key: cord-1009990-yyjepxc6
authors: Jaffe, Anna E.; Graupensperger, Scott; Blayney, Jessica A.; Duckworth, Jennifer C.; Stappenbeck, Cynthia A.
title: The Role of Perceived Social Norms in College Student Vaccine Hesitancy: Implications for COVID-19 Prevention Strategies
date: 2022-01-26
journal: Vaccine
DOI: 10.1016/j.vaccine.2022.01.038
sha: 8492a705f3e5337c77681a5816a7009ff37bdb30
doc_id: 1009990
cord_uid: yyjepxc6

Among US adults, the highest rates of hesitancy to receive the COVID-19 vaccine are among young adults aged 18 to 25. Vaccine hesitancy is particularly concerning among young adults in college, where social interactions on densely populated campuses can lead to substantial community spread. Given that many colleges have opted not to mandate vaccines, identification of modifiable predictors of vaccine hesitancy – such as perceived social norms – is key to informing interventions to promote vaccine uptake. To address this need, we examined predictors of and explicit reasons for vaccine hesitancy among 989 students aged 18 to 25 recruited from four geographically diverse US universities in the spring of 2021. At the time of the survey, 57.3% had been vaccinated, 13.7% intended to be vaccinated as soon as possible, and 29.0% were vaccine hesitant. Common reasons for hesitancy were wanting to see how it affected others first (75.2%), not believing it was necessary (30.0%), and other reasons (17.4%), which were examined via content analysis and revealed prominent safety concerns. Despite these varied explicit reasons, logistic regressions revealed that, when controlling for demographics and pandemic-related experiences, perceived descriptive and injunctive social norms for vaccine uptake were each significant predictors of vaccine hesitancy (ORs = 0.35 and 0.78, respectively). When both norms were entered into the same model, only perceived descriptive norms uniquely predicted vaccine hesitancy (OR = 0.37; 95% CI: 0.29 – 0.46). Findings suggest perceived social norms are strongly associated with vaccine-related behavior among young adult college students. Correcting normative misperceptions may be a promising approach to increase vaccine uptake and slow the spread of COVID-19 among young adults.

Beginning in early 2020, and spanning the entirety of 2021, the COVID-19 pandemic associated with the rampant spread of the SARS-CoV-2 virus has been an unmitigated public health crisis (Desai & Patel, 2020) . The pandemic has been associated with concerns beyond those directly related to the virus, such as job loss (Kochnar, 2020) and financial instability (Menasce Horowitz et al., 2021) , increased intimate partner aggression (Parrott et al., 2021) , loneliness and other mental health problems (Graupensperger, Cadigan, et al., 2021; Killgore et al., 2020; Liu et al., 2020; Luchetti et al., 2020; Palgi et al., 2020; Weissbourd et al., 2021) , and changes in substance use behaviors (Clay & Parker, 2020; Jaffe et al., 2021; Satre et al., 2020) . In the US, the Centers for Disease Control and Prevention (CDC) have endorsed empirically-supported prevention strategies to reduce person-to-person spread of the virus (e.g., wearing face masks, social distancing; CDC, 2020) and state governments have inconsistently implemented ever-evolving physical distancing measures in response to surges in new cases (e.g., shelter-in-place orders, travel restrictions, curfews). Despite these efforts, new cases in the US rose in the latter half of 2021, approaching daily-case rates similar to the previous peak observed in December of 2020 (World Health Organization [WHO] , 2021). Beyond the immediate impact of the virus, continued spread and the prolonging of the pandemic is concerning because the SARS-CoV-2 virus has and will continue to mutate as long as ongoing transmission persists (del Rio et al., 2021) .

It has become clear that the most promising strategy for combatting COVID-19 is prophylactic vaccines which can facilitate sufficient herd immunity (Corey et al., 2020; Graham, 2020) . Experts estimate that approximately 70-90% of people need to be vaccinated to achieve herd immunity (Anderson et al., 2020) . However, slow vaccine uptake rates and subsequent virus mutations have meant a 'moving goalpost' scenario whereby exact estimates of required vaccine coverage are unknown and herd immunity may no longer be feasible (Tkachenko et al., 2021) . Nevertheless, increasing COVID-19 vaccine uptake rates is the most critical public health priority given that simulation analyses show the current rate of vaccination is insufficient for preventing exacerbation of the pandemic and further mutation of more contagious SARS-CoV-2 variants (Sah et al., 2021) .

Even prior to the COVID-19 pandemic, the WHO considered vaccine hesitancydelayed acceptance or refusal of vaccines despite availability -as a top-ten global health threat (WHO, 2019) . The threat currently posed by vaccine hesitancy, specific to COVID-19, positions this hurdle among the most salient issues facing our society (Wiysonge et al., 2021) .

