Whose opinion matters? The relationship between injunctive norms and alcohol consequences in college students

Abstract

Harm reduction approaches may benefit from research extending the exploration of predictors of alcohol use per se to those components most directly related to alcohol-related harm. This investigation evaluated the relationship between perceived injunctive norms of alcohol use (level of approval of drinking behaviors in specific situations) and the experience of alcohol-related consequences as a function of typical student reference groups at increasing levels of similarity to the respondent: based on race, gender, Greek status, and combinations of these dimensions, as well as parents, close friends, and the students' own attitudes. Participants were 3753 students (61% female) from two campuses who completed an online survey. Preliminary analyses determined that there were no differences in the relationship between perceived injunctive norms and consequences across the eight student groups of varying specificity, thus all eight levels were combined into one variable of perceived student injunctive norms. However, the relationship between this variable and consequences was weaker than the perceived attitudes of more proximal referents (parents, close friends, as well as their own personal attitudes). Subsequent analyses predicting consequences while controlling for demographic variables and drinking level, revealed that perceived injunctive norms for students, parents, and close friends as well as personal attitudes each significantly predicted consequences. Results suggest an important role for perceived injunctive norms in the experiencing of consequences over and above the amount of consumption and point to types of injunctive norms feedback that might form effective interventions (i.e., incorporating close friend and parent feedback as well as general student feedback).

Introduction

An increasing amount of research is being generated on alcohol use among adolescents and college students. A primary reason for the influx of resources aimed at understanding this dynamic is the alcohol-related negative consequences that are experienced by drinkers and non-drinkers alike. In 2001, an estimated 500,000+ unintentional injuries and more than 1700 U.S. college student deaths were alcohol-related, an increase of nearly 6% from 1998 (Hingson, Heeren, Winter, & Wechsler, 2005). The proportion of college students who reported driving while under the influence of alcohol also increased from 26.5% to 31.4% during the same three years. Excessive drinking, in particular, increases the risks of fatal and non-fatal injuries, academic failure, violence and other crime and unsafe sexual behavior (Goldman, 2002, Presley et al., 1997, Wechsler et al., 2000, Wechsler et al., 2001, White and Jackson, 2004-2005) and is associated with long term repercussions (White & Jackson, 2004–2005). Finally, acute alcohol intoxication is associated with increased accidental and self-inflicted injuries (Rehm et al., 2003). While these statistics are cause for concern, the effects of alcohol use among college students extend farther than the drinking individual. Second-hand consequences are experienced by other students, as well as neighbors in the surrounding communities of college campuses (e.g. physical, verbal and sexual assaults, vandalism, aggressive confrontation, sleep disturbances, etc.) (Wechsler et al., 2002). For example, an estimated 600,000+ students are hit/assaulted by another drinking student each year (Hingson et al., 2005). The negative impact that can result from alcohol use and misuse among college students warrants research that can be incorporated into the formation of prevention and intervention programs.

Most current alcohol treatment programs for college students include moderate drinking and harm reduction as their primary goals (Barnett et al., 2004). Harm reduction approaches have nearly replaced abstinence only approaches and are designed to ‘meet people where they're at' in an effort to reduce identifiable harms associated with alcohol use (Marlatt et al., 1998). The sought after reduction in harm however, is nearly always viewed as a function of reductions in alcohol use as the primary outcome variable. As a result, a large body of research and application has been devoted to identify significant predictors of alcohol use that can be targeted to achieve reductions in drinking and thus, decrease harm associated with use. However, as of 2002 and despite the massive proliferation of college prevention programs, there has not been an overall reduction in problems related to alcohol use among college students (Wechsler et al., 2002).

More recently, White (2006) conducted a rigorous review of the promising technique known as personalized feedback interventions (PFIs) for reducing the harms associated with alcohol use. PFIs provide students with feedback about their own drinking patterns relative to college norms, as well as information about other aspects of their drinking behaviors, related problems and/or perceived risks. White concluded that students receiving PFIs reduced drinking and related problems more than those not receiving a PFI, thus supporting the efficacy of this method. Due to the inefficiencies associated with general prevention programs to reduce problems associated with use and the apparent potential for the PFI technique, the harm reduction approach may further benefit from research extending the exploration of predictors of alcohol use per se to those components most directly related to alcohol-related harm. Moreover, research has found the correlation between drinking quantity and frequency and alcohol-related negative consequences in this population to rarely exceed the moderate range of .6 (Larimer et al., 2001, Turner et al., 2000). This suggests that substantial variance in the experience of alcohol problems on college campuses cannot be explained by drinking behavior alone. The current study sought to investigate this theoretical extension by elucidating direct pathways to alcohol-related negative consequences via an established construct associated with alcohol use; social norms. By identifying the association between perceived norms and alcohol problems, prevention science can be more fully tailored to meet the goal of reduced alcohol-related risk among college students.

