Humility in novice leaders: links to servant leadership and followers’ satisfaction with leadership

ABSTRACT Across two studies, we investigated connections between leader humility, servant leadership, and follower satisfaction. Leaders were resident and spiritual life advisors in on-campus housing at a university. In Study 1, we assessed leaders’ self-reported expressed humility, intellectual humility, and servant leadership at three times across six months. Leader humility did not change across time, but aspects of both forms of leader humility were positively associated with servant leadership. In Study 2, leaders were rated by their followers. Leaders reporting more respect for diverse viewpoints, a characteristic of intellectual humility, had followers who were more satisfied with their leaders’ interpersonal leadership and justice orientation. This research suggest potential benefits to cultivating humility among novice leaders in some contexts.

Humility has been a topic of interest in leadership studies. Collins (2001), for example, discovered companies that ascend from 'good to great' stock values have CEOs with a paradoxical combination of humility and strong professional will. Leaders, and the groups they manage, may languish if leaders are self-centered or arrogantly overestimate their knowledge and ability. In contrast, leaders who recognize the limits of their knowledge and an are open to learning are well-positioned to attend to their followers and become better leaders. This topic has often been examined in corporate settings, where leaders who express humility have been shown to increase team performance (Owens & Hekman, 2012;Rego et al., 2017) and subordinates' job satisfaction (Owens et al., 2013).
In the current research, we examined humility among novice leaders in a community leadership context, where followers may particularly benefit from leaders who are caring and willing to work behind the scenes. Our goal was to assess whether leader humility would be associated with leaders engaging in more servant leadership across time, and whether leader humility would predict followers' satisfaction with leaders' performance.

Humility defined and its measurement
Broadly defined, humility involves being down-to-earth, unassuming, and willing to admit limits or mistakes (vis a vis being arrogant, self-centered, or conceited; Tangney, 2000;Van Tongeren et al., 2019). The terms humble and humility can be traced to Latin words humus 'earth' or humilis 'on the ground.' Humble persons are metaphorically well-grounded and don't think too highly (or lowly) about themselves. However, because secondary definitions of humility include servility or timidity, psychologists and philosophers are careful to distinguish authentic humility from selfabasement  and intellectual humility from intellectual servility (Haggard et al., 2018;Whitcomb et al., 2015).
In the following studies, we examined whether two forms of humility -expressed and intellectual humilitywere associated with servant leadership (Studies 1 & 2) and followers' satisfaction with their leaders (Study 2). Expressed humility (EH; Owens et al., 2013), a form of interpersonal humility in the social domain, involves a willingness to see the self accurately, expression of appreciation for others' contributions, and teachability (Owens & Hekman, 2012;Owens et al., 2013). Intellectual humility (IH) is marked by an awareness of the limits of one's knowledge and is often conceptualized as including a willingness to revise one's views given an evidentiary basis (Haggard et al., 2018;Krumrei-Mancuso & Rouse, 2016;Leary et al., 2017). Some have conceptualized interpersonal features of IH, such as a respectfulness of diverse viewpoints or a non-defensiveness about intellectual disagreements (Krumrei-Mancuso & Rouse, 2016;McElroy et al., 2014). EH and IH are unique in that EH represents the interpersonal expression of general humility, whereas IH is centered on people's beliefs about what they know. Yet, the constructs also have overlapping content, given that EH, as a general form of humility, encompasses some qualities of the subdomain of IH, and IH (like EH) can be expressed interpersonally. According to McElroy-Heltzel et al. (2019), measures of both EH and IH tap the same facets of: openness/lack of superiority, other-oriented unselfishness, and willingness to admit mistakes/teachable. We posit both qualities of humility could predispose one to engage in servant leadership and to be an effective leader.
We had leaders self-report their levels of humility -a method that has been a point of discussion in the literature. Some have raised concerns about selfreporting humility (e.g., Davis et al., 2010;Tangney, 2000). However, concerns have lessened over time (e.g., Worthington & Allison, 2018), as research has supported the construct validity of self-reported humility (Ashton et al., 2014) and humility measures have correlated negligibly with desirable responding (e.g., Haggard et al., 2018;Leary et al., 2017). In the current research, we examined whether self-reported humility would predict both self-and other-reported leadership qualities and outcomes, to evaluate whether links between humility and leadership would transcend common method variance.

