Intellectual Humility and Responsiveness to Public Health Recommendations

Abstract

We examined the association between intellectual humility (IH)—a willingness to consider credible new information and alternative views and revise one’s own views if warranted—and adherence to experts’ health behavior recommendations in the U.S. during the COVID-19 pandemic. Study 1 (N = 541) results showed that people higher in IH are more likely to engage in recommended health behaviors (e.g., mask-wearing, social distancing)—even when controlling for political affiliation. Additional analyses focused specifically on mask-wearing produced initial evidence consistent with mediation of the IH-mask-wearing relationship by the beliefs that mask-wearing 1) is an effective way to slow the spread of COVID-19 and 2) protects others. Based on the pathway from IH to mask-wearing through a concern for others found in Study 1, Study 2 further examined the relationship between IH and prosocial tendencies. The results from Study 2 (Ns for correlation coefficients ranged from 265 to 702) showed an association between IH and several values and traits that reflect a concern for others (e.g., agreeableness, benevolence). These findings suggest that IH may influence behavior through both intra- and interpersonal mechanisms. Implications of these findings for the health-behavior domain are discussed.

Public health recommendations and mandates are an efficient and effective means of curbing the spread of infectious diseases. However, their effectiveness is reduced, potentially dramatically, when people do not comply (Kleinsinger, 2018). Adherence to recommendations is voluntary, relying on people’s willingness to sacrifice some measure of comfort and convenience in order protect themselves and others from disease. Adherence to public health mandates, though required, is often not closely monitored in the United States, and nonadherence is unlikely to result in external punishments for people who disregard them. As such, adherence to mandated behaviors is, in essence, voluntary. In short, the effectiveness of public health efforts to curb the spread of diseases such as COVID-19 and influenza is highly dependent on people’s willingness to voluntarily follow recommended or mandated behaviors.

The reasons for nonadherence to public health recommendations and mandates include psychological reactance, low trust in science, and lack of knowledge (Albarracin et al., 2021; Kleitman et al., 2021; Levkovich, 2020; Mallinas et al., 2021; Plohl & Musil, 2021; Stosic et al., 2021). Of relevance to the present research is a willingness by some people to trust their personal beliefs about disease transmission over experts’ opinions and recommendations. Indeed, people often claim with great certainty that their personal views about health information are more accurate than views offered by experts (Evans & Hargittai, 2020; Hoorens, 1994; Hoorens et al., 2022; Weinstein, 1987), disputing the effectiveness of proposed safety measures for which evidence is strong and clear. Their nonadherent position implies an unwillingness to admit they could be wrong in their personal views and experts could be right in recommending specific mitigation behaviors.

The COVID-19 pandemic has presented an unprecedented opportunity for behavioral scientists to examine the influence of dispositions, social influences, and health-related information on engagement in voluntary health-protective behavior. In the present study, we consider one such influence, intellectual humility, a recognition of one’s intellectual limitations, including an understanding that one could be wrong about important matters and, therefore, should thoughtfully consider credible and relevant information and change one’s view if warranted (Hoyle & Davisson, in press; Leary et al., 2017; Porter et al., 2022). Despite a surge in interest in the intellectual humility (IH) construct, to date IH has not been examined with reference to health information and behavior.

