Friends Talk to Friends About Drinking: Exploring the Role of Peer Communication in the Theory of Normative Social Behavior

Abstract

Peer communication represents 1 mechanism through which norms are disseminated in social groups. The theory of normative social behavior (TNSB) posits that group identity, outcome expectations, and injunctive norms moderate the relation between descriptive norms and behaviors. This article extends the purview of the TNSB by conceptualizing peer communication as another moderator in the relationship between descriptive norms and behaviors. A survey was conducted among college students (N = 675) to measure their normative perceptions, peer communication, alcohol consumption, alcohol-related consequences, and intention to drink alcohol. As hypothesized, descriptive norms, peer communication, and the interaction between these two variables were significantly associated with consumption, even after controlling for known predictors of both consumption and various mechanisms of normative influences. Controlling for prior consumption, peer communication was also a significant predictor of intention to consume alcohol in the future. This model explained approximately 63% of the variance in intention. This study found a significant relationship between peer communication and alcohol drinking behaviors and intentions after controlling for perceived norms.

Notes

¹Tests of the influence of normative mechanisms on both behavior and behavioral intentions were also conducted in this sample through regression equations. The main effects of the normative mechanisms (injunctive norms, outcome expectations, and group identity) were similar in this study to those reported by (Rimal & Real, 2003).

²This study was conducted in 2001 and used measures from an earlier study conducted in 1998 (Rimal & Real, 2003).

³We do not report the Cronbach's alpha for the index comprising total alcohol consumption because we believe that this index is misleading for our purpose. Cronbach's alpha assumes that all measures comprising an index are mutually supportive, not mutually exclusive, whereas our index comprises mutually exclusive measures both across contexts and across beverage types. Our measure is simply the total amount of alcohol consumed “this past weekend.” Because the unit for the number of drinks consumed for each beverage type was standardized according to the alcohol content—12 ounce of beer = 4 ounces of wine = 1.5 ounces of liquor (Dawson, 1998; Russell et al., 1991)—a simple addition resulted in the total alcohol content across beverage type and social context. It is also important to note that we assessed the extent to which respondents drank alone in one context and with friends in another. Talking to friends was significantly correlated with drinking in each context (at home, r = .44, p < .001; in a bar, r = 38, p < .001; at a party, r = .47, p < .001).