Estimates of vaccine hesitancy have been heterogenous and dynamic, though data consistently show a non-trivial proportion of people who remain hesitant to receiving a COVID-19 vaccine (see Aw et al., 2021) . One group that is of particular concern is young adults, who have the lowest rates of vaccine uptake (CDC, 2021) and the highest levels of vaccine hesitancy (Hamel et al., 2021) relative to other adult age groups in the US.

Although research on recent variants continues to emerge (Riley et al., 2021) , there is evidence suggesting young adults may be at lower risk of developing severe symptoms and complications associated with COVID-19 (Zhu et al., 2020) . Because symptoms are often minor or not present at all, young adults may be less likely to isolate and therefore more likely to unknowingly spread the virus (Farber & Johnson, 2020) , resulting in the poorest adherence to other mitigation strategies relative to other age groups (Jørgensen et al., 2020) .

As such, increasing vaccination rates among young adults may be a key step toward reducing community transmission, including spread to vulnerable and high-risk individuals.

COVID-19 incidence rates have been particularly high on US college campuses (Dickler, 2020) . The densely-populated structures on college campuses that require close contact (e.g., lecture halls, classrooms, residence halls, Greek housing) place colleges at heightened risk for community spread (Losina et al., 2021; Lu et al., 2020) . College students are also motivated to socialize and drink alcohol, which has also been associated with poorer adherence to social distancing measures . Thus, college students represent a high-risk subgroup for community transmission in which reduced vaccine hesitancy and increased vaccine uptake would be important.

The most common framework for operationalizing vaccine hesitancy is the 5C's model of individual-level determinants of vaccine hesitancy: confidence, complacency, convenience (or constraints), risk calculation, and collective responsibility (Betsch et al., 2018 ). Although these individual-level reasons for vaccine hesitancy are critical to address, they may appeal less to young adults, who are largely motivated by social factors (Wiysonge et al., 2021) . As such, addressing vaccine hesitancy among young adults may require a deeper consideration of social influences beyond the traditional 5C's model (Rimal & Storey, 2020) .

Young adults' health behaviors and attitudes are powerfully influenced by the behaviors and attitudes of their peers (i.e., social norms; Berkowitz, 2004) . Indeed, social norms are central to several behavioral theories such as Social Norms Theory (Berkowitz, 2004) and Theory of Reasoned Action (Fishbein & Ajzen, 2011) . Social norms are distinguished into two primary sources of influence: (1) perceived descriptive norms that entail perceptions of others' behavior, and (2) perceived injunctive norms that entail perceptions of others' attitudes or opinions towards a behavior (Cialdini et al., 1990) . Both perceived descriptive and injunctive norms are robust predictors of a wide range of healthrelated behaviors, such as seatbelt adherence , sunscreen use (Mahler et al., 2008) , alcohol use (Graupensperger et al., 2020; , and risky sexual behavior .Young adults are particularly susceptible to perceived social norms as they have a drive for peer approval (Burnett et al., 2011) and are motivated to adhere to behaviors and attitudes of others as a means of fitting-in and being accepted by peers (Helms et al., 2014) .

Specific to vaccinations, perceptions of social norms are related to college students'

intentions to receive vaccines for influenza (Quinn et al., 2017) and Human Papillomavirus (Stout et al., 2020) . Moreover, there is emerging evidence that perceived social norms may play a key role in COVID-19 vaccination uptake. A recent quasi-experimental study reported a strong positive association between perceptions of the proportion of others who would get a COVID-19 vaccine and one's own intentions to get vaccinated (Agranov et al., 2021) . A 10% increase in perceptions of others' vaccine intentions was associated with a 6.8% increase in one's own propensity to vaccinate, on average. Similarly, US adults who reported greater expectations that friends and family (Latkin et al., 2021) or people in their county would get vaccinated were more likely to express positive vaccine intentions.

Specific to college students, a recent study found that perceived norms were a strong predictor of students' own vaccination intentions: those who believed a greater proportion of young adults would get vaccinated were more likely to report intentions to get the COVID-19

vaccine themselves (Graupensperger, Abdallah, et al., 2021) .