Servant leadership defined and its outcomes
By definition, servant leaders are inclined to put others first and focus on followers' well-being (Greenleaf, 1977). Servant leaders patiently guide, attend to, and aid their followers. Several servantleader qualities-such as empathy, listening, a commitment to others' growth, and community building-can contribute to leaders' effectiveness (Spears, 2010). Servant leadership is also associated with stronger relationships between leaders and their followers, including greater trust and loyalty, as well as personal and professional benefits to followers (see Krumrei-Mancuso, 2020, for review). Much of this research has been conducted in for-profit organizations, but research from community settings mirrors these findings. For example, in an educational context in Indonesia, teachers' ratings of their supervisors' servant leadership strongly predicted trust/ loyalty (Sendjaya et al., 2017).

Theorized connections between humility, servant leadership, and satisfaction with leaders
There are theoretical and empirical reasons to hypothesize a connection between humility and the nature of leadership (Owens & Hekman, 2016). Below, we review prosocial correlates of humility that could predispose leaders to serve followers and to be appreciated by followers. We also consider the role of humility in the formation of social bonds and the smoothing of social interactions, which may be particularly relevant to the social aspects of leadership.

Humility and prosociality
Broadly, theorists posit humility promotes cooperation and reciprocal altruism in dyads and small groups (see Ashton & Lee, 2007). For example, humility correlates with numerous positive and prosocial qualities (Leman et al., 2016) implicated in interpersonal cooperation and teamwork (Ashton & Lee, 2007;Barrick et al., 2001) such as agreeableness, openness, perspective-taking, helpfulness (LaBouff et al., 2012), generosity (Exline & Hill, 2012), and social relationship quality (Peters et al., 2011). IH, broadly defined, has also predicted more prosocial values such as empathy, benevolence, and lower powerseeking (Krumrei-Mancuso, 2017). Because humility promotes a prosocial orientation to aid others, we predicted humility would be positively associated with servant leadership and followers' satisfaction with their leaders.

Humility, social bonds, and social oil
More specifically, we theorized humility, servant leadership, and follower satisfaction with leaders are connected, in part, because humility can facilitate social bonding  between leader and followers (cf. Van Tongeren et al., 2019). Humble leaders may care for those they serve in ways that create or strengthen bonds. In contrast, a sense of superiority could signal inaccessibility and prevent the formation of healthy leader-follower relationships.
Expressing humility could also function like a metaphorical social oil  that smooths social interactions and reduces interpersonal friction. Humility involves a lowering of the self that could be disarming and put others at ease. This may also extend to IH, as one study found that students with some instruction about IH engaged in more compromise seeking than those who did not receive instruction in IH (Meagher et al., 2019).
Literature from organizational settings suggests that leaders who express humility tend to validate followers, which can increase the perception of the leader as sincere and promote follower engagement (Owens & Hekman, 2012). Krumrei-Mancuso (2020) positioned humility as an operating mechanism of servant leadership that contributes to a host of positive outcomes, such as increasing mutual respect and trust and improving communication. EH and IH may be vehicles through which leaders are able to transcend self in a way that promotes taking an other-oriented approach to leadership (i.e., servant leadership).
Given conceptual connections between EH/IH and prosocial qualities, we theorized EH/IH would promote servant leadership. Further, we thought the social bonding and social oil hypotheses of humility would be particularly relevant to the interpersonal aspects of leadership, involving the relationship and interactions between leaders and followers. Further, humility may be especially impactful for novice leaders and peer leadership, which is relevant to this study of student leaders tasked with helping fellow students grow, develop, and get along in residential communities. In this setting, leader humility could aid in the development of secure social bonds, interpersonal cooperation, and sense of belonging. This is supported by a small pilot study in which new community leaders who responded with IH during the first few weeks on the job displayed more servant leadership over time (Krumrei-Mancuso, 2018).

Overview
In two studies, we investigated connections between leaders' humility and servant leadership (Studies 1 & 2) and followers' ratings of satisfaction with leader performance (Study 2). We hypothesized EH and IH would predict increased servant leadership over time (Study 1) and EH and IH would predict greater satisfaction with the leader's performance among followers (Study 2). Both studies were approved by Pepperdine University's Institutional Review Board (Protocol ID: 17-06-574). Participants were provided with written information necessary to give their consent to participate in the research prior to proceeding to the surveys.

