social norms experiment

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• J Stud Alcohol Drugs

Experimental Test of Social Norms Theory in a Real-World Drinking Environment

Mark b. johnson.

a Pacific Institute for Research and Evaluation, Calverton, Maryland

Social norms theory articulates that behavior is influenced by perceptions of behavioral norms. Social norms interventions attempt to modify perceptions of what behavior is normative as a means of influencing actual behavior. Social norms interventions have been widely used on college campuses to reduce the level of student drinking. The effectiveness of these interventions has been mixed. A social norms program might fail because the intervention operations failed to sufficiently implement social norms theory in the real world or because of the theory’s limitations. Our research involves an experimental examination of the impact of social norms information on actual drinking behavior within a real-world drinking environment.

Nearly 3,000 participants were interviewed and randomly assigned to one of nine social norms feedback conditions before heading to bars and nightclubs in Tijuana, Mexico. These same participants were resampled, interviewed again, and subjected to breath alcohol analysis when they returned to the United States.

We found that persons whose perceptions of normative drinking changed (became more accurate) during their visit to Tijuana consumed relatively less alcohol. We also found that providing participants with social norms feedback produced more accurate perceived norms. However, the effect sizes were too small to produce statistically significant results showing that social norms feedback could effectively reduce drinking via changing normative perceptions.

Conclusions:

Our research demonstrated that providing social norms feedback changed perceived drinking norms and that changes in perceived norms were correlated with reduced drinking. Effect sizes, however, were quite small.

A lcohol use is highly prevalent among U.S. college students ( Substance Abuse and Mental Health Services Administration, 2009 ), with an estimated 1,825 alcohol-related deaths and 500,000 alcohol-related injuries among college students annually ( Hingson et al., 2009 ). Social norms marketing—campaigns to correct misperceptions about alcohol use—is one popular strategy for tackling the problem of student drinking. The National Institute on Alcohol Abuse and Alcoholism (2002) described social norms marketing as having “logical and theoretical promise” but lacking convincing and comprehensive evaluations (p. 20). Nevertheless, an estimated 48% of all 4-year colleges and universities in the United States have implemented a social norms marketing campaign on their campuses in an effort to reduce college drinking ( Wechsler et al., 2003 ; also see Keeling, 2000 ).

Social norms theory

Social norms interventions are predicated on the fact that people are influenced by the real, imagined, or implied behavior of their peers. Because drinking tends to be a social activity and is done in groups of peers, alcohol consumption should be subject to intragroup processes, pressure to conform ( Asch, 1956 ; Cialdini and Trost, 1998 ; Deutsch and Gerard, 1955 ), and the desire to increase group identification ( Mackie et al., 1990 ; Tajfel, 1982 ; Tajfel and Turner, 1986 ). People’s drinking behavior and perceptions about existing norms regarding alcohol consumption ( Clapp and McDonnell, 2000 ; Perkins, 1985 , 1997 ; Perkins and Wechsler, 1996 ) should be influenced.

Considerable evidence has shown, however, that across contexts and demographic groups, people misjudge the actual normative drinking tendencies and perceive that peers consume alcohol more frequently and in greater volume then they actually do ( Berkowitz, 2004a , 2004b ; Borsari and Carey, 2003 ; Perkins, 2002 ). Given that group members’ behavior can be influenced by the perceived norms of their group, and people often overestimate normative drinking, it follows that excessive drinking may be engendered, at least in part, by pressure or motivation to match behavior to perceived drinking norms that are inaccurately high. Social norms interventions attempt to reduce drinking by correcting this misperception and communicating to drinkers that normative consumption is actually lower than they perceive. The desired consequence of more accurate drinking perceptions is that normative influence should facilitate more moderate alcohol consumption.

Efficacy of social norms interventions

Social norms interventions typically rely on mass-marketing campaigns or brief personal interviews that include normative feedback to correct inaccurate perceived norms. Many empirical evaluations of these approaches have been supportive. For example, in a pre–post study, Mattern and Neighbors (2004) observed reductions both in perceived norms and in reported drinking following a social norms intervention conducted in a college residence hall. Comparable results were observed following a multimodal social norms campaign targeting student athletes ( Perkins and Craig, 2006 ). Another campus-wide social norms program ( Foss et al., 2000 ) used breath-test surveys of students (not merely self-reported drinking) to document reductions in student alcohol consumption following the campaign.

Experimental and quasi-experimental evaluations also support the efficacy of social norms approaches. One study of a multicampus mass-marketing campaign ( DeJong et al., 2006 ) found a small but statistically significant reduction in self-reported drinking at schools that used social norms strategies (relative to controls), and another statewide evaluation found significant changes in perceived norms and self-reported driving after drinking in counties that received the intervention relative to counties that did not ( Perkins et al., 2010 ). Importantly, a recent meta-analysis of 22 studies ( Moreira et al., 2009 ) that met the criteria of random assignment to experimental condition found that social norms messaging (relative to a control) reduced short-and moderate-term heavy drinking and alcohol-related problems.

There are limitations to what the literature reports about the validity of the social norms theory. This is true both for studies that have demonstrated reductions in alcohol misuse and for studies that have failed to find significant effects ( Carter and Kahnweiler, 2000 ; Clapp et al., 2001, 2003 ; Granfield, 2002 ; Wechsler et al., 2003 ; Werch et al., 2000 ). First, social norms programs could fail to show program efficacy for several reasons. Null findings could be attributable to limitations of the social norms theory or to inadequate implementation of the theory in natural drinking environments ( Berkowitz, 2004b ; Thombs et al., 2004 ). Unfortunately, little of the social norms research has collected information on mediating variables that would allow more direct testing of the proposed theory; neither have studies typically collected information on process data that might inform the fidelity of the study operations.

Second, even studies that attributed significant reductions in alcohol misuse to social norms interventions may suffer from plausible alternative explanations. The research to date has relied almost exclusively on drinking data derived from self-reported consumption, which may be subject to demand characteristics or socially desirable responding. This is particularly plausible regarding social norms interventions, given that they involve informing heavy drinkers that their behavior is not normal.

The research described herein addresses some limitations of the literature. The research involved experimental tests of social norms theory in a natural drinking environment using objective measures of alcohol consumption (e.g., breath-based blood alcohol concentrations [BrACs] expressed in units of grams/deciliter). We did not evaluate a program; rather, we conducted an experiment designed to test whether the theory underlying social norms campaigns is supported.

Potential moderators of social norms feedback

In addition to testing whether providing social norms feedback has affected perceived norms and subsequent drinking behavior, research has examined whether variables might moderate the effect of this feedback (e.g., Borsari and Carey, 2003 ; Thombs et al., 2004 ). Factors that increase the persuasiveness of the social norms feedback should have a greater effect on behavior.

For a normative message to change a recipient’s initial perceived norms, recipients must first attend to the normative message and become aware of the discrepancy between their beliefs and the true norm. Therefore, in our study we manipulated the salience of the social norms feedback under the hypothesis that social norms messages that stress this discrepancy should be more effective in correcting perceived norms. Second, social norms messages that are personally relevant to recipients may be more effective. Thus, we manipulated whether participants received accurate drinking norm feedback in reference to people in general or whether they received norms regarding specific gender and racial/ethnic groups. We hypothesized that the normative information would be more effective if it referred to a group with whom participants more closely identified (e.g., LaBrie et al., 2010 ; Lewis and Neighbors, 2007 ; Neighbors et al., 2010 ). Third, participants might not believe social norms information if it is inconsistent with their actual experiences. Therefore, we instructed some participants to seek out examples that confirmed the social norms information (whereas other participants received no such instructions). We hypothesized that participants who sought out specific examples consistent with the social norms information would be more influenced by the feedback that they received.

Data were collected using the portal survey method (e.g., Johnson and Clapp, 2011 ; Lange et al., 2006 ; Voas et al., 2006 ) used at the San Ysidro border crossing between San Diego, CA, and Tijuana, Mexico. The portal survey method requires access to the “portal” of a high-risk drinking environment where participants can be sampled before entering the drinking environment and then resampled as they exit the environment. We have used the border crossing between San Diego and Tijuana in several studies because Tijuana represents a high-risk drinking environment, and the flow of pedestrian traffic through the border-crossing portal is well controlled, allowing for easy sampling of participants.

Participant recruitment

Data were collected on Fridays and Saturdays 1 weekend per month by two or three small teams of data collectors. During the “entry” portion of the study, participants were recruited just before they crossed into Mexico between 10 P.M. and 1 A.M. on survey nights. Groups of participants were randomly sampled from the flow of foot traffic as they approached the border crossing and were invited to take part in the study.These potential participants were offered a $20 retail store gift card (to be given after completing the “exit” portion of the survey) for taking part in the study.

Entry survey

Once they provided informed consent, the participants completed a questionnaire. Items included were demographics, recent drinking quantity and frequency, how they arrived at the border crossing (e.g., driver, passenger), and their drinking intentions for the evening (i.e., not to drink, to get buzzed, to get drunk, or to get very drunk). In the last part of the questionnaire, participants provided information on their perceived social norms about the typical (normative) drinking levels of persons who visit the bars and clubs of Tijuana. Participants were asked to describe their perceptions of (a) the percentage who do not drink while in Tijuana, (b) the percentage who drink enough to become legally drunk while in Tijuana, (c) the average number of drinks typically consumed while in Tijuana, and (d) the percentage who consume 10 or more drinks while in Tijuana.

After completing the questionnaire, participant groups were randomly assigned to an experimental condition and were given social norms information corresponding to the condition to which they were assigned. Following receipt of the experimental instructions, participants were asked to provide a breath sample to measure alcohol consumption at entry. Finally, each participant was given a hospital-style identification bracelet bearing a unique identification number, which we used to match entry data with exit data for each participant.

Experimental variables

Participant groups were randomly assigned to one of nine experimental conditions, only two or three of which were randomly selected to be administered each survey night. Entire participant groups were assigned to the same experimental condition based on which survey team recruited them. Individuals in the same group were never assigned to different conditions.

One condition served as a nonintervention control, and the participants in the eight remaining conditions received accurate normative information. The method of providing social norms feedback in the remaining eight conditions varied via the manipulation of three independent variables: (a) the salience of the discrepancy between actual and perceived norms, (b) the specificity of the reference group to which the normative information applied, and (c) the believability of information (whether participants received experimental instructions to seek out examples of nondrinkers and light drinkers in Tijuana).

Social norms feedback information was given to participants in a single, brief dose via graphs generated on a tablet-style computer. A separate graph was used to show the actual normative drinking levels for each of the four drinking items (percentage nondrinkers, percentage legally drunk, average standard drinks consumed, percentage consuming ≥10 drinks). In the low-salience condition, participants simply saw four graphs that displayed the actual, accurate values (based on pilot data) for each of the four items. In the high-salience condition, each graph not only revealed the accurate norms but also included the mean perceived norms (also based on pilot data), thus stressing the discrepancy between actual drinking norms and the public’s perceptions.

