What are other parents saying? Perceived parental communication norms and the relationship between alcohol-specific parental communication and college student drinking

Abstract

This study examined parents' normative perceptions of other college parents' alcohol-specific communication, and how parents' perceived communication norms and alcohol-specific communication relate to student drinking outcomes. A sample of 457 student-parent dyads were recruited from a midsize university. Students completed Web-based assessments of alcohol-related attitudes and behaviors. Parents completed alcohol-specific measures of communication norms and parent-child communication, including communication content (i.e., targeted communication) and frequency of communication. Results indicated that parents overestimated how much other parents talked to their college students about the frequency and quantity of alcohol use, but underestimated how often parents initiated conversations about alcohol. In a path model, perceived communication norms positively predicted both targeted communication and frequency of communication. Perceived communication norms and targeted communication negatively predicted students' attitude toward alcohol use. In contrast, more frequent communication predicted students holding more approving attitudes toward alcohol. The relationship between parents' perceived communication norms and students' drinking behaviors was mediated by the parental communication variables and student attitudes. Tests of indirect effects were undertaken to examine meditational processes. The findings underscore relations involving parental perceived communication norms and parents' own alcohol communication and their children's drinking outcomes. The complex relationships of different types of parental communication and student outcomes warrant further research.

Figure 1

Initial path model. Standardized coefficients are presented. All paths have statistically controlled for gender and race on the endogenous variables. E = predictive error. For diagrammatic clarity, the following paths were estimated but not displayed: predictive path from V2 to V6 (β = − .03, ns), predictive path from V3 to V5 (β = .07, ns), and correlation between E2 and E3 (r = .32, p < .001). *p < .01. **p < .001.

Figure 2

Final path model. Standardized coefficients are presented. All paths have statistically controlled for gender and race on the endogenous variables. E = predictive error. For diagrammatic clarity, the following paths were estimated but now displayed: correlation between E2 and E3 (r = .32, p < .001). *p < .01. **p < .001.