In the present research, we explore various processes by which IH might influence people’s adherence to public health recommendations and mandates. We focus on two possible mechanisms. The first concerns a greater tendency by people high in IH to rely on credible evidence to shape their opinions and views. The tendency of people high in IH to process information more readily and deeply is well-established (Krumrei-Mancuso et al., 2020; Leary et al., 2017; Porter et al., 2022). Intellectually humble people are more open to information that conflicts with what they believe and willing to reconsider their existing beliefs. For example, compared to people low in IH, people high in IH prefer information that provides a balanced as opposed to one-sided view (Leary et al., 2017). In addition, when presented with new information either opposing or matching their own opinion, people high in IH spend more time processing information opposing their own view and do so more carefully than information that matches their own view (Deffler et al., 2016). Importantly, people high in IH are not indiscriminately open to new information. Rather, high IH people are able to distinguish high from low quality evidence (Leary et al., 2017). In addition, intellectually humble people seem to be more skeptical about false information and motivated to investigate suspicious claims, including misinformation about COVID-19, through fact-checking (Koetke et al., 2021). IH should be related to greater consideration of logical evidence such as the effectiveness of mitigation behaviors. Because the evidence favoring recommended and mandated mitigation behaviors during the COVID-19 pandemic is of high quality and endorsed by people with relevant expertise, we would expect people high in IH to be more likely to follow recommendations and mandates. This is likely due, in part, to their willingness to believe expert opinion that following these measures will protect themselves and others.

The second mechanism by which IH might influence people’s willingness to follow public health recommendations and mandates is consideration of the impact of their personal behavior on the health of others. Validation studies of IH measures have suggested that IH is associated with a number of prosocial tendencies, including perspective taking, empathy, gratitude, altruism, and benevolence (Krumrei-Mancuso, 2017; Krumrei-Mancuso, 2018; McElroy et al. 2014;). In particular, people high in IH are, on average, more agreeable and kinder than their low IH counterparts (Krumei-Mancuso, 2018; Leary et al., 2017). Because mitigation behaviors could be motivated by concern for others as well as oneself (Chan, 2021), we reasoned that people high in IH, because they are somewhat higher on characteristics that reflect a concern for others, would be more likely than their low IH counterparts to follow public health recommendations and mandates issued during the COVID-19 pandemic.

The Present Studies

In light of evidence that IH is related to both evidence-based reasoning and prosocial intentions, the present study tests the hypothesis that IH is positively associated with engagement in COVID-19 protective behaviors. In Study 1, we surveyed a large sample of adults about their engagement in recommended or mandated behaviors aimed at curbing the spread of COVID-19. In addition to reporting their engagement in those behaviors, participants provided information about their views on wearing a face covering (i.e., mask-wearing). These views assessed beliefs about who has the right to choose to wear a face covering (e.g., individuals, the government, businesses), the effectiveness of mask-wearing in slowing the spread of COVID-19, and who benefits from mask-wearing (e.g., oneself, other people). Of these views on masking, believing that wearing a mask is an effective way to slow the spread of COVID-19 suggests greater attention to high quality, credible information. Believing that wearing a mask is a collective choice to protect others reflects a prosocial orientation. Participants also completed a brief measure of IH. We used these data to examine the simple associations between IH and a set of COVID-19 protective behaviors and, for mask-wearing, two pathways by which IH might be associated with a greater tendency to wear a mask, a simple and effective means of curbing the spread of COVID-19: (a) reliance on credible information and (b) prosocial tendencies. In Study 2, we drew on unpublished data from our own prior work to examine the association between IH and a set of values and traits consistent with a tendency to be concerned for others.

Study 1

In Study 1, we first examine the simple associations between IH and a set of health-protective behaviors recommended or mandated by health experts in the U.S. during the COVID-19 pandemic. We expected positive associations between IH and engagement in these behaviors, particularly those for which evidence of effectiveness is strongest and which have the potential to protect others as well as oneself. We collected additional information about beliefs related to mask-wearing and used these data in exploratory analyses designed to examine the two proposed pathways from IH to adherence.

Method

Between April 3, 2020 and January 1, 2021, we recruited parents or guardians from households that included an adolescent participant in an ongoing longitudinal study to complete an online survey assessing demographics, individual differences, and experiences with COVID-19. Participants who completed this initial survey—72% did so in April or May of 2020—received a $30 gift card to the retailer of their choice. Between August 3, 2020 and January 14, 2021¹, all respondents to the initial survey received a supplemental survey that included items about pandemic-related protective behaviors. Participants who completed the supplemental survey—82% did so in August of 2020—received a $10 gift card. All procedures and materials were approved by the Duke University Institutional Review Board (protocol #2017-0332).