In the low-specificity condition, the social norms feedback given to participants was based on the means of the total sample of pilot study subjects (collected previously at the Tijuana border crossing). In the moderate-specificity condition, participants were given information on the perceived norms for each race/ethnicity and gender combination against the actual normative value for each item. This allowed participants to see that even people similar to themselves held inaccurate norms. Finally, in the high-specificity condition, we provided both the actual norms and perceived norms for each Sex × Race/Ethnicity combination.

These first two experimental variables were crossed to create four social norms feedback conditions: (a) low-salience, low-specificity; (b) high-salience, low-specificity; (c) high-salience, moderate-specificity; and (d) high-salience, high-specificity reference groups. Figure 1 presents social norms feedback graphs that were shown to participants under each condition. During the study, participants saw one graph for each of the four drinking items, but because of space issues, we provide graphs for only one of the four drinking items.

Percentage of people who visit Tijuana and do not drink alcohol: Experimental social norms feedback provided to participants

Exit survey

At 1:00 a.m. , data collectors relocated to the outside of the border-processing center where individuals pass to reenter the United States. These data collectors attempted to resample the same participants (identified by hospital-style identification bracelets) to complete the second part of the study. Resampled participants were interviewed and asked about drinking behavior while in Tijuana, means of transportation home, and their current perceived norms (on the four drinking items). Interviewers also asked participants to provide an anonymous breath sample. Finally, data collectors removed participants’ identification bracelets and gave each a $20 gift card.

Data collection occurred over 52 weekend nights from January 2008 to August 2009, during which 2,972 individuals (from 935 peer groups) were contacted and invited to participate. Of those contacted, 128 (4.3%) refused to participate, and 499 (16.8%) completed the entry survey but did not return for the exit survey. A total of 2,218 participants (74.6% of those recruited) completed both the entry and exit surveys and provided entry and exit breath tests. The proportion who did not return for the exit survey did not vary by conditions ( p = .84).

Our working sample was 61.8% male, 59.8% of whom were younger than age 21 and 20.5% of whom drove to the border crossing. The majority were Hispanic (69.6%), with a minority of White (non-Hispanic) (9.7%), racial minority (non-Hispanic) (14.0%), and multiracial or “other” (6.8%) participants. Most participants (80.0%) entered with a blood alcohol concentration of 0 (mean entry BrAC = .010).

Social norms

Participants were asked to provide their perceptions of normative drinking levels during the entry survey (before the experimental manipulations) and again during the exit survey. Table 1 shows the (descriptive) actual and perceived norms of our sample. The table shows that participants’ perceived norms overestimated the actual drinking in Tijuana. The last column shows the social norms feedback (based on pilot data) that was provided to participants. The feedback closely matched the actual normative behaviors measured during the study.

Actual and perceived social norms

Notes: BrAC = breath alcohol concentration.

Analytic goals

Our analyses addressed three broad goals relevant to understanding the efficacy and limitations of providing accurate social norms information to change behavior. The first goal was to test the hypothesis that changes in perceived drinking norms were correlated with changes in actual drinking. Specifically, we hypothesized that persons whose perceived drinking norms changed toward accuracy (for whatever reason) would drink relatively less. Because we measured perceived norms during entry as well as exit from Tijuana, we could compute the change in perceived norms between the two time points.

The second goal was to test whether the experimental interventions significantly affected participants’ perceived drinking norms. This involved comparing control participants with participants who received social norms feedback on degree of change in perceived drinking norms. Additionally, this goal involved comparing mean “changein-perceived-norm” scores among the different salience, reference-group specificity, and believability conditions. Finally, our third goal was to test whether the social norms interventions significantly reduced participants’ BrACs and, specifically, whether this effect was mediated by the change in perceived norms.

Analytic approach

Our portal survey methodology produces data with a nested structure because individual participants are nested within peer groups, thus making the assumption of independent observations untenable. Accordingly, the research goals were addressed using generalized linear mixed modeling (SAS 9.2, SAS Institute, Inc., Cary, NC), with peer group modeled as a random variable.

Hypothesis 1: Changes in perceived drinking norms will predict drinking behavior.

We computed the value for each of the four drinking items, scored so that larger (positive) values reflected reductions in perceptions of excessive drinking. We subjected the four difference scores to a principal components analysis and retained a factor score for the first factor (the only factor with an eigenvalue > 1). This factor score was a standardized composite measure where a 0 indicated mean average change in perceptions. Increasingly positive factor scores indicated that participants’ drinking perceptions changed toward greater accuracy. Negative scores reflected minimal change toward accuracy or change away from accuracy.

We predicted participants’ exit BrACs from their changein-perceived-norms factor scores. Our analyses also included entry BrAC, sex, race/ethnicity, age, drinking intentions, driver status, and drinking history as covariates. Drinking history was represented as a factor score reduced from three quantity and frequency drinking items. The primary finding of this analysis was a statistically significant main effect of the change-in-perceived-norms factor score on exit BrACs, F (1, 610) = 5.6, p < .05. The slope was negative (un-standardized B = –.003) and significant above and beyond the covariates. To the extent that participants’ perceived norms changed in the direction of less excessive drinking (greater accuracy), their own alcohol consumption was significantly lower. Every change in perceived norms of 1 SD was associated with a reduction in BrAC of .003; for example, change from two standard deviations above the mean to two below the mean was associated with a reduction in BrAC of .012.

Hypothesis 2: Social norms feedback will influence participants’ perceived drinking norms.

These analyses included tests of the experimental variables’ impact on participants’ perceived drinking norms. The standardized change-inperceived-norms variable served as the dependent measure, although it was rescaled for interpretability when making sense of results.

The analyses focused on five main tests: social norms versus control, low versus high salience, low versus moderate specificity, moderate versus high specificity, and confirmation instructions versus no instructions. These tests were conducted via a series of planned contrasts within an omnibus model, F (8, 635) = 5.4, p < .01, that included the nine-level experimental condition variable as the effect of interest. The omnibus model included sex, race/ethnicity, driver status, age category, drinking history, and drinking plans as control variables, as well as peer group as a random variable. Because of space limitations, we did not interpret the covariates.

The first contrast, which compared the mean change-inperceived-norms score for the control condition with the eight social norms feedback conditions, was statistically significant, t (635) = 4.2, p < .01. The second contrast compared the low-salience, low-specificity conditions with the high-salience, low-specificity conditions. The difference in mean change-inperceived-norms scores between the two was not significant ( p = .82). The third contrast, comparing the high-salience, low-specificity conditions with the high-salience, moderate-specificity conditions, also was not significant ( p = .32). The fourth contrast, moderate-reference-group specificity versus high-reference-group specificity (both high salience), was statistically significant, t (635) = 2.16, p < .05. Finally, the fifth contrast, which compared the four no-instruction conditions with the four confirmation-instruction conditions, was statistically significant, t (635) = –2.22, p < .05.

Our change-in-perceived-norms factor score was standardized so that a 0 indicated mean change in norms, not no change. To make this variable easier to interpret, we rescaled the factor score to the same metric as the change in perceived norms regarding the percentage of Tijuana visitors who consumed 10 or more drinks ( M = 14.5, SD = 26.8). Accordingly, a factor score of 1.0 was rescaled to reflect a change in perceived norms by 41.3 percentage points (toward perceptions of less drinking). A factor score of –1.0 meant that perceived norms changed by 12.3 percentage points toward perceptions of more perceived drinking.

Table 2 reveals the model-estimated change-in-perceived-norms score (rescaled) for each of nine experimental conditions. Specific hypothesis were tested by contrasting the means of different combinations of these nine groups. For example, a contrast described as (A + B vs. C + D) would test the null hypothesis that the mean score of participants assigned (in aggregate) to conditions A and B was equal to the mean score of participants assigned (in aggregate) to conditions C and D, with + signs indicating which groups of participants were combined before scores were averaged. The statistically significant contrast between the mean of the control condition participants and the mean of participants assigned to the eight social norms conditions (A vs. B through I) reflects a change of 16.1 percentage points (toward less drinking) by those who received the feedback relative to only 4.1 percentage points by the control.

Model-estimated change-in-perceived-norm scores per experimental condition

The significant contrast between the moderate-specificity and high-specificity reference group conditions (F + G vs. H + I) was not in the intended direction. Participants in the high-specificity conditions changed their perceived norms 5.3 percentage points less than did those in the moderate condition. We hypothesize that more specific information would be more persuasive.

Finally, the significant confirmation-instruction contrast (B + D + F + H vs. C + E + G + I) showed that participants who were instructed to seek examples of nondrinkers while in Tijuana changed their perceived norms 3.8 percentage points more (toward more accuracy) than did groups that did not receive these instructions. An exploratory analysis examined the effect of the confirmation instructions within each of the four experimental blocks. The analysis found that confirmation instructions had a significant impact on change in norms only in the high-salience conditions with low or moderate specificity ( p < .01) but not in the low-salience condition ( p = .51) or high-salience, high-specificity condition ( p = .71) (see Table 2 for means).

Hypothesis 3: Providing accurate feedback on drinking norms will reduce alcohol consumption.

If providing participants with accurate drinking norms information caused significant changes in perceived norms, and if changes in perceived norms were associated with lower drinking, it follows that providing social norms feedback should significantly decrease BrACs of participants returning from Tijuana. The largest effect of the experimental variables on perceived norms involved the omnibus contrast between the control group and all of the social norms feedback conditions combined. If any comparison among conditions produced significant differences in exit BrAC, it seemed reasonable that this comparison would be most likely. The analysis of control versus social norms feedback conditions on exit BrACs controlled for demographics and entry BrAC. The planned contrast, however, failed to find a significant difference between the conditions on exit BrACs ( p = .42). We excluded nondrinkers from the analysis because it seemed that they would be unaffected by the intervention even if drinkers were affected, but this effect, too, was non-significant ( p = .50). We similarly conducted tests including only participants with BrACs of .02 and .04 and higher, and in no case did the effect approach statistical significance. Thus, the chain between the norms feedback, changes in perceived norms, and exit BrAC could not be confirmed. No other planned contrasts (of the salience, reference-group specificity, and confirmation-instruction conditions) produced a statistically significant difference in exit BrACs.

For descriptive purposes, in Table 3 , we present raw data (not model adjusted) depicting perceived norms at entrance, perceived norms at exit, and actual drinking behavior separately for control participants and participants who received social norms feedback. Data for the experimental group exclude cases from the high-specificity conditions. The data suggest more accurate pre-versus post-change for participants in the social norms condition but few differences in behavior (expect, perhaps, for the percentage consuming ≥10 drinks).