Study 1
The aim of Study 1 was to examine whether EH and IH would be associated with more servant leadership across time among a naturally occurring sample of novice community leaders. We expected this to be the case on the basis of theory and a pilot study (Krumrei-Mancuso, 2018).

Participants
The sample consisted of 101 Resident and Spiritual Life Advisors hired by the Residence Life program of a Christian university to work 20 hours per week providing leadership in university campus housing. The response rate was 95% of the total pool of leaders. The leaders were students themselves and lived in designated housing areas where their responsibilities included building community, developing relationships with and caring for residents, providing peer counseling, providing spiritual support, organizing events, attending staff meetings, completing paperwork, enforcing community standards, and monitoring the residential area for safety and maintenance needs. The leaders completed a 16-day training program prior to commencing the leadership positions, in which they engaged in team building; received information about policies, procedures, and expectations; and practiced with applied scenarios relevant to their leadership responsibilities. Although servant leadership is consistent with the overall goals of the leadership positions, it was not explicitly incorporated into the hiring process or addressed during the training program.
The leaders ranged in age from 18 to 24 years (M age = 20.07, SD age = .96). The sample was 65.3% female and 34.7% male, mirroring the student body as a whole. The sample was 56.4% White, 20.8% Latinx, 14.9% Asian, 13.9% Black or African-American, and 7.9% multi-racial. Consistent with the hiring goals of the Residence Life program, all leaders identified as Christian.

Servant leadership
We made use of the 4-item Altruistic Calling subscale of the Servant Leadership Questionnaire (Barbuto & Wheeler, 2006), which reflects core aspects of servant leadership and is fitting for non-corporate, peer-based leadership. The scale was designed as an other-report measure, but used in a self-report format within the current study (sample item: 'I do everything I can to serve my residents.'). Items were rated on a Likert scale from (1) strongly disagree to (5) strongly agree.

Intellectual humility
We used the Comprehensive Intellectual Humility Scale (Krumrei-Mancuso & Rouse, 2016) for a broad assessment of IH that includes a focus on both intrapsychic and interpersonal aspects of being intellectually humble. The scale represents a higher order factor consisting of four subscales: (1) lack of intellectual overconfidence (e.g., 'When I am really confident in a belief, there is very little chance that belief is wrong.' reverse scored), (2) openness to revising one's viewpoint (e.g., 'I am open to revising my important beliefs in the face of new information.'), (3) respect for others' viewpoints (e.g., 'I welcome different ways of thinking about important topics.'), and (4) independence of intellect and ego (e.g., 'When someone contradicts my most important beliefs, it feels like a personal attack.' reverse scored). Items were rated on a Likert scale from (1) strongly disagree to (5) strongly agree.

Expressed humility
The Expressed Humility Scale assesses expressed behaviors demonstrating humility in the context of organizational leadership (Owens et al., 2013). The scale was designed as an other-report measure, but used in a self-report format in the current research. The scale represents a higher order factor consisting of three subscales: (1) willingness to view oneself accurately (e.g., 'I admit it when I don't know how to do something.'), (2) appreciation of others' strengths and contributions (e.g., 'I show appreciation for the unique contributions of others.'), and (3) teachability (e.g., 'I am willing to learn from others.'). Items were rated on a Likert scale from (1) strongly disagree to (5) strongly agree.

Social desirability
We used three items of the Socially Desirable Response Set-5 (Hays et al., 1989) to assess participants' tendency to respond to survey items in a way deemed favorable by society. Items were rated on a Likert scale ranging from (1) definitely false to (5) definitely true.