Participants

All households in the parent study from which adolescent data were obtained (N = 891) were invited to participate. A parent or guardian from 690 households completed the first survey; a subset of 557 completed the supplemental survey as well. Six participants did not provide adequate usable data on the first survey (i.e., stopped responding before reaching the COVID-19 items at the end of the survey) and therefore were excluded from the sample. An additional nine participants were excluded because their data on the supplemental survey were unusable (i.e., stopped responding before reaching the protective behavior section at the end of the survey). An additional 11 participants were excluded because their age (14–21 years) suggested that they were an adolescent participant in the longitudinal study rather than the parent or guardian. Because the analyses required variables from both surveys, the analysis sample included only those participants with usable data on both surveys (N = 541).

Age ranged from 30 to 70 years (M = 47.01, SD = 6.78). Approximately 90% of the participants identified as female. In terms of race/ethnicity, 67% of the participants identified as White, 20% as Black, and 12% identified as another race/ethnicity. For analyses that include race/ethnicity as a covariate, two dummy codes comparing White and Black and White and Other were included in the model. Additional descriptive information about the sample is provided in the supplement, Table S1.

Measures

Demographics and Covariates

Participants reported their age, sex (coded as 0 = male, 1 = female), race/ethnicity, years of schooling completed, and monthly household income. They also reported their subjective social status (MacArthur Scale of Subjective Social Status; Adler et al., 2007). To account for political differences that might explain adherence, we assessed political ideology, political party affiliation, and percent of voters in the participant’s county of residence who voted for Donald Trump in the 2016 presidential election (descriptive statistics for these variables are provided in the supplement, Table S2).

Intellectual Humility

Individual differences in IH were assessed using the six-item General Intellectual Humility Scale (Leary et al., 2017). Example items are, “I reconsider my opinions when presented with new evidence” and “I accept that my beliefs and attitudes may be wrong.” Participants indicated their responses on five-point scales with endpoints labeled strongly disagree and strongly agree.

Engagement in Protective Behaviors

Participants responded to five items assessing the degree to which they engaged in protective behaviors aimed at reducing COVID-19 transmission: staying home (“I am complying with the rules and suggestions of the government and healthcare system to remain at home to try to contain the virus”), wearing a face mask (“I wear a face covering in public places or indoors around people not in my household”), keeping physical distance (“I do my best to keep a distance of 6 feet between me and people not in my household”), hand washing (“I wash my hands before/after eating, after visiting public places, and after handling mail/packages”), and avoiding touching one’s face (“I touch my face with my hands”). The five-point response scale for the staying home item ranged from not at all true of me to very true of me. The items assessing masking, physical distancing, and hand washing used the same five-point response scale of never, rarely, some of the time, most of the time, and always. The five-point scale for the face touching item scale was anchored by never and very often and reverse-scored to reflect avoidance of face touching.

To assess avoidance of risky situations, participants were provided with a list of 20 activities that involve at least some risk of exposure to COVID-19. These ranged from relatively low risk activities such as “get takeout from a restaurant” and “eat at a restaurant outdoors/on a patio” to activities relatively high in exposure-risk such as “attend a wedding or funeral” and “eat at an indoor restaurant” (the complete list is included in the supplement, Table S3). The list was preceded by the question, “What is the likelihood you will do any of the following activities while the coronavirus pandemic is ongoing?” to which participants chose a response from 1 (highly unlikely) to 5 (highly likely). To create a composite measure of risk avoidance, we reversed likelihood ratings, weighted each rating by the degree of exposure risk, and created a sum. Weights were assigned based on published ratings by public health experts (e.g., Texas Medical Association COVID-19 Task Force, Michigan State Public Health Experts, and major print news organizations), which were available for 17 of the 20 activities. Weights reflecting exposure risk ranged from 1 to 5 (low, low-moderate, moderate, moderate-to-high, high). Additional information is provided in the supplement (see note to Table S3).