Perceived and actual norms by three conditions

This research experimentally tested the effect of providing accurate drinking norm feedback on perceived norms and on objectively measured alcohol consumption in the context of a natural drinking environment. Although the research occurred in the field, this study was a test of theory, not an evaluation of an intervention. We argue, however, that field experiments are an important but overlooked step in the translation of basic research to practical scientific applications. By administering social norms feedback via brief interviews in a real-world environment, we hoped to provide definitive support for the theoretical mechanisms posited to underlie social norms interventions.

A fundamental aspect of social norms theory is that perceived norms influence behavior and that changing norms should change behavior. Our research demonstrated a significant relationship between changes in perceived norms and alcohol consumption (measured from breath samples). Although this relationship cannot be shown to be causal, the presence of a correlation is necessary for a causal relationship to exist, and the lack of significant correlation would be inconsistent with the theory.

The experimental manipulation of social norms feedback significantly affected participants’ perceived norms. Those who received accurate normative information later changed their norms, on average, to be more accurate (toward perceptions of less excessive drinking). Our efforts to moderate the effect of the social norms feedback by increasing the salience and specificity of the norms message were unsuccessful. In fact, providing participants with highly detailed information appeared to reduce the effect. Our only explanation for this is that the high-salience, high-specificity condition offered too much information that was too complex to be processed by participants in the field.

The confirmation instruction manipulation did moderate the effectiveness of the social norms intervention. Participants who were instructed to seek examples of nondrinkers and light drinkers in the bars and clubs they visited experienced significantly more change in their norms than did participants who did not receive such instructions. This finding is consistent with the idea that the efficacy of social norms feedback may be undermined by individual experience. Encouraging individuals to find firsthand experiential evidence that confirms that normative information appeared to increase its effectiveness. Future research will need to explore this effect more carefully.

Theory, effect sizes, and practical significance

Our desire was to demonstrate via mediational analysis that our experimental manipulations reduced drinking by changing perceived norms. However, direct tests of social norms feedback failed to predict significant changes in participants’ BrACs. The relationship between changes in perceived norms and exit BrACs, although statistically significant, was quite weak. This weakness reduced the likelihood of detecting significant mediation.

It is important to note, however, that the social norms communication administered via our experimental manipulation was quite brief. Successful social norms marketing programs ( DeJong et al., 2006 ) typically involve considerably greater exposure to the norms message (i.e., years) than our study did, and even successful social norms feedback provided through brief motivational interviews ( Carey et al., 2006 ; Murphy et al., 2001 ) involves 30 or more minutes of interaction. Furthermore, these face-to-face normative-feedback efforts typically are personalized for the recipient. Nevertheless, even the 3-minute exposure of general social norm information in our study appeared to influence participants’ perceived norms. Second, participants were sampled en route to a drinking environment. Even with a relatively open border between San Diego and Tijuana, trips into Mexico require some effort and often entail waiting in line at Customs to reenter the United States. Thus, these trips are purposeful, and people who cross into Mexico and plan to drink alcohol likely placed some importance on that behavior. A person’s drinking intentions strongly predict his or her actual alcohol consumption ( Lange et al., 2006 ). Therefore, it is not unexpected that, in this context, changes in norms only weakly predicted changes in drinking behavior. It might be different if the research were dealing with persons whose drinking decisions were more discretionary. Accordingly, social norms strategies might more effectively shape an individual’s drinking intentions.

The research does underscore the importance of message recipients actively processing the normative information. Moreira et al.’s (2009) review suggests that norms-feedback interventions tend to be more effective when participants are engaged in the information presentation, either during face-to-face interviews or over the web (where, presumably, if participants were willing to go to the website to access the information, they had some motivation to attend to it). In our study, the instruction to seek out nondrinkers and light drinkers—which was designed to make the normative information more experiential (thus personal) and less didactic—did influence perceived norms and is consistent with evidence regarding the value of personal engagement. We hypothesized that directly manipulating the salience of the inaccuracy between perceived and actual norms would facilitate engagement with the information, but we have no evidence that our operation had the psychological effect that we intended.

Future endeavors

Field experiments provide opportunities to conduct basic tests of theory in environments with high ecological validity. Such studies may aid in the transition of science to practice by demonstrating that theoretical relationships persist in realworld environments and are not enveloped by countervailing factors and influences. This research found that, even in natural drinking environments, changes in perceived norms correlate with reduced drinking levels and that it is possible to change people’s normative perceptions by providing them with information on actual drinking norms.

One important question is “At what point in the drinking-decision process does the perceived norm influence behavior?” Whether individuals make drinking plans with norms in mind or whether they manage their behavior on the fly and rely on normative information for a given context suggests different obstacles for implementing successful social norms strategies. If perceived norms typically influence the construction of drinking intentions and guide behavior outside of the drinking environment, then the value of confirmation-type instructions in a standard social norms marketing campaign is less clear.

Another larger, theoretical question might be “What is the efficacy of social norms information that has been provided to a particular population?” Presumably, perceived norms operate like any other cognition and follow general rules of cognitive functioning (e.g., accessibility, relevance). Thus, even if a social norms campaign successfully communicates accurate drinking information to an audience, and even if this information is stored in explicit memory, there is no guarantee that this information will be accessible at the times when it has the best opportunity to influence behavior (e.g., in drinking environments or when drinking plans are being formed). Thus, it might be worth studying the accessibility of normative information at different times among populations that have been exposed to accurate social norms feedback. Given that alcohol consumption has been found to influence what cognitions are accessible in memory (e.g., the alcohol myopia model; Steele and Josephs, 1990 ), the primarily obstacle to changing behavior via normative feedback may not be how the information is communicated but whether environmental features reduce its accessibility and consequent use.

Social norms marketing strategies have been widely used to reduce unhealthy behaviors (e.g., excessive drinking, smoking, drunk driving), but the results have been inconsistent. Even when evaluations of large-scale social norms marketing programs demonstrate significant changes in behavior, it is often unclear whether these changes were mediated by perceived norms or whether it was some other mechanism. There still is considerable opportunity, however, to conduct basic research on the processes through which normative feedback might influence behavior. Focused experimental studies may eventually inform public health practitioners regarding how to develop effective social norms programs that reliably reduce unhealthy behavior.

This research was supported by National Institute on Alcohol Abuse and Alcoholism Grant R01 AA015716.