Given that there may be a range of explicit reasons young adults have for vaccine hesitancy in the specific context of the COVID-19 pandemic, one might question whether perceived social norms would be a meaningful predictor of vaccine hesitancy across individuals with varied rationales. Among college students who did not intend to get a COVID-19 vaccine as of November 2020, 85.2% reported they were afraid or nervous of unknown side effects, 68.5% did not trust that the vaccines would be sufficiently tested, 29.6% believed a vaccine would give them COVID-19 or make them sick otherwise, and over a quarter (25.9%) did not think the vaccines would work (Graupensperger, Abdallah, et al., 2021) . These reasons for vaccine hesitancy among college students were highly similar to other studies from the US, UK, and Taiwan, which have highlighted concerns about vaccine safety, side effects, perceptions that others need it more, and distrust of vaccines Robertson et al., 2021; Small et al., 2021; Tsai et al., 2021; Williams et al., 2020) .

However, given the importance of peer influences on young adults and college students in particular, perceived norms may be a unifying factor that drives attitudes and behaviors for the COVID-19 vaccine across individuals with a wide range of explicit reasons for hesitancy.

Perceived social norms regarding others' vaccination behaviors and attitudes have been indicated as a potentially salient predictor of young adults' vaccine uptake (Graupensperger, Abdallah, et al., 2021; Latkin et al., 2021) , which could have important intervention implications. However, several noteworthy limitations should be addressed with additional research. First, existing studies on the role of perceived social norms for intentions to receive the COVID-19 vaccine were conducted with data prior to public availability of COVID-19 vaccines (i.e., December of 2020), so the associations between perceived norms and behavior must be re-evaluated during a period in which vaccines were more available to young adults. Secondly, Graupensperger, Abdallah, and colleagues' (2021) college student sample was collected at one university in a metropolitan area where vaccine uptake has been exceptionally high (Murthy et al., 2021) ; thus, research must be extended to a more geographically diverse sample, including more rural populations where vaccine hesitancy has been relatively higher (Kricorian et al., 2021) . Third, Graupensperger and colleagues' college study included few covariates. Notably, it has since become evident that COVID-19 attitudes and vaccine hesitancy has been a highly politicized issue in the US, such that those who identify as a Democrat have much more favorable attitudes toward COVID-19 vaccines than those who identify as a Republican (Fridman et al., 2021; Milligan et al., 2021) . Further, there is emerging evidence that other demographic variables may be related to COVID-19 vaccination uptake. For example, women (Detoc et al., 2020) and sexual minorities report stronger intentions to receive a COVID-19 vaccine than men and heterosexual individuals, respectively. Thus, identifying additional correlates to vaccine hesitancy necessitates re-evaluating the relative influence of perceived social norms while also accounting for key covariates in college students.

To address these gaps, the current study examined associations between college students' vaccine hesitancy and perceived descriptive and injunctive norms within the context of the COVID-19 pandemic. First, we aimed to set the stage by characterizing college students' varied and explicit reasons for their vaccine hesitancy during a time when vaccines were widely available for young adults. Although one study examined reasons for vaccine hesitancy among college students in November 2020 prior to vaccine availability (Graupensperger, Abdallah, et al., 2021) and another study examined reasons for vaccine hesitancy among US young adults in March 2021 during the early months of the public vaccine roll-out , we are aware of no studies to date that have examined explicit reasons for vaccine hesitancy among college students in the US since the vaccines were made publicly available.

Second, we aimed to clarify whether perceived social norms is a significant predictor of vaccine hesitancy even in the context of a multitude of explicit reasons for hesitancy. If so, findings would highlight the importance of perceived social norms for young adults, beyond the 5Cs model of vaccine hesitancy, and point to the potential for normative feedback to be an effective intervention strategy across a wide range of vaccine-hesitant young adults.

Perceived descriptive norms were operationalized as perceptions of vaccine uptake among people the participant "knows and talks to", similar to Latkin and colleagues' (2021) conceptualization of social norms as anticipated vaccine uptake among friends and family.

Perceived injunctive norms were operationalized as perceived approval of COVID-19

vaccines among the typical student at the participant's university. In both cases, perceived norms pertain to a referent group that is relatively proximal to the participants, which tend to be more influential than distal referent groups (Lac & Donaldson, 2018) . Given geographic differences in vaccine acceptance (Salomoni et al., 2021) , this study builds upon existing findings by recruiting students from urban and rural universities in different US regions.

Moreover, the extent that perceived norms relate to vaccine hesitancy were examined aboveand-beyond the effect of a thorough set of covariates. Specifically, we hypothesized that students who perceived greater vaccine uptake (descriptive norms) and approval (injunctive norms) would be less likely to report vaccine hesitancy, even after controlling for demographic characteristics and other COVID-related experiences (e.g., personal history of COVID-19, perceived risk, fatigue and stress related to the pandemic). In this way, we aimed to make a conceptual contribution to the literature by clarifying that, in light of the diverse reasons for hesitancy during this pandemic, social norms continue to play a prominent role in vaccine uptake.