Views on Masking

After reporting on their engagement in protective behaviors, participants reported their specific views on masking. Participants indicated the extent of their agreement (1, strongly disagree, to 5, strongly agree) with whether mask-wearing is (1) an individual choice that protects my health, (2) a collective choice to protect others/society, (3) something I have the right to choose whether or not to do, (4) something that government officials have the right to require people to do, (5) something that businesses have the right to require people to do, and (6) an effective way to prevent the spread of COVID-19.

Results and Discussion

For descriptive purposes, we first examined the correlations between IH and the set of demographic variables and potential covariates. IH was not significantly correlated with age (r = .01, p = .719) or gender (r = .00, p = .921). IH was significantly but weakly correlated with years of education (r = .11, p = .006) and monthly income (r = .09, p = .030). People high in IH tended toward more liberal political ideology (r = .24, p < .001) and reported less social mobility (r = −.11, p = .006). The correlation between IH and socioeconomic status was near zero and nonsignificant (r = −.03, p = .430). A multiple regression analysis predicting IH from dummy coded political party affiliation indicated a significant association (multiple R = .27, p < .001), with Republicans scoring lower (M = 3.30) on IH than Democrats (M = 3.75) or Independents (M = 3.73). A similar analysis for dummy coded race/ethnicity revealed a weak and nonsignificant association (multiple R = .10, p = .103). A multilevel model predicting IH from percentage of votes for Donald Trump in participants’ county of residence in the 2016 election accounting for clustering by county revealed a near-zero association (r = .02, p = .660).

Descriptive statistics for IH and the mitigation behaviors and correlations between them are shown in Table 1. Except for avoiding face touching, self-reported compliance with the mitigation behaviors was high. The average IH score was half a scale point above the midpoint, slightly lower than observed in prior uses of the General Intellectual Humility Scale (Leary et al., 2017). IH correlated with extent of engagement in four of the six mitigation behaviors. The pattern of correlations between behaviors show moderate positive associations between staying at home, social distancing, wearing a face covering, and avoiding risky situations. Washing hands and avoiding touching one’s face were weakly correlated with each other and the other mitigation behaviors.

Table 1.

Means, Standard Deviations, and Intercorrelations of Intellectual Humility and Self-Reported COVID-19 Mitigation Behavior

Variable	M	SD	IH	Home	Distance	Wash	Face	Mask
Intellectual humility	3.61	0.69
Stay at home	4.34	0.97	.14
Social distancing	4.46	0.73	.16	.19
Washing hands	4.55	0.63	.06†	.10	.23
Avoiding touching face	3.25	0.92	−.00†	.13	.14	.25
Wearing face covering	4.56	0.81	.17	.31	.41	.27	.21
Avoiding risky situations	11.16	3.50	.20	.25	.43	.26	.19	.43

Note. Means could range from 1 to 5 except for avoiding risky situations, a weighted mean, which could range from 3.35 to 16.76. All correlation coefficients differ from zero at p < .001 except the two with the † superscript.

Ns ranged from 612 to 621.

Associations for staying at home (sr = .16, p < .001) and social distancing (sr = .12, p = .006) changed little with the addition of the covariates (age, sex, race/ethnicity, years of education, monthly household income, and political party affiliation). Associations for wearing a face covering (sr = .11, p = .014) and avoiding risky situations (sr = .08, p = .039) were reduced but remained significant. The lone significant predictor among the covariates was party affiliation, which descriptive analyses showed to be correlated with IH. We retained party affiliation as a covariate in subsequent analyses focused on wearing a face covering.