• Asch SE. Studies of independence and conformity: A minority of one against a unanimous majority. Psychological Monographs. 1956; 70 (9):1–70. [ Google Scholar ]
• Berkowitz AD. An overview of the social norms approach. In: Lederman L, Stewart L, Goodhart F, Laitman L, editors. Changing the culture of college drinking: A socially situated prevention campaign. Mulgrave, Australia: Hampton Press; 2004a. [ Google Scholar ]
• Berkowitz AD. The social norms approach: Theory, research, and annotated bibliography. Trumansburg, NY: Author; 2004b. Retrieved from http://www.alanberkowitz.com/articles/social_norms.pdf . [ Google Scholar ]
• Borsari B, Carey KB. Descriptive and injunctive norms in college drinking: A meta-analytic integration. Journal of Studies on Alcohol. 2003; 64 :331–341. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Carey KB, Carey MP, Maisto SA, Henson JM. Brief motivational interventions for heavy college drinkers: A randomized controlled trial. Journal of Consulting and Clinical Psychology. 2006; 74 :943–954. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Carter CA, Kahnweiler WM. The efficacy of the social norms approach to substance abuse prevention applied to fraternity men. Journal of American College Health. 2000; 49 :66–71. [ PubMed ] [ Google Scholar ]
• Cialdini R, Trost M. Social influence:Social norms, conformity and compliance. In: Gilbert DT, Fiske ST, Lindzey G, editors. Handbook of social psychology. 4th ed. Vol. 2. New York, NY: McGraw-Hill; 1998. pp. 151–192. [ Google Scholar ]
• Clapp JD, Lange JE, Russell C, Shillington A, Voas RB. A failed norms social marketing campaign. Journal of Studies on Alcohol. 2003; 64 :409–414. [ PubMed ] [ Google Scholar ]
• Clapp J, McDonnell AL. The relationship of perceptions of alcohol promotion and peer drinking norms to alcohol problems reported by college students. Journal of College Student Development. 2000; 41 :19–26. [ Google Scholar ]
• Clapp JD, Russell C, DeJong W. Done 4 did zip: Evaluating a failed social norms marketing campaign. Anaheim, CA: Presented at the Fourth National Conference on the Social Norms Model; 2001, July. [ Google Scholar ]
• DeJong W, Schneider SK, Towvim LG, Murphy MJ, Doerr EE, Simonsen NR, Scribner RA. A multisite randomized trial of social norms marketing campaigns to reduce college student drinking. Journal of Studies on Alcohol. 2006; 67 :868–879. [ PubMed ] [ Google Scholar ]
• Deutsch M, Gerard HB. A study of normative and informational social influences upon individual judgment. Journal of Abnormal and Social Psychology. 1955; 51 :629–636. [ PubMed ] [ Google Scholar ]
• Foss RD, Marchetti LJ, Holladay KA. Development and evaluation of a comprehensive program to reduce drinking and impaired driving among college students (DOT HS 809 396) Washington, DC: National Highway Traffic Safety Administration; 2000. Retrieved from http://www.nhtsa.dot.gov/people/injury/research/ReduceCollege/tech_dog.htm . [ Google Scholar ]
• Granfield R. Believe it or not: Examining to the emergence of new drinking norms in college. Journal of Alcohol and Drug Education. 2002; 47 :18–31. [ Google Scholar ]
• Hingson RW, Zha W, Weitzman ER. Magnitude of and trends in alcohol-related mortality and morbidity among U.S. college students ages 18-24: 1998-2005. Journal of Studies on Alcohol and Drugs, Supplement No. 2009; 16 :12–20. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Johnson MB, Clapp JD. Impact of providing drinkers with “know your limit” information on drinking and driving: A field experiment. Journal of Studies on Alcohol and Drugs. 2011; 72 :79–85. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Keeling RP. Social norms research in college health. Journal of American College Health. 2000; 49 :53–56. [ PubMed ] [ Google Scholar ]
• LaBrie JW, Hummer JF, Neighbors C, Larimer ME. Whose opinion matters? The relationship between injunctive norms and alcohol consequences in college students. Addictive Behaviors. 2010; 35 :343–349. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Lange JE, Reed MB, Johnson MB, Voas RB. The efficacy of experimental interventions designed to reduce drinking among designated drivers. Journal of Studies on Alcohol. 2006; 67 :261–268. [ PubMed ] [ Google Scholar ]
• Lewis MA, Neighbors C. Optimizing personalized normative feedback: The use of gender-specific referents. Journal of Studies on Alcohol and Drugs. 2007; 68 :228–237. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Mackie DM, Worth LT, Asuncion AG. Processing of persuasive in-group messages. Journal of Personality and Social Psychology. 1990; 58 :812–822. [ PubMed ] [ Google Scholar ]
• Mattern JL, Neighbors C. Social norms campaigns: examining the relationship between changes in perceived norms and changes in drinking levels. Journal of Studies on Alcohol. 2004; 65 :489–493. [ PubMed ] [ Google Scholar ]
• Moreira T, Smith LA, Foxcroft DR. Social norms interventions to reduce alcohol misuse in university or college students. Cochrane Database of Systematic Reviews, Issue. 2009; 3 :Article No. CD006748. [ PubMed ] [ Google Scholar ]
• Murphy JG, Duchnick JJ, Vuchinich RE, Davison JW, Karg RS, Olson AM, Coffey TT. Relative efficacy of a brief motivational intervention for college student drinkers. Psychology of Addictive Behaviors. 2001; 15 :373–379. [ PubMed ] [ Google Scholar ]
• National Institute on Alcohol Abuse and Alcoholism. A call to action: Changing the culture of drinking at U.S. colleges (NIH Publication No. 02–5010) Bethesda, MD: National Institutes of Health; 2002. Retrieved from http://www.collegedrinkingprevention.gov/media/TaskForceReport.pdf . [ Google Scholar ]
• Neighbors C, LaBrie JW, Hummer JF, Lewis MA, Lee CM, Desai S, Larimer ME. Group identification as a moderator of the relationship between perceived social norms and alcohol consumption. Psychology of Addictive Behaviors. 2010; 24 :522–528. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Perkins HW. Religious traditions, parents, and peers as determinants of alcohol and drug use among college students. Review of Religious Research. 1985; 27 :15–31. [ Google Scholar ]
• Perkins HW. Designing alcohol and other drug prevention programs in higher education: Bringing theory into practice (pp. 177–206) Newton, MA: The Higher Education Center for Alcohol and Other Drug Prevention; 1997. College student misperceptions of alcohol and other drug norms among peers: Exploring causes, consequences, and implications for prevention programs. Retrieved from http://alcohol.hws.edu/perkinstheorychapter.pdf . [ Google Scholar ]
• Perkins HW. Social norms and the prevention of alcohol misuse in collegiate contexts. Journal of Studies on Alcohol, Supplement. 2002; 14 :164–172. [ PubMed ] [ Google Scholar ]
• Perkins HW, Craig DW. A successful social norms campaign to reduce alcohol misuse among college student-athletes. Journal of Studies on Alcohol. 2006; 67 :880–889. [ PubMed ] [ Google Scholar ]
• Perkins HW, Linkenbach JW, Lewis MA, Neighbors C. Effectiveness of social norms media marketing in reducing drinking and driving: A statewide campaign. Addictive Behaviors. 2010; 35 :866–874. [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Perkins HW, Wechsler H. Variation in perceived college drinking norms and its impact on alcohol abuse: A nationwide study. Journal of Drug Issues. 1996; 26 :961–974. [ Google Scholar ]
• Steele CM, Josephs RA. Alcohol myopia. Its prized and dangerous effects. The American Psychologist. 1990; 45 :921–933. [ PubMed ] [ Google Scholar ]
• Substance Abuse and Mental Health Services Administration . Results from the 2008 National Survey on Drug Use and Health: National findings (NSDUH Series H-36, HHS Publication No. SMA 09–4434) Rockville, MD: Office of Applied Studies; 2009. Retrieved from http://oas.samhsa.gov/nsduh/2k8nsduh/2k8Results.cfm . [ Google Scholar ]
• Tajfel H. England. Cambridge University Press; 1982. Social identity and intergroup behavior. Cambridge. [ Google Scholar ]
• Tajfel H, Turner JC. The social identity theory of inter-group behavior. In: Worchel S, Austin LW, editors. Psychology of intergroup relations. Chicago, IL: Nelson-Hall; 1986. [ Google Scholar ]
• Thombs DL, Dotterer S, Olds RS, Sharp KE, Raub CG. A close look at why one social norms campaign did not reduce student drinking. Journal of American College Health. 2004; 53 :61–68. [ PubMed ] [ Google Scholar ]
• Voas RB, Furr-Holden D, Lauer E, Bright K, Johnson MB, Miller B. Portal surveys of time-out drinking locations: A tool for studying binge drinking and AOD use. Evaluation Review. 2006; 30 :44–65. [ PubMed ] [ Google Scholar ]
• Wechsler H, Nelson TE, Lee JE, Seibring M, Lewis C, Keeling RP. Perception and reality: A national evaluation of social norms marketing interventions to reduce college students’ heavy alcohol use. Journal of Studies on Alcohol. 2003; 64 :484–494. [ PubMed ] [ Google Scholar ]
• Werch CE, Pappas DM, Carlson JM, DiClemente CC, Chally PS, Sinder JA. Results of a social norm intervention to prevent binge drinking among first-year residential college students. Journal of American College Health. 2000; 49 :85–92. [ PubMed ] [ Google Scholar ]

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• News & Views
• Published: 17 December 2018

SOCIAL PSYCHOLOGY

Social norms and how they impact behaviour

• Katherine J. Reynolds 1  

Nature Human Behaviour volume  3 ,  pages 14–15 ( 2019 ) Cite this article

2102 Accesses

15 Citations

14 Altmetric

Metrics details

• Human behaviour

There is wide interest in the social norms construct across psychology, economics, law and social marketing. Now a study investigates an important missing piece in the social norms’ puzzle: what is the underlying process that explains how norms impact behaviour? The answer: self–other similarity (self-categorization) and internalization.

This is a preview of subscription content, access via your institution

Relevant articles

Open Access articles citing this article.

Co-evolution of conditional cooperation and social norm

• Balaraju Battu

Scientific Reports Open Access 03 October 2023

Access options

Access Nature and 54 other Nature Portfolio journals

Get Nature+, our best-value online-access subscription

$29.99 / 30 days

cancel any time

Subscribe to this journal

Receive 12 digital issues and online access to articles

$119.00 per year

only $9.92 per issue

Buy this article

• Purchase on SpringerLink
• Instant access to full article PDF

Prices may be subject to local taxes which are calculated during checkout

Hallsworth, M. et al. Lancet 387 , 1743–1752 (2016).

Article   Google Scholar  

Nolan, J. M., Schultz, P. W., Cialdini, R. B., Goldstein, N. J. & Griskevicius, V. Pers. Soc. Psychol. Bull. 34 , 913–923 (2008).

Reynolds, K. J., Subasic, E. & Tindall, K. Soc. Pers. Psychol. Compass 9 , 45–56 (2015).

Cialdini, R. B. Pre-suasion: A Revolutionary Way to Influence and Persuade 1st edn (Simon & Schuster, New York, 2016).

Turner, J. C. & Reynolds, K. J. in Handbook of Theories in Social Psychology (eds Van Lange, P., Kruglanski, A. & Higgins, T.) 399–417 (Sage, Thousand Oaks, 2012).

Reynolds, K. J. & Branscombe, N. R. in The Psychology of Change: Life Contexts, Experiences, and Identities (eds Reynolds, K. J. & Branscombe, N. R.) 265–281 (Psychology Press, London, 2015).

Pryor, C., Perfors, A. & Howe, P. D. L. Nat. Hum. Behav . https://doi.org/s41562-018-0489-y (2018).

Turner, J. C. Eur. J. Soc. Psychol. 35 , 1–22 (2005).

Nyborg, K. et al. Science 354 , 42–43 (2016).

Article   CAS   Google Scholar  

Download references

Author information

Authors and affiliations.

Research School of Psychology, The Australian National University, Canberra, Australian Capital Territory, Australia

Katherine J. Reynolds

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Katherine J. Reynolds .

Ethics declarations

Competing interests.

The author declares no competing interests.

Rights and permissions

Reprints and permissions

About this article

Cite this article.

Reynolds, K.J. Social norms and how they impact behaviour. Nat Hum Behav 3 , 14–15 (2019). https://doi.org/10.1038/s41562-018-0498-x

Download citation

Published : 17 December 2018

Issue Date : January 2019

DOI : https://doi.org/10.1038/s41562-018-0498-x

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Scientific Reports (2023)

Trust in science, social consensus and vaccine confidence

• Patrick Sturgis
• Ian Brunton-Smith
• Jonathan Jackson

Nature Human Behaviour (2021)

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

• Faculty Resource Center
• Biochemistry
• Bioengineering
• Cancer Research
• Developmental Biology
• Engineering
• Environment
• Immunology and Infection
• Neuroscience
• JoVE Journal
• JoVE Encyclopedia of Experiments
• JoVE Chrome Extension
• Environmental Sciences
• Pharmacology
• JoVE Science Education
• JoVE Lab Manual
• JoVE Business
• Videos Mapped to your Course
• High Schools
• Videos Mapped to Your Course

Chapter 4: Understanding and Influencing Others

Back to chapter, the stanford prison experiment, previous video 4.2: fundamental attribution error, next video 4.4: conformity.

There’s a dark side to the summer of 1971. Police “arrested” a number of college students who responded to an ad seeking volunteers for a psychological study of prison life.

Little did they know, in the basement of the Stanford Psychology Department, they were about to embark upon one of the most famous and controversial psychological studies: The Stanford Prison Experiment, led by Philip Zimbardo and colleagues.

After filling out an informational questionnaire, participants were randomly assigned to one of two groups: Prisoners—who were realistically picked up and booked by police—or Guards, who were in direct control of the inmates.

Depending on the assignment, they were dressed to fit their role —a set of expectations defining how those in the social position should behave.

For example, each guard was given a nightstick and whistle, along with mirrored aviator glasses, to emphasize their status and authority.

In addition, two researchers were present and oversaw the day-to-day operations: the lead experimenter, Zimbardo, acted as the prison superintendent, while another researcher, David Jaffe, was the Warden.

During one orientation session for the guards, Zimbardo vaguely outlined behaviors that they should conform with, like limiting the prisoners’ freedom and using their power to evoke fear and dominate non-violently.

In a follow-up session, the Warden provided more explicit directions to encourage the creation of a prison environment, for the collective good of the experiment.

With the correctional authorities in place, the prisoners entered the mock jail, which contained three cells and a closet for solitary confinement. They were stripped down, immediately degraded, and blindfolded to confuse them. Also, as part of their assigned role, they were outfitted in numbered gowns and nylon-stocking caps to depersonalize them.

On the second day, the prisoners started to rebel by blockading the cell door with the bed. As punishment, the ring leader was placed in confinement. This event precipitated the guards now turning on the other inmates, and the level of cruelty escalated.