Procedures
Data were collected at four time points via online surveys. The first assessment (Time 0) took place during a leadership training program immediately prior to the start of the leadership positions. At this time, we gathered demographic and social desirability data. Servant leadership and humility were assessed at Times 1 through 3. Time 1 data were collected after 3 weeks of leadership, Time 2 data were collected after 6 weeks of leadership (3 weeks after Time 1), and Time 3 data were collected after 6 months of leadership (4.5 months after Time 2). We made use of multilevel analyses because time points (level 1) were nested within leaders (level 2). Table 1  Given that EH and IH tap overlapping as well as unique content, they were strongly correlated (r = .54, p < .01 at Time 1; r = .52, p < .01 at Time 2; r = .44, p < .01 at Time 3). For this reason, we examined collinearity statistics at each time point for all subscales of the two measures, but found that multi-collinearity was not problematic (Variation Inflation Factors ranging from 1.23 to 3.11 and Tolerance ranging from .32 to .81 across time points). Based on this we included EH and IH in the same models in the primary analyses.

Preliminary analyses
We examined EH, IH, and servant leadership differences on the basis of gender, race, ethnicity, age, and social desirability. The only significant finding was that t-tests revealed those who identified as Latinx displayed slightly more EH than individuals who did not identify as Latinx at Time 1 (M = 41.65 versus 40.51, p = .03) and Time 3 (M = 42.25 versus 40.28, p = .02). Based on this, we included ethnicity as a covariate. In addition, we made an a priori decision to include social desirability tendencies as a control variable, given that all data were gathered through self-report.

Leaders' humility as a predictor of servant leadership
We used SPSS Advanced Statistics 25 to conduct random effects linear mixed models to assess associations between humility and servant leadership, controlling cluster effects of time (level 1) nested within individuals (level 2). We used maximum likelihood estimation to facilitate likelihood ration tests to compare fit between models.
We mean centered all continuous variables to improve interpretability. We also mean centered ethnicity, so that the intercept would represent the value of servant leadership averaged between Latinx and non-Latinx participants. The covariates were included as fixed effects in the model. We used a subject identifier as a random effect to account for leaders' inter-individual differences in servant leadership.
The results are displayed in Table 2. The ICC of the unconstrained (null) model indicated that 26% of the variance in servant leadership could be explained by inter-individual differences of the leaders. In Model 1, we added variables representing each time point as fixed effects in the analysis, to evaluate linear trends of servant leadership. This indicated there was no significant change in servant leadership from Time 1 to Time 2 or 3. In Model 2, we added ethnicity and social desirability as control variables, which did not significantly improve model fit (χ 2 = .59, p = .74). In Model 3, we added all subscales of EH and IH. This significantly improved model fit (χ 2 = 47.21, p < .001). Two specific aspects of humility were predictive of greater servant leadership. For every unit increase in teachability (a characteristic of expressed humility), there was a .34 unit increase in servant leadership and for every unit increase in respect for others' viewpoints (a characteristic of intellectual humility), there was a .17 unit increase in servant leadership.
As a local effect size estimate for humility, we examined the reduction of variance components between Model 2 and Model 3, to examine how much of the variance in servant leadership that was attributable to inter-individual differences of the leaders was accounted for by adding EH and IH to the model. This indicated that after taking leaders' ethnicities and social desirability response tendencies into account, 15.92% of the variance in leaders' servant leadership could be accounted for by humility.

Discussion
In this study, we found positive associations between humility and servant leadership among novice community leaders in the first six-months of a leadership position, even when controlling social desirability and ethnicity. When examining all forms of humility together, being teachable, an aspect of expressed humility, was the strongest predictor of servant leadership. Notably, this involves being willing to learn and take advice from others, as well as being open to the ideas of others, which bridges to the concept of IH. Respect for others' viewpoints was also associated with more servant leadership. This is a characteristic of being intellectually humble that involves appreciating that there are diverse ways of thinking about topics as well as the ability to respect individuals who think differently from oneself. The relatively stronger covariance with servant leadership of the EH subscale than the IH subscale may be due to the fact that the measures of EH and servant leadership focus more on behaviors than the measure of IH, which includes items about beliefs and attitudes, in addition to behaviors. A humility-servant leadership connection fits with existing literature on the role of servant-leadership (Spears, 2010) and humility (Owens et al., 2013) in organizations. Further, EH and IH were not associated with social desirability, which is consistent with studies showing comprehensive measures of IH are negligibly correlated with desirable responding (Haggard et al., 2018;Leary et al., 2017).
We found small ethnic differences in humility, which could be explored in future research in the context of cultural humility (Hook et al., 2013). We were unable to locate any other empirical research about humility in Latinx communities or racial-ethnic differences in EH.
Use of a naturally-occurring sample of leaders and the longitudinal design are strengths of Study 1, however, we did not find systematic change in self-reported humility or servant leadership across time. The community leader training program was not explicitly designed to increase humility or servant leadership, and the sixmonth interval may not have been long enough to observe naturally occurring changes in these variables. A previous study examining whether teaching IH in a college course or not would increase IH also observed that self-reported IH did not increase over time, but peer-rated IH did increase over time (Meagher et al., 2019). Exclusive reliance on self-report is a limitation of Study 1. In Study 2, we examined whether leaders' selfreported humility would predict followers' estimates of their leaders' servant leadership and followers' reports of their satisfaction with their leaders' job performance.