As described in the method section, we measured six views associated with mask-wearing. These items allowed us to explore potential mechanisms that underlie the association between IH and wearing a face covering. Descriptive statistics and intercorrelations among the six views are shown in the supplementary materials (Table S4). We first regressed self-reported mask-wearing on the six views and IH controlling for party affiliation. (Though some of the views about mask-wearing were highly correlated with each other, regression diagnostics show that this is not distorting the regression results. The largest value of the variance inflation factor was 3.14 (generally, values greater than 5 are considered potentially problematic). The smallest tolerance value was .32 (generally, values less than .25 are considered evidence of distorted estimates of regression weights). The model was highly significant, F(10, 583) = 29.28, p < .001, explaining about a third of the variance in mask-wearing (R² = .33). Among the six views, the view that mask-wearing protects others (b = .18, p < .001) and is an effective way to prevent transmission (b = .32, p < .001) were significant predictors. Consistent with a statistically mediated effect, the IH effect was nonsignificant (b = .02, p = .641). To model that pattern, we moved to structural equation modeling.

We estimated a model in which each of the six views about wearing a face covering mediate the relation between IH and mask-wearing controlling for the other views. The model and results are displayed in Figure 1. As was evident in the multiple regression model, the direct path from IH to mask-wearing is no longer significant with views in the model. Although IH was related to five of the six views, only two of those views were significantly related to mask-wearing. Specifically, the IH-mask-wearing relation was mediated by the view that a face covering should be worn to protect others (p = .005 for test of indirect effect) and that wearing a face covering is an effective way to limit the spread of COVID-19 (p = .002 for test of indirect effect).

Figure 1.

Mediational model exploring the potential role of beliefs about wearing face covering in the association between intellectual humility and wearing face covering. Path coefficients are standardized maximum likelihood estimates. Paths indicated by broken lines are nonsignificant. Significant mediators are set in bold. Paths from intellectual humility to beliefs are controlling for political party affiliation.

The results show that high IH people are more likely to follow health behaviors recommended or mandated during the COVID-19 pandemic. Further examination of masking behavior in particular showed that the relationship between IH and mask-wearing is mediated by believing that wearing a mask is an effective way to slow the spread of the coronavirus and protects others. These findings align well with previous research identifying perceptions of effectiveness and other-focused identity as motivators in taking collective action (e.g., van Zomeren et al., 2008).

Our interpretation of these results assumes that participants’ beliefs in the effectiveness of wearing a mask shows their reliance on credible evidence. Though the effectiveness of face masks in slowing the spread of the coronavirus is evidenced by credible research endorsed by infectious disease experts, people may believe mask-wearing is effective even if that belief does not originate from exposure to credible evidence. For example, people may believe wearing a face mask is effective because an authority figure told them so. Even so, attention to the effectiveness of mask-wearing suggests at least some processing of evidence rather than a sole reliance on peripheral features of the belief. Moreover, the documented tendency for high IH people to discriminate between high- and low-quality information; true and false information; and credible and non-credible information sources suggests that the link between IH and beliefs about effectiveness reflects attention to credible evidence (Koetke et al., 2021; Krumrei-Mancuso et al., 2020; Leary et al., 2017; Porter et al., 2022). However, additional evidence is required to support this assumption.

A potential limitation of Study 1 is the reliance on single items to assess engagement in protective behaviors and views on masking. Due to the ever-changing nature of the pandemic and respective health recommendations, as well as the timing of the data collection, using ad hoc measures was necessary as no validated measures were available. Future research may benefit from establishing and using multi-item measures to assess these self-reported beliefs and behaviors.

In addition, the sample in Study 1 was largely female. Research has shown that women are more likely than men to comply with recommended health behaviors in general (Sobol et al., 2020; Vaidya et al., 2012). Moreover, this same pattern of gender differences has been shown with health recommendations specific to the COVID-19 pandemic (e.g., hand washing, staying at home)—even though COVID-19 poses a greater risk of mortality to men than women (Chen et al., 2020; Lin et al., 2021; Wenham et al., 2020). However, research has not shown consistent gender differences in intellectual humility (Zmigrod et al., 2019), self-reports or behavioral measures of prosociality (Olsson et al., 2021), or individual differences in a tendency to process information deeply, such as need for cognition (Cacioppo & Petty, 1982). Nevertheless, future research should examine whether the links between IH and engaging in health-protective behaviors are the same for men and women.