Several guards… but not all …upped their performance to act tough—to fit the preconceived expectations of their label—all under the watchful eyes of the Warden and Superintendent. These leaders promoted toughness as a shared attribute of conforming to the in-group to achieve the goal of exposing the toxicity of the penal system.

Due to the increased and creative oppression placed on the prisoners, the experiment was terminated early, after six days instead of two weeks.

Ultimately, the guards behaved in a stereotypical tough-guy manner, striving to fit the normative guidance they were given. Sometimes, toxic situations can bring out the worst in someone, especially in cases where the individual identifies with a leader and the group’s collective cause.

The famous and controversial Stanford Prison Experiment , conducted by social psychologist Philip Zimbardo and his colleagues at Stanford University, demonstrated the power of social roles, social norms, and scripts.

Social Roles

One major social determinant of human behavior is our social role— a pattern of behavior that is expected of a person in a given setting or group (Hare, 2003). Each one of us has several social roles. You may be, at the same time, a student, a parent, an aspiring teacher, a son or daughter, a spouse, and a lifeguard. How do these social roles influence your behavior? Social roles are defined by culturally shared knowledge. That is, nearly everyone in a given culture knows what behavior is expected of a person in a given role. For example, what is the social role for a student? If you look around a college classroom you will likely see students engaging in studious behavior, taking notes, listening to the professor, reading the textbook, and sitting quietly at their desks. Of course, you may see students deviating from the expected studious behavior such as texting on their phones or using Facebook on their laptops, but in all cases, the students that you observe are attending class—a part of the social role of students.

Social roles, and our related behavior, can vary across different settings. How do you behave when you are engaging in the role of son or daughter and attending a family function? Now imagine how you behave when you are engaged in the role of employee at your workplace. It is very likely that your behavior will be different. Perhaps you are more relaxed and outgoing with your family, making jokes and doing silly things. But at your workplace you might speak more professionally, and although you may be friendly, you are also serious and focused on getting the work completed. These are examples of how our social roles influence and often dictate our behavior to the extent that identity and personality can vary with context (that is, in different social groups; Malloy, Albright, Kenny, Agatstein, & Winquist, 1997).

Social Norms

As discussed previously, social roles are defined by a culture’s shared knowledge of what is expected behavior of an individual in a specific role. This shared knowledge comes from social norms—a group’s expectations of what is appropriate and acceptable behavior for its members—how they are supposed to behave and think (Deutsch & Gerard, 1955; Berkowitz, 2004). How are we expected to act? What are we expected to talk about? What are we expected to wear? In our discussion of social roles, we noted that colleges have social norms for students’ behavior in the role of student and workplaces have social norms for employees’ behaviors in the role of employee. Social norms are everywhere including in families, gangs, and on social media outlets. 

Because of social roles, people tend to know what behavior is expected of them in specific, familiar settings. A script is a person’s knowledge about the sequence of events expected in a specific setting (Schank & Abelson, 1977). How do you act on the first day of school, when you walk into an elevator, or are at a restaurant? For example, at a restaurant in the United States, if we want the server’s attention, we try to make eye contact. In Brazil, you would make the sound “psst” to get the server’s attention. You can see the cultural differences in scripts. To an American, saying “psst” to a server might seem rude, yet to a Brazilian, trying to make eye contact might not seem an effective strategy. Scripts are important sources of information to guide behavior in given situations. Can you imagine being in an unfamiliar situation and not having a script for how to behave? This could be uncomfortable and confusing. How could you find out about social norms in an unfamiliar culture?

Zimbardo’s Stanford Prison Experiment

In the summer of 1971, an advertisement was placed in a California newspaper asking for male volunteers to participate in a study about the psychological effects of prison life. More than 70 men volunteered, and these volunteers then underwent psychological testing to eliminate candidates who had underlying psychiatric issues, medical issues, or a history of crime or drug abuse. The pool of volunteers was whittled down to 24 healthy male college students. Each student was paid $15 per day and was randomly assigned to play the role of either a prisoner or a guard in the study. Based on what you have learned about research methods, why is it important that participants were randomly assigned?

A mock prison was constructed in the basement of the psychology building at Stanford. Participants assigned to play the role of prisoners were “arrested” at their homes by Palo Alto police officers, booked at a police station, and subsequently taken to the mock prison. The experiment was scheduled to run for several weeks. To the surprise of the researchers, both the “prisoners” and “guards” assumed their roles with zeal. In fact, on day 2, some of the prisoners revolted, and the guards quelled the rebellion by threatening the prisoners with night sticks. In a relatively short time, the guards came to harass the prisoners in an increasingly sadistic manner, through a complete lack of privacy, lack of basic comforts such as mattresses to sleep on, and through degrading chores and late-night counts.

The prisoners, in turn, began to show signs of severe anxiety and hopelessness—they began tolerating the guards’ abuse. Even the Stanford professor who designed the study and was the head researcher, Philip Zimbardo, found himself acting as if the prison was real and his role, as prison supervisor, was real as well. After only six days, the experiment had to be ended due to the participants’ deteriorating behavior. 

The Stanford prison experiment demonstrated the power of social roles, norms, and scripts in affecting human behavior. The guards and prisoners enacted their social roles by engaging in behaviors appropriate to the roles: The guards gave orders and the prisoners followed orders. Social norms require guards to be authoritarian (such behavior was reinforced; see Haslam, Reicher, & Van Bavel, 2018) and prisoners to be submissive. When prisoners rebelled, they violated these social norms, which led to upheaval. The specific acts engaged by the guards and the prisoners derived from scripts. For example, guards degraded the prisoners by forcing them do push-ups and by removing all privacy. Prisoners rebelled by throwing pillows and trashing their cells. Some prisoners became so immersed in their roles that they exhibited symptoms of mental breakdown; however, according to Zimbardo, none of the participants suffered long term harm (Alexander, 2001).

This text is adapted from OpenStax, Psychology. OpenStax CNX.

Suggested Reading

Haslam, S. A., Reicher, S., & Van Bavel, J. J. (2018, June 27). Rethinking the nature of cruelty: The role of identity leadership in the Stanford Prison Experiment. https://doi.org/10.31234/osf.io/b7crx

Alexander, M. (2001, August 22). Thirty years later, Stanford prison experiment lives on.  Stanford Report . Retrieved from http://news.stanford.edu/news/2001/august22/prison2-822.html.

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here .

Loading metrics

Open Access

Peer-reviewed

Research Article

Social norms don’t always work: An experiment to encourage more efficient fees collection for students

Affiliation Department of Anthropology, University College London, London, United Kingdom

* E-mail: [email protected]

Affiliation School of Public Policy, University College London, London, United Kingdom

• Antonio Silva, 

• Published: May 24, 2017
• https://doi.org/10.1371/journal.pone.0177354
• Reader Comments

The use of social norms has become the tool of choice for behaviourally informed interventions. However, it is still not clear for what type of contexts and populations is this intervention effective. This randomised controlled trial with 4298 students tests the applicability of social norms to improve the late payment of university tuition fees. We find that providing information to late payers does not increase their likelihood of paying. This finding highlights how the use of social norms may not always be an effective tool in influencing behaviour.

Citation: Silva A, John P (2017) Social norms don’t always work: An experiment to encourage more efficient fees collection for students. PLoS ONE 12(5): e0177354. https://doi.org/10.1371/journal.pone.0177354

Editor: Pablo Brañas-Garza, Middlesex University, UNITED KINGDOM

Received: November 21, 2016; Accepted: April 26, 2017; Published: May 24, 2017

Copyright: © 2017 Silva, John. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: Data are held in a public repository Figshare: https://figshare.com/s/cde25efbb058429689a4 .

Funding: The authors received no specific funding for this work.

Competing interests: The authors have declared that no competing interests exist.

Introduction

The use of descriptive social norms has increasingly become the tool of choice for behaviourally-inclined policy-makers and administrators. The idea is simple: inform people how many other people are doing an activity—and if the proportion is high—the remainder will be more likely to conform. The idea was initially tested by Cialdini and colleagues [ 1 , 2 ] in a series of experiments involving litter and messages left in hotel rooms to encourage guests to recycle their towels [ 3 ]. The use of norms has been tested in a variety of contexts and behaviours, including encouraging people to settle fines and taxes, with a series of successful trials carried out by the UK tax authority [ 4 ], as well as trials on recycling and charitable giving [ 5 , 6 , 7 ].

One question that arises is the extent to which the use of descriptive social norms can be applied to a wider range of domains and populations. This study tests whether providing feedback on norms of university tuition fee payment encourages late-paying students to settle their fees. Students are often new to university so may not be aware of the norms of payment and, as a result, might underestimate the level of fee payment across the university. We hypothesised that late paying students are more likely to pay their fees if they are informed of the typical behaviour of other students; but we found that providing this information was not effective in improving the rate of payment of tuition fees.

In this paper, we briefly review the literature on social norms and the underlying theoretical framework, and set out the rationale and design of the study. We then report the results and discuss their implications for the research on social norms.

Research on social norms

Humans are social beings who take behavioural cues from those around them. These cues are derived from social norms—common practices driven by shared beliefs about what is typical and appropriate behaviour. These informal understandings of rules emerge from cultural and social contexts and indicate how to behave in society; but they do not necessarily involve active reflection, as they are often unconsciously internalised [ 8 , 9 , 10 ].

There are two kinds of social norms, descriptive and injunctive. Descriptive norms are the perception of what people do and injunctive norms are what people approve or disapprove of [ 1 ]. Here we focus on descriptive norms with the assumptions that most people want to adhere to how the majority of people behave and tend to emulate the behaviour of people with perceived shared characteristics or where social distance is thought to be low [ 11 ].

In the case of private behaviours—like paying tuition fees—a large number of people act independently and don’t necessarily communicate their behaviour to others. This information asymmetry can then result in an individual being unaware of the normative behaviour and erroneously assume a different behaviour. The supply of the correct norm can then allow the individual to make a more informed choice. In this way, the supply of norms can be understood as a moral expectation that people aim to live to up [ 12 ], which can from a game-theoretic perspective facilitate the coordination of interactions [ 13 ].

Over the past decades, numerous studies have shown that the use of social norms can lead to behaviour change. One of the first applications of this approach was by Cialdini [ 1 ], who reduces littering using descriptive norms by varying the amounts of litter on the floor and using injunctive norms by placing handbills with different messages on cars’ windshields. In another famous study, messages left in hotel rooms in the U.S.A. indicating how many other people have recycled their hotel towels reduced the likelihood of guests asking for new towels [ 3 ].

There has also been an increasing use of social norms in public policy. Tax collection, in the UK and US in particular, has been at the forefront of using social norms to improve payment compliance. Messages using local norms of how many people pay tax significantly increase the numbers of late payers paying their tax [ 4 , 14 , 15 ]. Social norms have shown to be effective in several other contexts, including reducing prescription of antibiotics [ 16 ], curbside recycling [ 5 ], charitable giving [ 6 ], and energy consumption [ 17 ].