Social norms theory posits that perceptions of what constitutes typical actions or beliefs of one's peer group, also known as perceived norms, influence behavior (Berkowitz, 2004). In the context of collegiate alcohol use, perceived norms are typically categorized into two types: descriptive norms refer to the perception of others' quantity and frequency of drinking (Borsari and Carey, 2001, Borsari and Carey, 2003), whereas injunctive norms relate to the perceived level of approval of specific alcohol-related behaviors in specific situations (Cialdini, Reno, & Kallgren, 1990). Research indicates that these perceived norms are among the strongest predictors of alcohol use for this population (for review see Borsari and Carey, 2003, Neighbors et al., 2007, Perkins, 2002). Of particular interest in this relationship however, is that to date, more than 25 studies have revealed misperceptions in peer drinking norms, with students consistently overestimating both descriptive and injunctive norms (Berkowitz, 2004, Perkins, 2002). For example, a nationwide study of college students found that 71% overestimate the amount of alcohol used by peers (Perkins, Haines, & Rice, 2005), and a meta-analysis of 23 college drinking studies revealed misperceptions in 91% of the measures investigated (Borsari & Carey, 2003). Moreover, these overestimations of descriptive norms have consistently been associated with heavy drinking (e.g. Borsari and Carey, 2001, Borsari and Carey, 2003, Lewis and Neighbors, 2004, Neighbors et al., 2007) and in a similar manner, overestimations of injunctive norms have been found to be correlated with personal drinking quantity, frequency, heavy drinking, and drinking to intoxication (Nagoshi, 1999, Perkins and Wechsler, 1996, Wood et al., 1992). A more recent study by Reed, Lange, Ketchie and Clapp (2007) found that injunctive norms were indeed predictive of alcohol consumption, yet this was found to be the case primarily when an individual strongly identified with the reference group. While extant research has identified several moderators of the norms-behavior link, the underlying mechanism by which normative influence is understood to function is via indirect peer influence in which the individual acts in a manner that's characteristic of the perceived group norm rather than their personal view or attitude (Berkowitz, 2004, Borsari and Carey, 2001).

Because of the consistency and magnitude with which perceived norms are overestimated, these constructs are frequently manipulated in PFIs. Typically such techniques seek to amend inaccurate normative misperceptions by exposing and augmenting existing healthy norms among one's referent group that individuals have falsely believed to be atypical (Berkowitz, 2004, Lewis and Neighbors, 2006). By revealing the actual, more modest norm, it is thought that students will then act or adjust one's beliefs to be more in line with the new, accurate, normative perception. Correcting overestimations mediates reductions in drinking.

While much research has been conducted with descriptive norms, decidedly less is known about how injunctive norms function in the college student drinking sphere. Given the similarities in the scope and functionality of descriptive and injunctive norms, the relative lack of research with injunctive norms, and the potential for its use in harm-reducing interventions, the current study focuses on the direct relationship between perceived injunctive norms and alcohol problems. More specifically, this research seeks to evaluate how perceptions regarding others' attitudes towards risky drinking behaviors are directly related to the amount of negative alcohol-related consequences an individual experiences, over and above what is accounted for by alcohol use.

To date, very few studies have examined how injunctive norms and alcohol problems are related or explored the potential utility of such a relationship. Early research evaluating the direct effect of an injunctive norm on alcohol problems found no independent influence on problems, but did find that it was predictive of alcohol use (Wood et al., 1992). Wood, Read, Palfai and Stevenson (2001) found that injunctive norms were related to alcohol problems however the injunctive norms measure was technically a social modeling construct that also incorporated a descriptive norms item (friend's alcohol use). This makes it difficult to be certain about the observed result. A more recent study attempted to disentangle the independent effects of group-specific descriptive and injunctive norms on drinking behavior and consequences among Fraternity and Sorority students (Larimer, Turner, Mallett, & Geisner, 2004). The authors found that participants' baseline perceptions of injunctive norms of other group members significantly predicted concurrent drinking and alcohol-related consequences as well as drinking and consequences assessed at one-year follow up. In contrast, baseline perceived descriptive norms did not exhibit a relationship with consequences at either time point or with drinking at the follow up. Note, in this study injunctive norms were defined for a more proximal reference group (members of one's own fraternity/sorority) than were descriptive norms (members of one's pledge class which could include pledges from other fraternities/sororities). However, it is possible that the influence of injunctive norms on behavior and subsequent consequences persists longer than the influence of descriptive norms. Thus, the current study builds on findings with promising avenues as noted in Wood et al., 2001, Larimer et al., 2004, by evaluating the direct effect of injunctive norms on alcohol-related consequences among a large representative sample of college students.

Because injunctive norms define the social approval by important others, the endurance and influence of these norms on drinking behavior and the ability to reduce negative consequences may rely on the individual's level of similarity to the reference group on which the norm is based (Larimer et al., 2004, Terry and Hogg, 1996, Trafimow and Finlay, 1996). In general, level of similarity appears to operate as a function of proximity. Recent work has revealed the critical importance of considering the specificity of the reference group in the relationship between injunctive norms and drinking (Neighbors et al., 2008). The authors found that the relationship varies considerably by the proximity of the reference group, and studies that have assessed injunctive norms more generally (Alva, 1998, Armitage and Conner, 2001, Larimer et al., 2004, Sheppard et al., 1988) may be masking the impact of specific reference groups. However, to date no research has evaluated this issue with respect to alcohol-related problems. It is important to understand the direct link between perceived injunctive norms for various reference groups and alcohol-related consequences in order to determine what types of normative education may be appropriate and effective (Mattern & Neighbors, 2004).

The relatively small amount of research looking at the relationship between injunctive norms and alcohol-related problems, in contrast to a host of research in the last decade using alcohol use as the dependent variable in examining the effects of perceived norms, is surprising given the large role that norms play in the design of current preventative programs at colleges. The present research evaluates the influence of varying levels of specificity of the reference group on the relationship between injunctive norms and alcohol-related problems, while controlling for overall alcohol use, within a large representative sample from two campuses. This research extends prior work in this area by broadening the focus beyond Greek-affiliated students and incorporating specificity of reference group via several combinations of student-affiliated referents, parents, and close friends, to help identify what types of future normative interventions may be most effective at reducing alcohol problems. We expected stronger positive associations between perceived injunctive norms and alcohol-related negative consequences as a function of level of specificity of reference group such that more proximal rather than more distal groups would exhibit stronger associations between perceived approval of drinking and consequences.