Study 2
Study 2 had two aims. First, we examined whether the relationship between humility and servant leadership from Study 1 would persist when using followers' ratings of their leaders' servant leadership rather than leaders' self-reported servant leadership.
Second, we investigated whether leaders' humility would predict followers' satisfaction with their leaders' performance. We expected this to be the case in light of previous studies showing connections between leader humility and ratings of leaders' job performance Owens et al., 2013). We used several dimensions of followers' satisfaction with a leader as indicators of leader performance (i.e., satisfaction with the leaders' administrative leadership, interpersonal leadership, and justice orientation).

Participants
During the academic semester in which the leaders completed their 6-month follow-up assessment for Study 1 (Time 3), campus residents (subsequently referred to as 'followers') were surveyed about their campus leaders (six weeks after Study 1, Time 3 data collection). Leaders from Study 1 who received ratings from campus residents in the followers' survey were included in Study 2. This consisted of 78 of the leaders from Study 1. T-test indicated no differences in servant leadership, EH, or IH between the leaders from Study 1 who were and were not included in Study 2 (p's ranging from .30 to .99). The leaders in Study 2 ranged in age from 18 to 24 years (M age = 20.12, SD age = .97). The leaders identified as 59% female and 33.3% male. Racially, 52.6% identified as White, 15.4% as Asian, 15.4% as Black or African American, and 10.3% as multiracial. In addition, 12.8% of the leaders identified as Latinx.
The leaders were rated by 215 followers, with an average of 2.76 raters per leader (SD = 1.86). The raters were 54.3% female and 33.3% male. Racially, 56.4% of the followers identified as White, 18.5% as Asian, 8.6% as Black or African American, 2.5% as Native Hawaiian or other Pacific Islander, and 2.1% as American Indian or Alaska Native. An additional 14.4% of the followers identified as Latinx. The raters were 45.3% first-year students, 14.0% second-year, 14.4% third-year, 8.2% fourth-year, and 6.6% graduate students. In addition, 6.2% indicated they were international students. Table 3 summarizes the psychometric properties of the measures used in Study 2. For all measures, higher scores indicated greater levels of the construct.

Humility
We used the Study 1, Time 3 administrations of the Expressed Humility Scale (Owens et al., 2013) and the Comprehensive Intellectual Humility Scale (Krumrei-Mancuso & Rouse, 2016) as assessments of leaders' selfreported EH and IH, respectively.

Servant leadership
Followers rated their respective leaders on servant leadership with the Altruistic Calling subscale of the Servant Leadership Questionnaire described in Study 1 (Barbuto & Wheeler, 2006; sample item: 'This leader does everything he/she can to serve me.'). Items were rated on a Likert scale from (1) strongly disagree to (7) strongly agree.

Satisfaction with leader's performance
To evaluate leader performance, we used items completed by followers as part of the national housing and residence life survey administered by Skyfactor Benchworks. On the basis of theory, we created three categories of leadership represented within the items. A group of four researchers and staff members of Residence Life independently rated each of the items for category fit. This resulted in the following subscales: (1) satisfaction with the leader's administrative leadership (e.g.,: 'How satisfied are you with your leader's enforcement of policies?'), (2) satisfaction with the leader's interpersonal leadership (e.g.,: 'How satisfied are you with your leader's efforts to get to know you?'), and (3) satisfaction with the leader's justice orientation (e.g.,: 'How satisfied are you with your leader's promotion of tolerance of others?'). Items were rated on a Likert scale from (1) strongly disagree to (7) strongly agree.
We conducted a confirmatory factor analysis in Mplus version 8 (N = 514), which indicated a model with the three theoretically derived latent variables (administrative leadership, interpersonal leadership, and justice orientation) provided reasonable fit to the data: χ 2 = 86.23 (p < .05), RMSEA = .09, CFI = .98, TLI = .97, SRMR = .02. Further, this three factor model fit the data better than a single factor with the same items (AIC of 11,743.05 for the single factor model versus 11,621.93 for the three factor model and BIC of 11,844.86 for the single factor model versus 11,736.47 for the three factor model).