Participants were college students ages 18 and older who were recruited in the spring of 2021 from either psychology or human development departmental participant pools at four public universities in the US. The research was advertised as "a study on how the COVID-19 pandemic is affecting college students' lives, including mental health, drinking, and sexual experiences." Participants provided informed consent and received class research credit for their time. Individuals were required to pass at least 3 of 4 attention check questions to be included. A total of 1,016 students participated. Given the current focus on young adult ages 18-25 (Arnett, 2007) , we excluded 26 older individuals (ages 26-61). One participant was excluded for not completing vaccine behavior questions.

The final sample was 989 college students from universities located in a mid-size

Midwestern city (n = 444), a large Southern city (n = 229), a large Northwestern city (n = 176), and a rural Northwestern town (n = 140). At the time of survey administration, the COVID-19 vaccine had been made available to some adults, with certain groups (e.g., at-risk adults and healthcare workers) receiving priority, but the roll-out across the US was ongoing. 

Participant were asked whether they have received the vaccine for COVID-19. coded as vaccine hesitant if they had not received the vaccine and did not intend to get it as soon as possible. Individuals who were vaccine hesitant were asked to select which of four reasons best described why they did not intend to get the vaccine as soon as possible (e.g., "I have a medical condition for which the vaccine has not yet been tested", "I don't think the vaccine is necessary"); an "other" option was also provided with a text response box to detail their rationale.

Perceived descriptive norms were assessed with the question, "Of the people you know and talk to regularly, how many have received the COVID-19 vaccine?" Response options were 0 = None, 1 = Very few, 2 = Many, 3 = Almost all, and 4 = Everyone I know.

Perceived injunctive norms were assessed by asking participants how much they think the typical college student at their university approves of getting the COVID-19 vaccine.

Response options ranged from 1 = Strongly disapprove to 7 = Strongly approve.

Participants were asked about their current gender identity, sexual orientation, and race/ethnicity. Although multiple categories were assessed (as detailed in the participant section above), responses were collapsed into dummy-coded variables for analytic purposes.

Specifically, gender identity (Cahill & Makadon, 2014 ) as a man (i.e., "male" or "transgender man" = 1) was compared to individuals who identified as women, other gender identities, and those who declined to state their gender (= 0). Exclusively heterosexual orientation (= 1) was compared to any identity as not exclusive heterosexual (= 0; Kinsey Institute, 2011).

Dummy-coded variables were created to represent the racial/ethnic categories (Hispanic,

were also asked if their political affiliation was "Democrat", "Republican", "Independent", or "Other"; Democrat was the specified as the reference group. Participants' current living situation (response options: "sorority or fraternity house", "residence halls/dorm room", "offcampus (but not with parents)", "off-campus (with parents)", "other") was recoded to represent whether participants were living with parents off-campus (= 1) or not (= 0).

Participants were asked several questions related to their personal experiences during the COVID-19 pandemic. Participants were coded as having ever tested positive if they indicated they had been tested for COVID-19 and "tested positive at least once" (Holmes et al., 2020) . In recognition that not all individuals had access to COVID-19 tests during times when tests were in short supply, we also included individuals who endorsed having "been presumed to be positive for COVID-19 (for example, I had a known exposure and/or symptoms consistent with or had a positive antibody test." Participants were also asked if a close friend or relative had passed away from COVID-19 or related complications.

Each of these experiences were coded such that 0 = No and 1 = Yes.

Participants reported what they believed to be their "personal risk for getting COVID-19 (or getting it again)" (National Library of Medicine, 2021) with response options ranging from 1 = Very low to 5 = Very high. To assess fatigue, participants were also asked how strongly they agreed or disagreed with the statement "I am tired of taking precautions against COVID-19". Response options ranged from 1 = Strongly disagree to 5 = Strongly agree.

COVID-19-related stress was assessed with the COVID Stress Scales . Participants were asked 24 questions about worries they might have experienced over the past 7 days (e.g., "I am worried about catching the virus", "I had trouble concentrating because I kept thinking about the virus"). Response options ranged from 0 = Not at all to 4 = Extremely. Total scores were summed and Cronbach's alpha in the current study was .95.

To characterize participants' varied and explicit reasons for their vaccine hesitancy, descriptive characteristics were examined vaccine behaviors and reasons for vaccine hesitancy. Open-ended text responses for vaccine hesitancy reasons were coded using directed content analysis (Hsieh & Shannon, 2005) . The first and third author began by

independently reading the open-ended responses to familiarize themselves with the data.