Study 1 revealed the importance of a concern for others in the relationship between IH and mask-wearing, such that people high in IH were more likely to wear a face covering if they believed doing so protects others. In Study 2, we turn to unpublished data from our previous research to more closely examine the relationship between IH and prosocial tendencies.

Study 2

In Study 2 we examined more closely the association between IH and consideration for others as reflected in the IH-related belief that mask-wearing protects others. Though past research has begun to establish a connection between IH and a concern for others (e.g., Krumrei-Mancuso, 2017), all of this research has been conducted with different measures of IH than the present research and most has focused on IH in the specific context of religious leadership (Krumrei-Mancuso, 2018; McElroy et al. 2014). To strengthen evidence favoring a pathway from IH to adherence to recommendations and mandates that protect others as well as oneself, we computed relevant correlations from our own unpublished data. The goal of these analyses was to strengthen the evidence of a connection between IH and concern for others in a broader context and to do so using the measure of IH we used in Study 1.

Method

Participants

Participants were college students enrolled in a four-year longitudinal study of student resilience and well-being (Hoyle et al., 2021). Sample sizes varied across assessments from which data were extracted for analyses presented here. Sample size at the first assessment was N = 702. A subset of these participants was enrolled in the longitudinal study and provided relevant data at the seventh (N = 277) and eighth (N = 265) assessments. At the first assessment, 61% of the sample was women. Participants could select more than one race/ethnicity category; 48% selected White, 28% selected Asian, 7% selected Black, and 18% selected Other, which included Multiracial.

Procedure and Measures

Participants completed a battery of questionnaires administered online each semester across four years, for a total of eight timepoints of data collection. On three of those occasions, they completed a measure of IH and one or more measures of individual differences in the general tendency to feel concern for others. About two months into their first year (Time 1), students completed measures of IH, personality, and values. Approximately two months into their fourth year (Time 7), students completed the same measures of IH and values along with a brief measure of prosociality. About one month before the end of their fourth year (Time 8), participants completed the IH and personality measures. IH was not assessed at other timepoints.

Intellectual humility.

As in Study 1, participants completed the General Intellectual Humility Scale (Leary et al., 2017). At the first, seventh, and eighth assessments, internal consistency estimates using coefficient alpha were .79, .82 and .85, respectively. Agreeableness. We assessed agreeableness using nine items from the Big Five Inventory (BFI; John et al., 1991). Example items, which followed the stem, “I am someone who …,” are “Is helpful and unselfish with others” and “Is considerate and kind to almost everyone.” The five-point response scale was anchored by strongly disagree and strongly agree. Internal consistency estimates of reliability were α = .80 and .83 at the first and eighth assessment, respectively.

Selfism.

Selfism items were written in the same format and with the same response scale as the BFI items and interspersed with other items constituting the scale. The eight items, which reflect a reinterpretation of a sixth personality domain sometimes labeled honesty-humility (Diebels et al., 2018), were written by members of the research team (Leary & Diebels, 2014). Example items are “Puts my own wants and needs first” and “Looks out for myself before others.” Internal consistency estimates at the two assessments were .79 and .78.

Universalism.

Participants completed an abbreviated version of the Portrait Values Questionnaire (PVQ; Schwartz et al., 2001). The PVQ asks respondents to rate their similarity to someone who holds particular values on five-point scales anchored by not at all like me and extremely like me. An example of the three expressions of universalism read, “This person thinks it is important that every person in the world be treated equally. This person wants justice for everybody, even for people he/she doesn’t know.” Internal consistency estimates were .57 and .62 at the first and seventh assessments, respectively.

Benevolence.