Despite most of the research in this area supporting the impact of social norms on behavioural change, there is some evidence that social norms are not always effective. In some cases, the provision of descriptive norms alone can highlight that a considerable number of people are, in fact, getting away with the non-normative behaviour, which reduces the effectiveness of the message [ 2 ]. Interventions based on average behaviours that inform people that they behave better than average can also result in a worsening of behaviour, as exemplified by energy conservation studies in which those with low levels of energy consumption increase their consumption after being informed that their average consumption is lower than their peers [ 18 , 19 , 20 ].

The effectiveness of social norms is also affected by the context of the interventions. For example, two studies of hotel towels, which were carried out in Europe [ 21 , 22 ], failed to replicate the results of Goldstein et al [ 3 ]. As Bonher & Schlüter [ 23 ] highlight, these results may be due to different baseline behaviours: while norms may be effective in a North American context where the baseline is lower, in Europe where the baseline is higher the effectiveness of the intervention is attenuated.

In summary, while the majority of published social norm studies show a positive impact in shifting behaviour [ 24 ], a closer reading suggests they are not always effective and that context matters. As Dolan et al ([ 25 ], p. 2) argue, ‘ the impact of norms on behaviors such as charitable giving and productivity might be quite different to that of other behaviors such as resource use . The key difference in these behaviors is understanding the production function of the behavior ’. In other words, the findings of existing social norms studies do not show external validity to all venues and contexts, in spite of many published and unpublished studies showing positive results [ 24 ]. Thus, it is important to test the effectiveness of social norms interventions in a broader set of contexts. In this study, we test the hypothesis that providing feedback on the average payment rates of university tuition fees as part of a payment reminder will encourage late paying students to settle their fees to the university.

Materials and methods

The study was designed and carried out at University College London (UCL), with the assistance of the Student Fees Office. The project was approved by UCL Research Ethics Committee (ID:3949/004). Consent was not obtained from the participants. Following the recommendation of UCL’s Research Ethics Committee, we sought permission from the student representative body, UCLU, to conduct the study.

The Fees Office sends out a reminder email to the students who have not paid their fees on time. The intervention kept the existing wording of the reminder email but inserted the sentence in bold “ OVER 90% OF UCL STUDENTS HAVE ALREADY PAID. PLEASE PAY THE AMOUNT DUE NOW.” which appeared in a prominent place near the top of the email (see Appendix for the full text of the treatment and control emails). We ran a two-arm trial over two years with the Fees Office sending reminders in four rounds of emails in November 2013, February 2014, November 2014, and February 2015 to unique late paying students of each round (no subjects appeared twice in our data). The control was the normal email, but adjusted so that the only difference between the treatment and control was the insertion of the new text. The allocation of treatment was randomised and blocked by age and gender to ensure balanced samples. The original sample includes a total of 4374 emails sent, from which 76 emails were removed from individuals who were in debt twice over the two-year period of the study, and as result received two emails. We removed the 2 nd email from the final analyses, after checking that receiving the treatment email twice had no significant effect on the payment rates (β = 0.25 [-0.92;1.42], p = 0.67). The final sample included 4298 individuals who were sent 812 emails in November 2013, 1459 emails in February 2014, 652 emails in November 2014, and 1375 emails in February 2015.

Table 1 shows the characteristics of the sample, which is balanced across treatment and control for all the variables that we have available. We analysed the data controlling for rounds, year, gender, age, and the amount of initial debt. The experiment was powered to detect a treatment effect of 5% in line with other studies using social norms to increase payment rates.

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

https://doi.org/10.1371/journal.pone.0177354.t001

We were not able to conduct a manipulation check to understand how the different messages affected the students. However, there was a surge in payments the following day after both the control and treatments emails were sent (27% of all payments occurred by the following day with the rate rapidly declining after the first few days), suggesting that students do read the emails and take action.

Model 1 shows logistic regression coefficients for treatment effect and Model 2 also includes covariates for gender, age, initial amount of debt, year (2013/2014 or 2014/2015), and round of reminders (November or February).

https://doi.org/10.1371/journal.pone.0177354.t002

As Table 2 shows, we find no significant effect of receiving the modified social norm email on the payment of the tuition fees when compared to the normal control email. This non-significant result remains when including covariates in the analysis on gender, age, amount of debt, year, and round ( Table 2 ). We also tested for the sub-group effects on gender, age, amount of initial debt, rounds, and year by running interactions of each factor with the treatment, but found no significant effect of treatment on any of the sub-groups. The payment rates over time also do not change between control and treatment emails (see Appendix , Fig 1 ).

Decay in payment rates after email sent for control and treatment groups.

https://doi.org/10.1371/journal.pone.0177354.g001

We find that younger individuals are more likely to pay their late fees after receiving the reminder than older individuals. Individuals are also more likely to pay when receiving the 2 nd round emails in February compared to the 1 st round in November.

We find no impact of the use of social norms on the likelihood of payment of late university tuition fees. This study is a rare example of an intervention providing normative information that failed to change behaviour.

Below we propose a few potential explanations. Social norms are more likely to be effective in stable and homogeneous populations in which normative behaviour can provide a cue of what is the most efficient behaviour in a specific context. For example, providing the social norms of late tax payment at the local level were more effective in increasing the payment rate than providing the norm at the country level [ 4 ]. In contrast, the sample of our study—university students—is a heterogeneous and temporary population that may result in individuals not being influenced by normative information. In particular, students attending a London-based university come from a variety of countries and ethnic groups, especially at post-graduate level.

Whilst we find older students were less likely to pay on time, the intervention had no effect on any age group. However, the mean age of our sample is still lower than most of other social norm studies, which may indicate that the impact of this type of social interventions is less effective on younger people. The sample of late payers is also unusual as the majority of students get their fees paid automatically through their student loans, and as result are not part of our sample. The remaining may be late payers for several reasons, such as being international students who are not able to take out loans, students who decide not to take out loans, or students that simply don’t have the financial means to pay the fees. They may also be dependent on others to pay the fees whose behaviour they cannot control, such as parents, or a funding body that does not have a regular payment schedule. In summary, in addition to limited access to normative information, students are also likely to face budget and logistic constraints.

The effectiveness of social norms interventions depends on a population having a shared sense of what is the desirable form of behaviour. In heterogeneous populations, like the one in this study, this shared identity may not exist and anti-conformist bias may emerge, in which people prefer not to conform to the social norm as they don’t identify with the wider group [ 26 , 27 ]. While we have not been able to define the reason behind the lack of impact of providing feedback on the normative behaviour in this context, our study highlights how social norms don’t always work, and the importance of determining the appropriate contexts where they may be effective.

Treatment email

Student ID: 010207306–11/Nov/2013

Dear Student Name,

Your UCL Time Limited Enrolment has expired and our records show that you have not paid sufficient fees or provided the required sponsorship evidence to complete the enrolment process. The balance now due is £2,375.00.

OVER 90% OF UCL STUDENTS HAVE ALREADY PAID .

PLEASE PAY THE AMOUNT DUE NOW .

The easiest way to pay is on-line via the UCL website .

You can ask somebody to do this for you but they will need your Student Number (shown above). Alternatively, we accept payments by credit/debit card (sorry, no Amex, Electron or Diners Club), Bank transfer and UK Sterling Cheque. Full details of how and where to pay can be found at the Student Fees website

IF YOU FAIL TO ACT WITHIN THE NEXT TEN (10) WORKING DAYS WE WILL START TO APPLY SANCTIONS THAT WILL RESULT IN YOU BEING DENIED ACCESS TO SPECIFIC SERVICES & FACILITIES AND SUSPENSION OF YOUR REGISTRATION MAY THEN FOLLOW.

Students in debt to UCL are also to be prevented from attending a graduation ceremony, receiving official notification of results and being awarded their degree. If you have recently paid or provided sponsorship evidence, your Portico account will soon be updated. We advise that you check your Portico Account on a regular basis. If you forwarded evidence of external sponsorship (employer, government embassy etc) to UCL around the start of session that is not recorded on Portico, please email another copy to [email protected] ensuring the email subject is "Evidence of Sponsorship". If you enrolled for the 2013/14 session on a Time Limited basis, this is the second email reminder that has been sent to your UCL account.

Please accept our apologies if you have received this email in error in which case you must contact Student Fees office (details below) as a matter of urgency to resolve this matter.

Any enquiries should be directed to the Student Fees Office by telephone or in person. The office is open for personal callers Monday to Friday from 10am to 4pm.

University College London,

Student Fees & Credit Control Section

Gower Street

London, WC1E 6BT

Tel: +44 (0)20 7679 4125 or 4128

Control email

Acknowledgments.

We would like to thank Martin Lofthouse from UCL Fees Office for his support of the project as well as his colleagues Ahmed Tufael and Alan Hitchcock. We also thank Julian Carter from UCL’s Financial Analysis & Reporting (FAR) team for providing the matched data. We thank the editor and reviewers for helpful comments in revision.

Author Contributions

• Conceptualization: AS PJ.
• Data curation: AS PJ.
• Formal analysis: AS PJ.
• Investigation: AS PJ.
• Methodology: AS PJ.
• Project administration: AS PJ.
• Resources: AS PJ.
• Software: AS PJ.
• Supervision: AS PJ.
• Validation: AS PJ.
• Visualization: AS PJ.
• Writing – original draft: AS PJ.
• Writing – review & editing: AS PJ.
• View Article
• Google Scholar
• 4. Hallsworth, M, List, JA, Metcalfe, RD, Vlaev, I. The behavioralist as tax collector: using natural field experiments to enhance tax compliance. NBER Working Paper No. 20007. 2014; 1–44. Available from http://www.nber.org/papers/w20007 .
• 8. Sherif M. The psychology of social norms. New York: Harper & Brothers; 1936.
• 10. Hechter M, Opp K-D. (ed) Social norms. New York: Russell Sage Foundation; 2001.
• 13. Espinosa, M, Hernández, P. In: Branas, P, Cabrales, A editors. Experimental economics I: decisions. Coordination games. Basingstoke: Palgrave. 2015, pp. 53–71.
• 14. Coleman, S. The Minnesota income tax compliance experiment: state tax result. Minnesota Department of Revenue Apr. 1996); available at http://www.revenue.state.mn.us/research_stats/research_reports/19xx/research_reports_content_complnce.pdf
• 15. Coleman, S. The Minnesota income tax compliance experiment: replication of the social norms experiment. 2007; SSRN. Available at SSRN: http://ssrn.com/abstract=1393292
• 20. Ayers, I, Raseman, S, Shih, A. Evidence from two large field experiments that peer comparison feedback can reduce residential energy usage. 2009; NBER Working Paper No. 15386, http://www.nber.org/papers/w15386 .
• PubMed/NCBI
• 24. John, P, Sanders, M, Wang, J. The use of descriptive norms in public administration: A panacea for improving citizen behaviours? 2014. Available at SSRN: http://ssrn.com/abstract=2514536
• 25. Dolan, P, Metcalfe, RD. Neighbors, knowledge, and nuggets: two natural field experiments on the role of incentives on energy conservation. 2015; Available at SSRN: http://ssrn.com/abstract=2589269 .