During data review, preliminary themes were independently created and then the two authors met to discuss what emerged from the data. These themes were defined and refined and used to code the data. Few discrepancies emerged, and when they did, these were resolved through discussion.

To examine the role of social norms as predictors of vaccine hesitancy, a series of logistic regressions were conducted. First, unadjusted odds ratios were estimated for 

First, we descriptively examined vaccine behaviors. Across participants, 57.3% (n = 567) were fully vaccinated or in the process of completing all doses recommended. Another 13.7% (n = 135) intended to be vaccinated as soon as possible. The remaining 29.0% (n = 287) were vaccine hesitant, including 20.4% (n = 202) who indicated they might get vaccinated but not right away, and 8.6% (n = 85) who did not intend to ever get the vaccine.

The 287 participants who were vaccine hesitant selected all reasons that applied to their hesitancy. The most common reasons were wanting to see how it affects others in the community first (75.2%; n = 216) and not believing the vaccine was necessary (30.0%; n = 86). Others indicated they had a medical condition for which the vaccine had not yet been tested (7.0%; n = 20) or previously had a severe allergic reaction to vaccines (5.9%; n = 17).

Fifty participants (17.4%) indicated there was another reason for their hesitancy; 48 provided a text response.

See Table 1 for results of the content coding of the "other" reason text responses. The most frequently reported other reason was safety concerns (n = 17), including concerns that the vaccine was not yet approved by the U.S. Food and Drug Administration and fear of longterm side effects. Some reported intentions to get the vaccine later (n = 11) such as over summer break. Reasons for delaying included allowing others who need it to get the vaccine first or living in a country outside of the US where they perceived the vaccine supply to be limited or untrustworthy. Others reported ideological concerns (n = 9), including distrust of the government, "religious reasons", or family. Several participants reported general disinterest (n = 7), including not caring to get it, perceptions that the vaccine was not needed, or believing they would not have complications because they had tested positive for COVID-19 already. Finally, a few respondents (n = 3) reported a fear of needles.

Regarding perceived descriptive norms, the modal participant (47.9%; n = 474) perceived that "many" people they knew and talked to regularly had received the COVID-19

vaccine ( Regarding perceived injunctive norms, 82.7% (n = 818) perceived the typical student was at least somewhat approving of the vaccine (i.e., response of 5 or above on a scale from 1 to 7). Perceived injunctive norms were lower among vaccine hesitant individuals (M = 5.02, SD = 1.40) relative to others who had received or intended to receive the vaccine (M = 5.62, SD = 1.19), t(463.54) = 6.32, p < .001.

Associations between perceived social norms and vaccine hesitancy as revealed in logistic regression models can be seen in Table 2 (see Supplemental Table 2 for full model results with covariates). Unadjusted and adjusted estimates were similar, and revealed that even after controlling for demographic characteristics and COVID-related experiences, both perceived descriptive and injunctive vaccine norms were significant predictors of vaccine hesitancy. However, perceived descriptive and injunctive norms were correlated (r = 0.34, p < .001), and after controlling for descriptive norms, injunctive norms were no longer a significant predictor of vaccine hesitancy (OR = 0.89, p = .073). Descriptive norms remained a significant predictor of vaccine hesitancy, even after controlling for all covariates and injunctive norms (OR = 0.37, p < .001). Considered in aggregate, Tjur's (2009) R 2 indicated all model predictors, including both perceived descriptive and injunctive norms, explained 26.8% of the variance in vaccine hesitancy.

Building on research suggesting the importance of perceived social norms for vaccine intentions among US adults (Latkin et al., 2021) and college students (Graupensperger, Abdallah, et al., 2021) prior to the COVID-19 vaccine roll-out, the current study was conducted when vaccines had become more widely available and revealed that although there are varied explicit reasons for and predictors of vaccine hesitancy, perceived descriptive and injunctive norms continue to be important drivers of behavior. Thus, findings highlight the robust nature of vaccine-related social norms as a predictor of behavior in a geographically diverse sample of US college students. These perceived norms were each significant predictors after controlling for demographic characteristics (i.e., university, age, gender, sexual identity, race/ethnicity, political affiliation, living situation) and experiences related to COVID-19 (i.e., testing positive, known death, perceived risk, fatigue with precautions, stress). In fact, perceived social norms were the only COVID-related variables examined that were uniquely predictive of vaccine hesitancy, highlighting the importance of social influences on young adults' health behaviors (Rimal & Storey, 2020) .

When perceived descriptive and injunctive norms were considered simultaneously in a combined model, descriptive norms emerged as the only unique predictor of vaccine hesitancy. Similarly, in each model, the effect size was larger (i.e., further from an odds ratio of 1) for descriptive norms (ORs = 0.32 to 0.37) than injunctive norms (ORs = 0.71 to 0.89).