Procedures
Data were collected via online surveys. Followers rated their leaders in conjunction with a campuswide survey used to assess students' university and housing experiences. For the purpose of the current study, a supplement was added to the standard survey inviting followers to identify the leader in their housing area they knew best and complete measures about this leader. We made use of the leaders' selfreported EH and IH assessed at Time 3 of Study 1, as this assessment of humility was gathered in the same academic semester in which the followers completed their survey about their leaders. Followers rated their leaders six weeks after the leaders provided their selfratings. We made use of multilevel analyses because followers' reports (level 1) were nested within leaders (level 2).

Preliminary analyses
Descriptive statistics of the study variables are provided in Table 3. We examined EH, IH, and servant leadership differences on the basis of gender, race, ethnicity, age, and social desirability. Similar to Study 1, a t-test revealed that those who identified as Latinx displayed slightly more EH than individuals who did not identify as Latinx (M = 42.72 versus 40.11, p < .001). Based on this, we included ethnicity as a covariate in the primary analyses. No other differences were observed for EH, IH, or servant leadership on the basis of demographic factors or social desirability tendencies.

Leader-rated humility as a predictor of follower-reported outcomes
We used SPSS Advanced Statistics 25 to conduct random effects linear mixed models to examine whether the selfreported humility of leaders (level 2) was predictive of a number of follower-reported outcomes (level 1), including followers' experience of their leader as a servant leader and followers' satisfaction with the leader's performance. We used maximum likelihood estimation to facilitate likelihood ration tests to compare fit between models. We followed similar procedures to Study 1, including grand mean centering all variables to improve interpretability.
The results are displayed in Table 4. The ICCs of the unconstrained (null) models indicated that 7% to 29% of the variance in the outcomes could be explained by inter-individual differences between the leaders. This is fairly consistent with previous research, where ICC values in cross-sectional multilevel modeling in the social sciences tend to range from .05 to .20 (Peugh, 2010). In Model 1, we added leader ethnicity (level 2) as a control variable with fixed effects, which did not significantly improve model fit. In Model 2, we added our assessed characteristics of EH and IH. This significantly improved model fit for each outcome.
As a local effect size estimate for humility, we examined the reduction of variance components between Model 1 and Model 2, to examine how much of the variance in each outcome attributable to the clusters of followers rating each leader could be accounted for by the leaders' EH and IH. This indicated that after taking leaders' ethnicities into account, 51.72% of the variance in follower ratings of leaders' servant leadership could be accounted for by leader humility, 66.09% of the variance in follower satisfaction with leaders' administrative leadership could be accounted for by leader humility, all of the variance in follower satisfaction with leaders' interpersonal leadership could be accounted for by leader humility, and 34.59% of the variance in follower satisfaction with leaders' justice orientation could be accounted for by leader humility.
With ethnicity and all forms of humility as covariates in the model, only respect for others' viewpoints, a characteristic of being IH, was a significant predictor of followers' satisfaction with leaders' leadership. Specifically, for every unit increase in leaders' respect for others' viewpoints, there was a .45 unit increase in followers' satisfaction with leaders' interpersonal leadership and a .41 unit increase in followers' satisfaction with leaders' justice orientation.

