the theoretical models. Sample size for hypothesis testing was based a power analysis (α=0.05, f2=0.10, 1-β=0.080). Results: Based on the results of an dependent samples t-test (p<0.001), SMS driving behavior was classified as two separate actions: reading text messages while driving (RSMS) and writing text messages while driving (WSMS). Two models were built to test HBM predictors of these two behaviors. The specified RSMS model identified four significant HBM predictors: perceived susceptibility (β= -0.191), perceived benefits (β=0.175), perceived barriers (β=-0.251), and cues to action (β=0.306). Collectively, the model's predictors accou nted for 35.3% of the variance in RSMS driving behavior. The specified WSMS model identified three significant HBM predictors: perceived benefits (β=0.295), perceived barriers (β= -0.319), and cues to action (β=0.240). Combined, the model's predictors accounted for 45.9% of the variance in the WSMS driving behavior. Conclusions: Cues to action was the most powerful predictor of RSMS driving behavior and the least powerful predictor of WSMS driving behavior. For both RSMS and WSMS driving behaviors, barriers exceeded benefits suggesting participants gauge relative barriers to benefits when deciding to engage or refrain from SMS driving behavior. Road familiarity and traffic conditions appear to factor into this cognitive task.