Two PVQ items expressed the value of benevolence. An example item is, “It’s very important to this person to help the people around him/her. He/she wants to care for other people.” Values of coefficient alpha for the two-item set were .56 and .59, respectively.

Prosociality.

The questionnaire battery administered at the beginning of students’ fourth year included a single item from the prosocial subscale of the Self- and Other-Interest Inventory (Gerbasi & Prentice, 2013): “I am concerned with the overall best interest for everyone.” The seven-point response scale ranged from strongly disagree to strongly agree.

Results and Discussion

The correlations between individual differences in IH and constructs reflecting concern for others are presented in Table 2. The correlation for universalism, which reflects “understanding, appreciation, tolerance, and protection for the welfare of all people” (Schwartz, 2012) is moderate in magnitude despite the clear conceptual distinction in measurement and conceptualization of IH and the value. The remaining coefficients are small but suggest a clear association between IH and different expressions of the tendency to consider others (or, with selfism, to not routinely consider self before others). Coupled with findings from the few published studies that have considered IH with reference to these variables (e.g., Krumrei-Mancusco, 2017; Porter et al., 2022), these results support the consideration of a pathway from IH to behaviors that protect oneself and others that runs through the general tendency to consider the welfare of others. These results are based on the measurement of IH using the brief General Intellectual Humility Scale, extending prior work and validating the use of that measure in Study 1.

Table 2.

Correlations between General Intellectual Humility and Individual Differences that Reflect Consideration of Others

Variable	Time 1	Time 7	Time 8
Agreeableness	.25		.21
Selfism	−.23		−.22
Universalism Value	.36	.32
Benevolence Value	.17	.22
Prosociality		.15

Note. All coefficients differ from zero at p < .001 except the Time 7 correlation with prosociality, for which p = .01. Empty cells indicate that the variable in that row was not assessed at that time point. Time 1 was first semester of first year. Time 7 was first semester of fourth year. Time 8 was second semester of fourth year. IH was not assessed at other time points.

General Discussion

We examined the strength of and potential explanations for a connection between intellectual humility and adherence to health-protective behaviors recommended or mandated by scientists and public health officials acting on scientific evidence in Study 1. We found small-to-moderate associations between IH and four of the six mitigation behaviors considered. Subsequent analyses focused specifically on mask-wearing examined the association between IH and specific views about mask-wearing that might account for the IH-mask-wearing relation. Consistent with the findings showing an association between IH and a general concern for others, we found a positive association between IH and the view that mask-wearing protects others. In addition, we found a positive association between the view that mask-wearing protects others and mask-wearing behavior. We also found a positive association between IH and the view that mask-wearing is effective at curbing the spread of the virus and a positive association between this view and mask-wearing behavior. These findings extend a growing body of work on IH to the health-behavior domain and highlight two mechanisms by which IH might influence health-protective behavior. Study 2 further investigates the relationship between IH and prosociality suggested in Study 1 and demonstrated a reliable association between IH and several dispositions and values reflecting a concern for others.

Despite having emerged as a topic of empirical inquiry only a decade ago (for a brief history, see Davis et al., 2019), the intellectual humility construct has been examined in relation to a broad range of consequential attitudes and behaviors (for a review, see Porter et al., 2022), with a particular focus on politics, religion, and social issues (e.g., Bowes et al., 2020; Huynh & Senger, 2021; Krumrei-Mancuso & Newman, 2020; Krumrei-Mancuso & Newman, 2021; Senger & Huynh, 2021; Vaupotičet al., 2022). Our work demonstrates the relevance of IH for health behavior, specifically behavior focused on protecting the health of oneself and others when recommended by scientists and policy-makers acting on scientific evidence. The small-to-moderate associations between IH and the four health-protective behaviors for which the evidence of effectiveness is strongest held when accounting for alternative explanations for engaging in those behaviors.² Moreover, those associations spanned a set of behaviors that were only modestly associated with each other. Of particular note is that the association between IH and the politically-polarized behavior of mask-wearing remained significant controlling for the well-documented effect of political party affiliation (e.g., Bruine de Bruin et al., 2020; Kerr et al., 2021). The COVID-19 pandemic provided a unique opportunity to examine the role of IH in adherence to recommended and required health-protective behaviors. However, these COVID-19-specific behaviors differ from most health-protective behaviors in that mandates, strong social norms, and political polarization played large roles. Future research should examine how IH relates to engagement in other health-protective behaviors that are not affected by these factors.