These findings may highlight the relative importance of perceived peer behavior over attitudes for vaccine uptake, which is consistent with research examining normative influences on other health behaviors, such as alcohol use (Lac & Donaldson, 2018; Neighbors et al., 2008) . However, these findings from the simultaneous model should be interpreted with caution given limitations in the measure and differences in normative referent groups.

Specifically, behaviors were evaluated for the people participants knew and talked to regularly; attitudes were evaluated for the typical college student at their university. Although both referent groups are proximal, it is perhaps unsurprising that a stronger effect was found for a potentially closer referent group, consistent with social norms literature in other domains (Cox & Bates, 2011; Stevens et al., 2021) . During the pandemic when nearly a third of students in the current sample were living with their parents, individuals that students talk to regularly may not just be a closer subset of a larger college student referent group, but may also include family and friends outside of college. More research is needed to evaluate the role of perceived descriptive and injunctive norms for several referent groups as related to vaccine uptake.

Consistent with research conducted prior to public availability of the vaccine in the US (Robertson et al., 2021; Tsai et al., 2021; Williams et al., 2020) , participants who remained hesitant when the vaccine was publicly available reported a range of explicit reasons for this hesitancy. Several reasons endorsed involved a social component. Three quarters of hesitant participants indicated they wanted to see how it affected others first,

indicating social feedback about the vaccine safety may be important. Several participants also wrote in that they were hesitant because their family was strongly disapproving. This sentiment is consistent with the previously discussed importance of a close referent group guiding behavior. Other reasons endorsed ranged far beyond social factors, including medical, religious, and mental health (e.g., phobias). Yet, it is among these same participants that social norms were, on average, important predictors of vaccine hesitancy -highlighting that social influences are an important driver of behavior.

Taken together, findings have important intervention implications. Social norms for health-related behaviors tend to be misperceived -individuals often perceive that others engage in more risky behavior (e.g., alcohol use; Cox et al., 2019) and less protective behavior (e.g., adhering to COVID-19 CDC guidelines; than is actually the case. These misperceptions are particularly common for less visible behavior that takes place in private -like vaccinations -and is inferred through partial information (e.g., via media) and projections of one's own beliefs (Rimal & Storey, 2020) .

Indeed, a recent study revealed that college students, on average, underestimated the proportion of peers that were intending to receive a COVID-19 vaccine (i.e., descriptive norms) and underestimated how important their peers felt that getting vaccinated was (i.e., injunctive norms) (Graupensperger, Abdallah, et al., 2021) .

Correcting normative misperceptions has been a promising harm-reduction strategy across a broad range of health behaviors (Dempsey et al., 2018) , and may also be a viable strategy to promote vaccine uptake. For example, personalized normative feedback interventions provide tailored feedback that contrasts individuals' perceived norms to actual norms, highlighting discrepancies (Dotson et al., 2015) . Personalized normative feedback interventions have been widely utilized for increasing health behaviors such as sun protection (Reid & Aiken, 2013) , and reducing problematic behaviors such as gambling (Peter et al., 2019) and alcohol use (Labrie et al., 2013) . It follows that, for young adults in particular, norm-correcting strategies may be a prudent approach for increasing vaccine uptake.

Strengths of the current study include the multisite data collection across geographically diverse public universities, assessment of vaccine behavior and attitudes during the initial roll-out of the 

Extending research conducted prior to public availability of the COVID-19 vaccine, 29.0% of college students across four universities were not vaccinated and did not intend to be immediately vaccinated in the spring of 2021. Students reported a wide range of reasons for this vaccine hesitancy, from safety and medical concerns, to perceptions that the vaccine was unnecessary, to ideological concerns regarding distrust and religion. However, across these varied reasons, and when controlling for demographic characteristics and COVIDrelated experiences, perceived descriptive and injunctive social norms for vaccine uptake emerged as important predictors of vaccine hesitancy. Descriptive norms (i.e., perceptions that people whom students talked to regularly were already vaccinated) emerged as having a particularly robust association with vaccine hesitancy. Findings indicate that correcting normative misperceptions (e.g., highlighting that more of one's peers are receiving the vaccine than believed) may be a viable strategy to promote vaccine uptake among young adults.

None. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. 

Side effects "The vaccine does not feel safe to me yet because we do not know the long term effects of these. This is why I will not let them put that into my body."