Discussion
Although leader-reported humility was not predictive of followers' ratings of leaders' servant leadership, it was predictive of greater satisfaction among followers with their leaders' performance. As might be expected on the basis of the social bonding and social oil hypotheses of humility , leader humility was relevant to followers' ratings of the social aspect of leadership and how leaders treated their followers (e.g., getting to know .90 a a Spearman-Brown Coefficient followers, helping them, treating everyone fairly) rather than administrative aspects of leadership (e.g., being organized, enforcing rules and policies). Notably, out of all assessed aspects of expressed and intellectual humility, leaders having respect for diverse viewpoints stood out as the strongest predictor of followers' satisfaction with leaders' performance.
To build on these findings, future research could investigate whether a relational humility measure (Davis et al., 2011) or peer-ratings of humility predict followers' or supervisors' ratings of leader performance. Further, with a higher number of raters, round-robin methods could be used to tease apart unique relationship effects or to see if there is consensus among multiple peers about a leader's humility (Meagher et al., 2015). Leaders' influence may rise when there is a clear consensus among followers about a leader's humility and effectiveness.

General discussion
Drawing on longitudinal and multi-reporter research designs among novice community leaders, our studies showed moderate, positive associations between aspects of leader-reported humility and leader-reported servant-leadership, as well as follower-reported satisfaction with leaders. Although leader-reported humility was predictive of leaderreported servant leadership in Study 1, this did not replicate with followers' ratings of leaders' servant leadership in Study 2. The self-other knowledge asymmetry model helps explain how some characteristics are perceived better by oneself compared to others (Vazire, 2010). Perhaps the altruistic calling component of servant leadership is more difficult for followers to discern in leaders than for leaders to self-assess (e.g., whether the leader sacrifices his/ her own interests to meet the needs of followers). Future research would benefit from assessing more aspects of servant leadership to confirm whether the relationship between humility and servant leadership can transcend common method variance.
The aspects of humility most relevant to leadership outcomes were teachability (Study 1) and being able to appreciate diversity of thought (Studies 1 and 2). Thus, these two qualities may be particularly valuable for young community leaders working among peers. The Note. Coefficients are unstandardized and followed by standard errors in parentheses. All predictors/covariates were grand mean centered. ICC = intraclass correlation coefficient. * p < .05; ** p < .01; *** p < .001. fact that respect for others' viewpoints was robust in predicting leadership outcomes compared to more general appreciation for others' strengths highlights the relevance of humility in the domain of knowledge for beneficial leadership outcomes.
Given the observed connections between these aspects of leader humility, servant leadership, and followers' satisfaction with the leadership, it could be useful for future research to begin laying a foundation for cultivating humility in community leaders or leaders in other peer-based service roles. Studies could examine situations in which leaders are more or less successful being humble in order to develop exercises and activities aimed at nurturing humility. In a small study, students who completed a weeks-long workbook-style humility intervention (n = 26) reported increased humility-modesty, forgiveness, and patience (but not selfcontrol) compared to students in the control condition (n = 33; Lavelock et al., 2014). Although encouraging and an important start, more intervention studies are needed that address EH and IH rather than behavioral modesty, which is distinct from humility (Davis et al., 2010). Research validating interventions for developing IH is in its early stages. Short-term interventions aimed at boosting wisdom via perspective-taking appear to increase IH (see Kross & Grossmann, 2012) and college classes with an emphasis on IH appear to increase peerreported IH (Meagher et al., 2019). Intervention studies that focus on leadership contexts would be useful.
Given the Study 1 links between leader humility and servant leadership and the Study 2 links between leader humility and followers' satisfaction with leaders, future research could explore whether servant leadership mediates links between leader humility and followers' satisfaction with leaders. The literature provides a substantial basis from which to expect servant leadership may result in favorable ratings of leaders (e.g., Hoch et al., 2018;Sousa & Van Dierendonck, 2017;Van Dierendonck, 2011). Thus, servant leadership could be an explanatory mechanism by which leader humility results in greater follower satisfaction with leadership.
The current research raises additional questions about the role of leader EH and IH, such as whether leader humility may be associated with follower or organizational well-being, questions that may be examined in university contexts by exploring indicators such as campus morale or year-to-year retention rates. If humility strengthens relational bonds (see Van Tongeren et al., 2019), it could be connected to individual wellbeing through reduced conflict or stress. For example, in other contexts, humility appears to buffer negative effects of stressful life events on indicators of distress and lower well-being (Krause et al., 2016). In a university context, humility among peer-leaders may make them more approachable or patient, which could decrease resident stress and increase indicators of organizational culture and health (i.e., trust, perceived social support, satisfaction). Further, research has indicated that humility is particularly relevant to job performance when individuals are working with challenging clients . To the degree that some college students living in residential dorms could be challenging to work with, selecting or developing humble, nondefensive community leaders could help smooth interactions and resolve interpersonal conflicts.