The additional information we collected on people’s views of mask-wearing allowed us to examine potential explanations for the IH-mask-wearing association. Although the results of our mediational analyses do not allow for causal inferences given the research design, they point to two potential explanations that warrant more focused and rigorous research in the future. The least surprising of the potential explanations is the view that mask-wearing is an effective means of preventing the spread of COVID-19. This view is consistent with evidence published in the scientific literature (e.g., Andrejko et al., 2022; Leech et al., 2022) and popular media (e.g., Parker-Pope & Sheikh, 2021). It also is consistent with research showing that people high in IH are more likely than their low-IH counterparts to attend to credible information, even when it is inconsistent with their current knowledge or understanding (Deffler et al., 2016; Leary et al., 2017). During a period in U.S. history when trust in scientists and their work is declining (Kennedy et al., 2022), policy makers and health communicators face the difficult challenge of promoting evidence-based health behavior. The psychological mechanisms that contribute to humble thinking may provide clues to encourage thoughtful processing of public health information and recommendations. Future research should examine more closely why intellectually humble people were more likely to believe that wearing a mask is an effective way to slow the spread of the virus. This finding could be due to multiple features of deep processing, such as being more attuned to the logic of an argument or being more likely to listen to credible sources.

A potential novel explanation is the view that mask-wearing protects other people from infection. Although IH as typically defined is primarily an explanation of how people interact with information relevant to their beliefs and understanding, scattered evidence in prior work suggests that it is correlated with prosocial intentions (e.g., Krumrei-Mancuso, 2017; Porter et al., 2022). Evidence from our second study supported this connection, showing small-to-moderate correlations between IH and several values and traits that reflect a general concern for others. Our exploratory mediational analyses showed that both IH and mask-wearing were associated with the view that mask-wearing protects others, suggesting a potential mechanism by which IH influences certain health behaviors. This finding supports a compelling causal hypothesis that should be examined in future research—that IH causes compliance with public health directives because people high on IH have a greater concern for the well-being of others. Importantly, it suggests that any causal influence of IH on health-protective behavior might be attributable both to intra- and interpersonal mechanisms. And our findings suggest that, for political charged behaviors such as mask-wearing, its influence is independent of political party affiliation. Finally, the present research relied on self-reports of attitudes and behaviors. Future research should further examine these relationships behaviorally.

In summary, analyses of survey data from a large sample of adult community members showed a significant association between intellectual humility and responsiveness to public health recommendations and mandates during the COVID-19 pandemic. The analyses also suggested two potential explanations for the association involving mask-wearing, extending the growing literature on the relevance of IH for specific attitudes and behavior. In light of this work, researchers have begun to consider potential interventions that aim to increase IH (see Porter et al., 2022 for current approaches to intervention development) in the hope that an increase in IH will be accompanied by more thoughtful consideration of information, including its sources, and, as a result, that attitudes and behaviors are both well-informed and subject to updating as new information becomes available. Our findings suggest that such interventions hold promise as a means of promoting behavior that protects self and others from transmissible health threats as the likelihood of their occurrence increases (Carlson et al., 2022).

Highlights.

• Intellectually humble people more likely to follow expert-recommended behaviors

• Intellectual humility related to greater mask-wearing during COVID-19 pandemic

• Intellectual humility-masking link mediated by beliefs it cuts spread, helps others

• Intellectual humility related to values and traits that reflect concern for others