• Skip to primary navigation
• Skip to main content
• Skip to primary sidebar
• Skip to footer

Social Mettle

A List of Quirky Ideas for Social Norm Breaching Experiments

Suppose there is a game in which you are not allowed to say 'yes' or 'no', and have to answer only with another question. Using up the entire range of 'wh-questions', you can think of how entertaining this game can get. Social norm breaching is nothing different than this.

Suppose there is a game in which you are not allowed to say ‘yes’ or ‘no’, and have to answer only with another question. Using up the entire range of ‘wh-questions’, you can think of how entertaining this game can get. Social norm breaching is nothing different than this.

No Suggestions Implied! While a Superman costume flaunting an underwear is taken well, someone wearing undergarments over normal clothes (that too at a fancy dress party) becomes an act of breach! Unfair, isn’t it!

Eating with your hands while dining at a fancy restaurant, wearing your bathrobe to college one day, sending a reply via email when someone had called you on your cell phone, or talking to a stranger by getting very close to him/her, and such similar deeds are ones that a normal human being would usually avoid doing.

If you have tried any of these, you can be called a researcher who was engaged in studying social norm breaching experiments. Such kind acts of nonsense, which you may call bizarre, are not solely meant for entertainment purposes. This testing of socially accepted rules are mainly a part of the fields of sociology and social psychology. Yes, academics can get very interesting at times!

What are Breaching Experiments in Sociology?

Breaching experiments try to study the reactions of people when a social norm is broken or violated.

There are some unwritten rules that all of us follow in our day-to-day conduct. How one would (rather should) behave in a given situation is predefined and based on a lot of assumptions. These experiments try to break these ‘taken for granted’ social norms. Reactions of others to such tricks are also fun to look at. This concept is associated with the ethnomethodology theory of sociology , put forth by Harold Garfinkel.

An unexpected behavior or comment leaves the respondent completely puzzled, making the experiment successful. The approach behind such experiments highlights that, people continue to make a number of such rules everyday, and do not even realize it.

Experiment Ideas

Clearly, a breaching experiment is like asking for trouble. When the action is troublesome, it makes it visible that practices leading to social stability are so much ingrained into our minds. Breaching of norms has to be a deliberate act though; it is not an issue of conflicting opinions leading to disobedience of a given norm. You can try troubling others with the following ideas.

– To a casual question like ‘what’s up?’, you can say ‘the sky’. ‘How’s it going?’ can be replied to in an exhilarating manner, like ‘I didn’t see any ‘it’ going’. When people are not really interested in knowing about you, and they still ask those questions, you may actually stop them and really explain to them some random event going on in your life. (Be very sure about who you want to experiment with this though!)

– Some tests that college students were asked to take, involved behaving like a stranger or renter in one’s home. Talking only when asked about something, or being very polite, are some things their parents reacted to quite strongly.

– In the tic-tac-toe game, ask a person to play first. When he/she places an ‘X’ in a square, you place an ‘O’ on a line forming the matrix, and not in any square space. That person might get confused, or would exclaim, “Have you gone crazy?” Behaving according to the established practices of following given rules is so important here, even if it is a game. This exemplifies an established social order.

– At a decently crowded public place, get one of your friends to stand opposite you. You act like both of you are talking about something important. Then, act as if the both of you are holding a very thin and delicate cotton string in your fingertips, each one of you holding one end of it. Now, start to move away, very slowly, so that people feel that you are holding something very precious. Shout out words like, ‘easy’, ‘be careful’, or ‘watch out’. You may find a few people actually believing you and ducking while they pass through. Someone might even go around you, so as to not break that string. You would notice, it is very easy to create social norms.

Experiment Examples

Here are some examples of interpersonal conversations, mentioned in ethnomethodology literature as case studies of experimentation given by Garfinkel. These have been sourced from books like ‘Garfinkel and Ethnomethodology’ by John Heritage, and ‘Sociology in Perspective’ by Mark Kirby.

– The subject was telling the experimenter―a member of the subject’s car pool―about having had a flat tire while going to work the previous day.

S : I had a flat tire. E : What do you mean, you had a flat tire? She appeared momentarily stunned. Then she answered in a hostile way: ‘What do you mean? What do you mean? A flat tire is a flat tire. That is what I a meant. Nothing special. What a crazy question!’

– By asking ‘What do you mean?’, as a response to every statement, students were asked to continue the conversation.

S : Hi, Ray. How is your girlfriend feeling? E : What do you mean ‘How is she feeling?’. Do you mean physically or mentally? S : I mean how is she feeling? What’s the matter with you? (He looked peeved.) E : Nothing. Just explain a little clearer as to what you mean. S : Skip it. How are your Med School applications coming? E : What do you mean ‘How are they going?’ S : You know what I mean. E : I really don’t. S : What’s the matter with you? Are you sick?

– On Friday night, my husband and I were watching television. He remarked that he was tired. I asked, ‘How are you tired? Physically, mentally, or just bored?’

S : I don’t know, I guess physically, mainly. E : You mean that your muscles ache, or your bones? S : I guess so. Don’t be so technical. (After more watching) S : All these old movies have the same kind of old iron bedstead in them. E : What do you mean? Do you mean all old movies, or some of them, or just the ones you have seen? S : What’s the matter with you? You know what I mean. E : I wish you would be more specific. S : You know what I mean! Drop dead!

– The victim waived his hand cheerily.

S : How are you? E : How am I in regard to what? My health, my finance, my school work, my peace of mind, my … S : (Red in the face and suddenly out of control.) Look! I was just trying to be polite! Frankly, I don’t give a damn how you are.

The results from these cases proved that the experimenters could successfully break the norms. It was possible because of the fact that, any given conversation (or communication) takes place smoothly, ‘assuming the background knowledge’, which helps two people make sense of what the other means.

Well, if you’ve got the point now, you can be real ‘innovative and original’ with this act of breaching. Oh, but just be sure that you don’t mess with the wrong people at the wrong time.

Like it? Share it!

Get Updates Right to Your Inbox

Further insights.

Privacy Overview

A RESOURCE CREATED BY:

How Does Media Influence Social Norms? A Field Experiment on the Role of Common Knowledge.

A radio program in mexico decreased personal and perceived social acceptance of violence against women..

Cite this Article

Arias, Eric. “How Does Media Influence Social Norms? Experimental Evidence on the Role of Common Knowledge.”  Political Science Research and Methods , vol. 7, no. 3, July 2019, pp. 561–578.

Arias, E. (2019). How Does Media Influence Social Norms? Experimental Evidence on the Role of Common Knowledge.  Political Science Research and Methods ,  7 (3), 561–578. 

Arias, Eric. “How Does Media Influence Social Norms? Experimental Evidence on the Role of Common Knowledge.”  Political Science Research and Methods  7, no. 3 (July 2019): 561–78. 

• Introduction
• Methodology

Media has the power to influence individual beliefs, attitudes, and behaviors. Thus, social norms that influence and encourage gender violence have been targeted by policy interventions using mass media communication. Prior research suggests that media influences through two effects: the individual or direct effect (private) or the social or indirect effect (public). In the individual effect, media information about new norms may persuade individuals to accept them. In the social effect, the information creates common knowledge of a norm and enhances social coordination as individuals more readily accept the information if they believe others have also accepted it. This study examined whether media’s social mechanism has a stronger impact than its individual mechanism on changing violent attitudes against women.

This study conducted a natural and randomized experiment in the rural indigenous community of San Bartolome Quialana in Oaxaca, Mexico using a multi-part soap opera radio program telling a story of a relationship that slowly becomes violent. The soap opera was broadcast via a community loudspeaker and only reached a portion of the community due to topographical conditions (creating the natural experiment component). Households within the loudspeaker’s reach were either randomly invited to listen to the soap opera at a community meeting or were able to hear the public broadcast in their own homes. This design tested whether public transmission of the soap opera alone was sufficient to influence norms, and whether creating certainty about common knowledge from face-to-face interactions with community members enhanced the social effects. Individuals outside the loudspeaker’s reach were randomly and privately invited to listen to the soap opera using an audio CD to test the individual mechanism. A post-intervention survey measured norms, attitudes, and behavior regarding violence against women.

The social broadcast groups, both the village loudspeaker and community meeting broadcasts, were more likely to reject violence against women and more likely to support gender equality than the individual broadcast group. Surprisingly, the social broadcast groups were also less likely to think domestic violence will decline in the future.

• When asked about personal beliefs on domestic violence, individual broadcast group participants were less likely than those in the social broadcast groups to acknowledge that violence against women is a recurring problem in the community.
• When asked about perceived social rejection of domestic violence in the community, individual broadcast group participants were less likely than those in the social broadcast groups to think that others in the community reject domestic violence.
• When asked about their expectations for the future decline of domestic violence, participants in the social broadcast groups were less likely than those in the individual broadcast group to think that domestic violence will decline.
• When asked about transmitting values of gender equality to the next generation, individual broadcast group participants were less likely than participants in the social broadcast groups to say they would educate their children on gender equality.
• When asked about potential reactions to domestic violence, individual broadcast group participants were less likely than participants in the social broadcast groups to say they would intervene if they heard or witnessed domestic violence.
• When asked whether they would sign a petition to support the creation of a violence against women support group, individual broadcast group participants were less likely than participants in the social broadcast groups to sign the petition.

The findings indicate that information delivered socially, either publicly or through face-to-face interactions, is more influential than information delivered individually or privately. Furthermore, although some evidence suggested that the face-to-face interactions of community meetings can enhance social effects, delivering information publicly through the village loudspeaker broadcast was sufficient to influence attitudes and norms. Lastly, the treatment effects on pessimistic expectations about the future emphasize the importance of implementing interventions with clear institutional mechanisms through which individuals can act upon their updated beliefs.