Self-other ratings of humility
With regard to measurement, because humility in general, and IH in particular, involves inner thoughts, beliefs, and attitudes, self-reported humility can be very informative (cf. Vazire, 2010). The current research highlighted that self-reported humility was not only predictive of a self-reported outcome (Study 1), but also predictive of some other-reported outcomes (Study 2). The size of the relationship between self-reported respect for others' viewpoints, a characteristic of being intellectually humble, and other-reported leadership outcomes in the current research (B = .41 -.45) was greater in magnitude than some relationships observed in other multimethod humility studies where full scale humility scores were used. For example, among health care employees who work with challenging clients, employee self-reported honesty-humility predicted supervisor ratings of employees' job performance (B = .18;Johnson et al., 2011) and among participants who interacted in small groups, self-reported IH related to peer-reported openness (r = .33; Meagher et al., 2021).
In our examination of multiple forms of self-reported humility together in one model, many of them were not significant in predicting leadership outcomes. This may be due to some forms of humility being more robust than others in relation to leadership, however, it may also be that some forms of humility would be better assessed via the report of individuals or groups who know leaders well rather than through self-report. Expressed humility, specifically, was designed to be rated by informant reporters.

Trait, state, or domain-specific humility
Further, it is important to note that we assessed EH and IH as relatively enduring traits. However, humility could be experienced as a momentary state (Chancellor & Lyubomirsky, 2013;Kruse et al., 2017) or with regard to some domains of knowledge, but not others (Hoyle et al., 2016). For example, leaders could be humble about their religious or political views, but not express humility about a decision that resulted in an organizational failure. Viewing leadership as one's area of expertise could make it particularly challenging to be humble about one's leadership. Yet, humility specific to leadership knowledge, skills, and abilities could be relevant in promoting servant leadership and increasing leadership effectiveness.

Organizational settings and roles
Although we are cautiously optimistic about the value of humility in leaders, it is important not to overgeneralize our findings or draw conclusions about EH/IH in all leadership contexts. Our research was conducted on the campus of a private Christian university among students who were carefully selected and trained for peer-leadership positions. Further, given that the leaders identified as Christian, their religious values may have influenced their understanding of the concepts of humility and their desire to be humble and service-oriented leaders. Future research is needed to establish whether these findings would replicate among leaders at public universities, peerleaders in other positions (e.g., student government), or more seasoned leaders in public and private sectors.
In addition, although we frame humility as a positive quality in the context of community leadership in a university residential setting, there are leadership contexts or roles in which humility might not be as relevant or beneficial. A social perception seems to be that humility is more useful for religious leaders and less useful for jobs that require self-promotion (e.g., entertainer, sales; see Exline & Geyer, 2004). In fact, several measures of humility correlate negatively with the exhibitionistic tendency (Rowatt et al., 2006), therefore questions remain about the value of humility for leaders whose roles require them to be the center of attention, energize a crowd, or be especially charismatic. On the other hand, humility may be one key to success in leadership contexts that require teamwork, resolving interpersonal issues, and guiding followers.

Conclusions
We suspect the most effective leaders are able to discern how to express humility at the right time and place and in the right amount (e.g., see Whitcomb et al., 2015). Future research is needed to assess humility in more nuanced ways -such as a golden mean between the two extremes of servility and arrogance (Battaly & Nichols, 2016). Behavioral measures are also needed, such as assessing when leaders acknowledge limits or revise views after examining evidence (Haggard et al., 2018;Ng & Tay, 2020). Capturing humility in the moment will require creative methods and measurement techniques in real-life contexts or in situations where humility is strained. The current research indicates that when a leader exhibits certain forms of humility, this can aid in putting followers and their needs at the forefront and can contribute to followers' satisfaction with interpersonal aspects of the leader's performance. Future research may examine if there are additional beneficial downstream effects of leader humility.

Disclosure statement
We have no known conflict of interest to disclose.

Funding
This paper was supported by a generous grant from the John Templeton Foundation, Grant No. 60622, Developing Humility in Leaders.