The college years are a unique period of newfound independence and freedom commonly marked by increased alcohol use. Indeed, alcohol use is considered by many students to be an important component of the college experience (Osberg et al., 2010). More than half of college students report consuming alcohol at least once in the past month, and nearly one-third report consuming several drinks (4+ drinks for women or 5+ drinks for men) on at least one occasion in the past month (Schulenberg et al., 2021). Additionally, relative to same-aged peers, full-time college students are more likely to report past-month drinking and consuming 4+ drinks for women or 5+ drinks for men on one or more occasions in the past month (Schulenberg et al., 2021; Substance Abuse and Mental Health Services Administration, 2020). Heavy alcohol use is associated with a range of adverse consequences to health and functioning (Barton et al., 2016, Krieger et al., 2018, Shorey et al., 2014, 2015). In some cases, alcohol use can be fatal. An estimated 1,500+ deaths among college students are attributed to alcohol consumption annually (Hingson et al., 2017). The development and improvement of strategies designed to reduce alcohol use and related consequences in high-risk populations are contingent upon our understanding of the major factors contributing to these outcomes. Social norms are one of the most widely recognized factors associated with drinking among young adults (Dotson et al., 2015) and nearly all individually focused alcohol interventions for this population incorporate strategies designed to alter perceived social norms (Carey et al., 2016). Nevertheless, few longitudinal examinations of associations between perceived norms and drinking have distinguished between-person and within-person associations.

Social norms refer broadly to the conceptualizations of normal or expected behaviors in a given population and context (Sherif, 1936). While norms research has expanded to address multiple disciplines, research on social norms in health-related contexts is abundant. Two types of social norms have received extensive attention: descriptive and injunctive (Cialdini, 2012, Cialdini et al., 1991, Kallgren et al., 2000). Descriptive norms refer to the perceived prevalence of a given behavior, whereas injunctive norms refer to the extent to which others approve of a given behavior. In the context of alcohol consumption among young adults, an extensive body of research has demonstrated systematic discrepancies between perceived norms and actual norms, both descriptive and injunctive (Borsari and Carey, 2003, LaBrie et al., 2010, Neighbors et al., 2008). Specifically, college students were likely to overestimate the frequency of heavy episodic drinking among their peers (Pedersen et al., 2010) as well as drinking quantity among peers, such as drinks per occasion (Perkins et al., 2005) and drinks per week (Dumas et al., 2019, Neighbors et al., 2006). Overestimating perceived acceptability of alcohol use among peers (i.e., injunctive norms) is also prevalent (Borsari & Carey, 2003). Importantly, overestimation of peers’ behaviors and acceptability may influence one’s behaviors and acceptability. The Theory of Planned Behavior posits that normative beliefs are a core component of what guides behavior (Ajzen, 1991). Therefore, it seems logical that norms may be useful in predicting an individual’s actual drinking-related behavior. Indeed, descriptive and injunctive norms are powerful predictors of alcohol consumption. Neighbors and colleagues (2007) found that descriptive norms had a large positive effect on drinks per week while injunctive norms had smaller but significant positive effects. While the relationship between norms and drinking behaviors is well-established, the fluctuations of this relationship over time are less clear.

Research on longitudinal associations between norms and drinking offers an interesting insight into the strength and direction of norms’ effects over time. Descriptive norms have been found to predict alcohol consumption two months later such that perceptions of higher drinks per week among peers were associated with higher actual drinks per week (Neighbors et al., 2006). Similarly, injunctive norms predicted drinks per week one year later among sorority and fraternity members, but descriptive norms did not (Larimer et al., 2004). It is important to consider that these earlier studies only used two time points, which can make interpretation of longitudinal associations difficult (Singer & Willett, 2003). Later studies using more than two time points corroborate previous findings. For example, high descriptive norms at the end of students’ first semester of college (T1) predicted increases in alcohol use at the end of their second semester of college (T2) (Ferrer et al., 2011). Interestingly, high descriptive norms at the end of students’ second semester of college (T2) predicted decreases in alcohol use at the end of their third semester (T3) (Ferrer et al., 2011). In a study examining injunctive norms, increases in drinking at the 3-and-6-month follow-ups were predicted by the preceding timepoint’s injunctive norms (Lewis et al., 2015). Despite the documented findings regarding the fluctuating relationship between norms and alcohol consumption over time, this research is limited in that it cannot tease apart whether these effects are due to having higher norms compared to others in general (i.e., between-person effects) or due to normative changes over time (i.e., within-person effects).

Evaluating within-person effects can tease apart whether an individual believes others drink more at certain times and if this impacts their drinking. Dumas and colleagues (2019) tested the within-person effects of descriptive norms and found that students engaged in more heavy episodic drinking at time points in which they believed their drinking groups’ heavy episodic drinking was higher than usual. A similar effect has been shown for injunctive norms: Graupensperger, Jaffe, and colleagues found significant within-person associations for injunctive norms on a variety of drinking outcomes, like the number of weekly drinks (2021). Another study, using separate models for descriptive and injunctive norms, found within-person effects for both types of norms on college athlete drinking (Graupensperger, Turrisi, et al., 2021). Despite these steps made to examine the variability of norms and their effects on drinking, research on longitudinal models that contain both descriptive and injunctive norms is limited. Evaluation of both types of norms is important to gain a clearer picture of how social influences can affect drinking and can be beneficial for improving norms-based interventions.

While research has evaluated how norms and drinking affect each other over time via cross-lag models (Lewis et al., 2015, Litt et al., 2015), few studies evaluating norm fluctuation (i.e., between- and within-person analyses) exist. Those that currently exist only examine either descriptive (Dumas et al., 2019; Graupensperger, Turrisi, et al., 2021) or injunctive (Graupensperger, Jaffe, et al., 2021) norms, despite the concurrent influence of both norm types on drinking (Neighbors et al., 2007). The current study aims to expand upon this prior research by evaluating the influence of both descriptive and injunctive norm fluctuations on college drinking. To determine which and when norms are influential, we examined both descriptive and injunctive norms simultaneously and at both the between- and within-person levels. With the current study design, we were able to evaluate normative fluctuations over the span of a whole year. Based on the well-documented between-person effects of norms on drinking in cross-sectional studies (Lewis & Neighbors, 2006), we hypothesized the following:

H1: Individuals with higher descriptive norms in general will report more drinking, even after controlling for the effects of injunctive norms.

H2: Individuals with higher injunctive norms in general will report more drinking, even after controlling for the effects of descriptive norms.

Additionally, based on previous longitudinal findings on descriptive and injunctive norms separately (Dumas et al., 2019; Graupensperger, Jaffe, et al., 2021; Graupensperger, Turrisi, et al., 2021), we hypothesized that:

H3: Individuals will report drinking more at time points where their descriptive norms are higher than usual, even after controlling for the effects of injunctive norms.

H4: Individuals will report drinking more at time points where their injunctive norms are higher than usual, even after controlling for the effects of descriptive norms.

Exploratory analyses also examined the direction of longitudinal effects of norms and drinking variables. Individuals may adjust their drinking behaviors based on their perceptions of others’ drinking (i.e., conformity) or base their perceptions of others’ drinking habits on their own drinking behavior (i.e., projection; Lewis et al., 2015).