The author partnered with the UNESCO Office in Mexico to conduct a natural experiment and a randomized field experiment. The audio soap-opera intervention was developed with a regional NGO to challenge gender norms and discourage violence against women. The soap opera, which included 4 episodes of 15 minutes each, contained regional dialect and locations to help participants better relate to the story. The study design included four conditions: individual broadcast, community meeting broadcast, village loudspeaker broadcast, and baseline group. The individual broadcast group comprised of randomly selected individuals living outside the village loudspeaker’s reach who were given a CD-rom. The community meeting broadcast group comprised of randomly selected individuals living within the village loudspeaker’s reach and  listened to the broadcast together in a community center setting. The village loudspeaker broadcast group was composed of the remaining individuals who lived within the reach of the public village loudspeaker. Households outside of the loudspeaker´s reach who did not receive the CD-rom were used as the baseline group. Participants were asked a series of questions on six outcomes relating to domestic violence: three questions on their beliefs and estimation of others´ beliefs and actions around violence against women (personal beliefs, perceived social rejection of domestic violence, and future expectations) and three questions measuring individual attitudes and actions (transmission of values to the next generation, their potential reaction to a domestic violence episode, and if they would sign a petition creating a violence against women support group). Responses were then compared across all four treatment conditions.

• Top of Page
• Authors & Citations
• Related Studies

Related GAP Studies

In the uk, incidents of domestic abuse where perpetrators use alcohol increase after early soccer games. , football, alcohol, and domestic abuse.

A cash transfer program in Kenya significantly boosted school enrollment among adolescent girls and elevated expectations for completing secondary education. 

Effects of a single cash transfer on school re-enrollment during covid-19 among vulnerable adolescent girls in kenya: randomized controlled trial.

Member Article

Don’t throw in the towel: use social influence research.

• Social Influences

Commercial decision-makers commonly base important program or policy choices on thinking grounded in the established theories and practices of a variety of business-related fields (e.g., economics, finance, distribution, accounting, supply management). What is vexing is how seldom these decision-makers avail themselves of established psychological theories and practices.

Take, for example, hotels. Via a card strategically placed in their room, guests in many hotels are urged to reuse their towels to help conserve environmental resources. According to the Project Planet Corporation, which supplies such cards to hoteliers, most guests will recycle at least one towel sometime during their stay, provided that they are asked to do so. For a social influence researcher like me, of course, the question becomes: “What wording of the request card will most effectively spur the desired behavior?” Because hotel managers typically don’t know and don’t consult the psychological literature on social influence, my guess is that their message choices are less than optimal and, consequently, are costing them plenty in replacement of (rather than reuse of) towels.

To make this point, my graduate students Noah Goldstein and Vladas Griskevicius and I have been registering what the request cards typically say in a wide variety of hotels. Although there is some variation, the cards most frequently try to increase recycling by focusing guests on: 1) environmental protection and 2) environmental cooperation . That is, guests are almost invariably informed that reusing their towels will conserve energy and help save the environment. In addition, they typically are told that towel reuse will allow them to become cooperating partners with the hotel in furthering its conservation efforts; to encourage such cooperation, guests may be told that the hotel will donate some of the savings from its towel-reuse program to environmental causes. This is expected to increase recycling above the simple environmental-protection appeal.

Notable in its absence from these persuasive appeals is a focus on descriptive social norms — that is, what most other people are doing in this situation (Goldstein & Cialdini, in press; Schultz, 1999). Compared to the existing types of communications, we wondered what would be the effect of simply informing guests that the majority of their counterparts do reuse their towels when requested. To examine that question, we enlisted the aid of the management of the Holiday Inn of Tempe, Arizona, where we randomly placed cards with conceptually different recycling appeals in its 190 rooms. The room attendants were trained to record the relevant reuse data. The cards were identical in two respects. First, on the front, they informed guests that they could participate in the program by placing their used towels on the bathroom towel rack or curtain rod. Second, on the back, the cards provided information regarding the amount of energy that could be saved if most guests participated in the program.

The cards differed, however, in the persuasive appeals designed to stimulate recycling. One focused guests on environmental protection. It stated:

A second type of card focused guests on environmental cooperation. It stated:

A third type of card focused guests on the descriptive norms of the situation. It stated:

Before presenting the outcomes of this study, I’ll need to describe certain aspects of the data that will help in their interpretation. First, we only looked at the data from single guests, as it was only single-occupancy rooms that allowed us to assure that the person reading the card was the person making the reuse decision. Second, we only examined the data from the first night’s stay, as all subsequent data would be subject to statistical dependencies. Finally, we did not count as a reuse effort a towel that was hung on a door hook or doorknob, as we wanted to eliminate the likelihood of guests complying unintentionally with the request. For each of these reasons, we believe the overall percentage of reuse was artificially depressed. Nonetheless, there were interesting differences among the varying communications.

The environmental-protection message produced a reuse rate of about 38 percent. The environmental-cooperation message fared no better, stimulating reuse only 36 percent of the time. But, the descriptive-norms message was significantly more effective than either of the others, generating recycling activity about 48 percent of the time.

Two aspects of these data are worth noting. First, the most successful of the communications was the one that we have never seen among the wide range of such messages that we have registered. This simple-but-effective appeal didn’t emerge from a history of trial and error to become a hotel “best practice.” Instead, it emerged from knowledge of the psychological literature on social influence.

Second, in the environmental-cooperation condition, the hotel’s pledge to donate to an environmental cause if guests reused their towels didn’t increase recycling at all. Why not? Perhaps the guests didn’t trust the hotel’s promise. Or perhaps they didn’t trust the hotel to choose an environmental cause for them. But when we conducted this study, we thought the reason for a poor showing of the cooperation condition might be due to something else: There is no sense of social obligation to cooperate with someone who offers you something, provided that you perform a favor first. However, there is a powerful sense of social obligation — embodied in the norm for reciprocation — to cooperate with someone who does something for you first and then asks for a favor in return (Cialdini, 2001). That is, members of all human societies are trained from childhood to reciprocate the favors they receive from others (Gouldner, 1960). We have very nasty names for those who don’t play by this rule: We call them moochers, or takers, or ingrates. Or teenagers. And, no one wants to be labeled in those ways. Consequently, people will go to great lengths to reciprocate a benefactor.

Therefore, we thought that the cooperation message got the concept (of cooperation) right but got the sequence wrong. According to a social psychological analysis, the better way to induce the desired response would be for the hotel to give the donation first and then ask guests to cooperate in this effort by conserving resources.

To test our thinking, we included a fourth type of message in our design, a reciprocation appeal. It read as follows:

WE’RE DOING OUR PART FOR THE ENVIRONMENT. CAN WE COUNT ON YOU? Because we are committed to preserving the environment, we have made a financial contribution to a nonprofit environmental protection organization on behalf of the hotel and its guests. If you would like to help us in recovering the expense, while conserving natural resources, please reuse your towels during your stay.

The result was an almost 47 percent success rate, significantly greater than the cooperation condition. Once again, we see that a relatively minor change, informed by social psychological theory, can serve as a corrective to the existing practices of otherwise astute businesspeople who would never leave themselves comparably uninformed in other arenas of business practice.

• Cialdini, R. B. (2001). Influence: Science and practice (4th ed.) Needham Heights, MA: Allyn & Bacon.
• Goldstein, N. J., & Cialdini, R. B. (in press). Using social norms as a lever of social influence. In Pratkanis, A. R. (Ed.), Science of social influence . New York: Psychology Press.
• Gouldner, A. W. (1960). The norm of reciprocity: a preliminary statement.
• American Sociological Review, 25, 161-178.
• Schultz, P. W. (1999). Changing behavior with normative feedback interventions: A field experiment on curbside recycling. Basic and Applied Social Psychology, 21 , 25-38.

The article was very informative.I should consider replicating this research in Kenya

APS regularly opens certain online articles for discussion on our website. Effective February 2021, you must be a logged-in APS member to post comments. By posting a comment, you agree to our Community Guidelines and the display of your profile information, including your name and affiliation. Any opinions, findings, conclusions, or recommendations present in article comments are those of the writers and do not necessarily reflect the views of APS or the article’s author. For more information, please see our Community Guidelines .

Please login with your APS account to comment.

About the Author

ROBERT CIALDINI is a professor of psychology at Arizona State University and an APS Fellow and Charter Member.

Robert B. Cialdini and Jennifer L. Eberhardt on The 7 Principles of Influence

Jennifer L. Eberhardt and Robert B. Cialdini explore the power of influence and the importance of “shipping” psychological science to address real-world challenges.

We Are More Envious of Things That Haven’t Happened Yet

We are more envious of someone else’s covetable experience before it happens than after it has passed, according to research published in Psychological Science, a journal of the Association for Psychological Science. “Enviable events lose

Around the World, People Have Surprisingly Modest Notions of the Ideal Life

Rather than being “maximizers,” people seem to aspire to moderate ideal levels of traits, such as pleasure, intelligence, personal freedom, and longevity.

Privacy Overview

Social Norms Experiment

by Brenna Fisher | Sep 30, 2016 | 1 comment

In one of my sociology classes, sociological theory, we talk a lot about deviant behaviors and social norms in our society. This past week, my professor challenged us to break one of these norms, so I decided to go to the dining hall and stare at people intently when asking them for something. I would behave completely normally, except my eyes remained wide and unblinking during our entire conversation. My first encounter was with the employee behind the counter at the sandwich station. I walked up and told them what I wanted on my sandwich without blinking. They put it in the oven to toast it. Next, a second employee asked me what toppings I wanted, and I repeated my experiment. Lastly, I had to pay for my sandwich, and I again used the same strategy. Of course, I had to blink a couple times, but for the majority, I kept my eyes open. It was a fairly simple experiment, but one that was just deviant enough to raise some eyebrows.

People reacted fairly calmly to my experiment. I got some confused looks from all employees. All three scrunched up their eyebrows at me and seemed mildly concerned and confused. Employee #1 seemed most surprised and confused. This was most likely because he was the first to witness it, while employees #2 and #3 witnessed it happening before I reached their station. They continued to take my order and treated me the same as I usually am treated. Employee #2 was not too keen on looking up at me, and avoided my eye contact for the most part. When they thought I couldn’t see them, both employees behind the counter exchanged a confused glance. However, this was the most reaction I got from any of the three employees. I believe the fact that they reacted very minimally was due to other social norms in our culture. It is not normal to freak out if someone breaks a social norm. It is weird, and you are generally a little uncomfortable, but the norm in our culture is to remain calm and collected through most non-emergency situations.

I think the reason this is a norm is because it does make people uncomfortable. No one likes to be stared at. In fact, we are socialized from a young age that staring is “rude.” Indeed, it does make situations awkward when there is no natural break in eye contact and minimal blinking. The conversation seems very forward and focuses, even if you are simply ordering a sandwich. I don’t see anything wrong with this norm—it simply allows for more comfortable exchanges in day to day conversations and activities.

That is a very interesting experiment! It can be very boring at times when everyone follows the norms that are expected in a certain society. I try to be different sometimes to entertain myself and its always a pleasure to experience all the emotions and reactions of those people around me. I wish people embraced being awkward, it’d make life much simpler.