10 Lack of research/approval "I don't think the vaccine is accurate and safe, I believe a vaccine should take many trial and errors and that could take years." 7

International "I'm an international student from a third world country which has already started producing its own vaccines which to me is crazy so I'll get it once I'm in the U.S." Notes. N = 989 for the unadjusted model; N = 985 for the adjusted models (four participants were excluded due to missing data on political affiliation, one of whom also had missing data on predictors for known death from COVID and perceived personal risk). OR = Odds Ratio; CI = confidence interval. Bolded estimates are statistically significant at p < .05. Covariates for adjusted models included university, age, gender, sexual orientation, racial/ethnic identity, political affiliation, living with parents, history of COVID-19, friend/relative dying from COVID-19, perceived personal risk, tired of taking precautions, and COVID-related stress. Full results for adjusted models including estimates for covariates are shown in Supplemental Table 2 .

Young adult perspectives on COVID-19 vaccinations

The importance of social norms against strategic effects: The case of COVID-19 vaccine uptake

Challenges in creating herd immunity to SARS-CoV-2 infection by mass vaccination. The Lancet

Emerging adulthood: What is it, and what is it good for?

COVID-19 vaccine hesitancy -A scoping review of literature in high-income countries. Vaccines

Predicting COVID-19 vaccine take-up: Moving beyond demographics

Prepandemic cognitive function and COVID-19 vaccine hesitancy: Cohort study

Higher Education Center for Alcohol and Other Drug Abuse and Violence Prevention

Beyond confidence: Development of a measure assessing the 5C psychological antecedents of vaccination

Household Pule Survey (COVID-19)

The social brain in adolescence: Evidence from functional magnetic resonance imaging and behavioural studies

Sexual orientation and gender identity data collection in clinical settings and in electronic health records: A key to ending LGBT health disparities

Implementation of mitigation strategies for communities with local COVID-19 transmission

COVID Data Tracker

A focus theory of normative conduct: Recycling the concept of norms to reduce littering in public places

Alcohol use and misuse during the COVID-19 pandemic: A potential public health crisis? The Lancet Public Health

A strategic approach to COVID-19 vaccine R&D

Referent group proximity, social norms, and context: Alcohol use in a low-use environment

Do misperceptions of peer drinking influence personal drinking behavior? Results from a complete social network of first-year college students

Brief individual-focused alcohol interventions for college students

Confronting the Delta Variant of SARS-CoV-2, Summer 2021

A critical appraisal of the social norms approach as an interventional strategy for health-related behavior and attitude change

Stopping the spread of COVID-19

Intention to participate in a COVID-19 vaccine clinical trial and to get vaccinated against COVID-19 in France during the pandemic

Coronavirus cases on college campuses spike, linked to parties. CNBC: Health and Science

Stand-alone personalized normative feedback for college student drinkers: A meta-analytic review

Young adults' physical distancing behaviors during the initial months of the COVID-19 pandemic: Adherence to guidelines and associations with alcohol use behavior

New data shows young people need to take social distancing seriously: Younger people may be spreading the virus among themselves

Predicting and changing behavior: The reasoned action approach

COVID-19 and vaccine hesitancy: A longitudinal study

Rapid COVID-19 vaccine development

Social norms and vaccine uptake: College students' COVID vaccination intentions, attitudes, and estimated peer norms and comparisons with influenza vaccine

Multifaceted COVID-19-related stressors and associations with indices of mental health, well-being, and substance use among young adults

Changes in young adults' alcohol and marijuana use, norms, and motives from before to during the COVID-19 pandemic

Changes in college student alcohol use during the COVID-19 pandemic: Are perceived drinking norms still relevant? Emerging Adulthood

Young adults underestimate how well peers adhere to COVID-19 preventive behavioral guidelines

Longitudinal associations between perceptions of peer group drinking norms and students' alcohol use frequency within college sport teams

KFF COVID-19 Vaccine Monitor

Adolescents misperceive and are influenced by high-status peers' health risk, deviant, and adaptive behavior

Multidisciplinary research priorities for the COVID-19 pandemic: A call for action for mental health science. The Lancet Psychiatry

Three approaches to qualitative content analysis

Kinsey's heterosexualhomosexual rating scale. Kinsey Institute for Research in Sex, Gender, and Reproduction

Is the COVID-19 pandemic a high-risk period for college student alcohol use? A comparison of three Spring semesters

Compliance without fear: Predictors of protective behavior during the first wave of the COVID-19 pandemic

Loneliness: A signature mental health concern in the era of COVID-19

Unemployment rose higher in three months of COVID-19 than it did in two years of the Great Recession

COVID-19 vaccine hesitancy: Misinformation and perceptions of vaccine safety. Human Vaccines & Immunotherapeutics. Advance online publication

RCT of web-based personalized normative feedback for college drinking prevention: Are typical student norms good enough?

Testing competing models of injunctive and descriptive norms for proximal and distal reference groups on alcohol attitudes and behavior

COVID-19 vaccine delay: An examination of United States residents' intention to delay vaccine uptake

Unique predictors of intended uptake of a COVID-19 vaccine in adults living in a rural college town in the United States

Randomized controlled trial of a webdelivered personalized normative feedback intervention to reduce alcohol-related risky sexual behavior among college students

Explicit and intrinsic intention to receive covid-19 vaccination among heterosexuals and sexual minorities in taiwan

Normative misperceptions of peer seat belt use among high school students and their relationship to personal seat belt use

Factors associated with depression, anxiety, and PTSD symptomatology during the COVID-19 pandemic: Clinical implications for U.S. young adult mental health

College campuses and COVID-19 mitigation: Clinical and economic value

Are college campuses superspreaders? A data-driven modeling study

The trajectory of loneliness in response to COVID-19

Social norms information enhances the efficacy of an appearance-based sun protection intervention

Disaster Information Management Research Center (DIMRC)

A year into the pandemic, long-term financial impact weighs heavily on many Americans

COVID-19 vaccine acceptance: Influential roles of political party and religiosity

Disparities in COVID-19 vaccination coverage between urban and rural counties -United States

The relative impact of injunctive norms on college student drinking: The role of reference group

The loneliness pandemic: Loneliness and other concomitants of depression, anxiety and their comorbidity during the COVID-19 outbreak

Intimate partner aggression during the COVID-19 pandemic: Associations with stress and heavy drinking. Psychology of Violence

A meta-analysis of brief personalized feedback interventions for problematic gambling

The influence of social norms on flu vaccination among African American and White adults

R: A language and environment for statistical computing. R Foundation for Statistical Computing

Correcting injunctive norm misperceptions motivates behavior change: A randomized controlled sun protection intervention

REACT-1 round 12 report: resurgence of SARS-CoV-2 infections in England associated with increased frequency of the Delta variant

Construction of meaning during a pandemic: The forgotten role of social norms

Predictors of COVID-19 vaccine hesitancy in the UK household longitudinal study

Accelerated vaccine rollout is imperative to mitigate highly transmissible COVID-19 variants

Hesitant or not hesitant? A systematic review on global covid-19 vaccine acceptance in different populations

Addressing problems with alcohol and other substances among older adults during the COVID-19 pandemic

College students' COVID-19 vaccine beliefs and intentions: Implications for interventions

Predicting substance use at a youth mass gathering event: The role of norms and the importance of their source

Self-efficacy and HPV vaccine attitudes mediate the relationship between social norms and intentions to receive the HPV vaccine among college students

Development and initial validation of the COVID Stress Scales

Coefficients of determination in logistic regression models -A new proposal: The coefficient of discrimination

Time-dependent heterogeneity leads to transient suppression of the COVID-19 epidemic, not herd immunity

Acceptability of covid-19 vaccines and protective behavior among adults in taiwan: Associations between risk perception and willingness to vaccinate against covid-19

Loneliness in America: How the pandemic has deepened an epidemic of loneliness and what we can do about it. Harvard: Making Caring Common Project

Ten threats to global health in 2019

Weekly epidemiological update on COVID-19 -10

Towards intervention development to increase the uptake of COVID-19 vaccination among those at high risk: Outlining evidence-based and theoretically informed future intervention content

Vaccine hesitancy in the era of COVID-19: Could lessons from the past help in divining the future? Human Vaccines and Immunotherapeutics

A comparative study of chest computed tomography features in young and older adults with Corona Virus Disease (COVID-19)

"I am waiting to get it over summer break." 4 Let others first "Give time for people who are more vulnerable to get it first." 4

Distrust "I will not take a vaccine the government says to take." 3 Religion "For religious reasons." 4Family "My mother doesn't believe in the vaccine, and I'm trying to make her happy for the time being." 2

Don't care or want to "I really just do not care that much." 4Not needed "Basically the same as the flu shot--not necessarily required but recommended."

Among US adults, the highest rates of hesitancy to receive the COVID-19 vaccine are among young adults aged 18 to 25. Vaccine hesitancy is particularly concerning among young adults in college, where social interactions on densely populated campuses can lead to substantial community spread. Given that many colleges have opted not to mandate vaccines, identification of modifiable predictors of vaccine hesitancy -such as perceived social norms -is key to informing interventions to promote vaccine uptake. To address this need, we Keywords: COVID-19, coronavirus, SARS-CoV-2, vaccine uptake, vaccine hesitancy, social norms

☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: