Contrasting self-report and consensus ratings of intellectual humility and arrogance

Abstract

Despite a growing interest in intellectual humility (IH) and intellectual arrogance (IA), adequate measurement remains a challenging issue. This paper presents a pair of studies that compare two strategies: self-assessments and relational measures of group consensus. In Study 1, unacquainted participants provided round-robin judgments following a set of collaborative tasks. A social relations analysis revealed no consensus for either construct, making the relational measure untenable. However, a round-robin design following months of cooperative course work (Study 2) produced consensus for both constructs. Self-reported IH in both studies was positively associated with self-enhancement, despite the construct’s definitional association with accurate self-appraisals, whereas relational IH was not. These studies reveal key ways in which personal and relational assessments can differ.

Full text

Contrasting self-report and consensus ratings of intellectual humility
and arrogance
Benjamin R. Meagher
a,
⇑
, Joseph C. Leman
a
, Joshua P. Bias
b
, Shawn J. Latendresse
a
, Wade C. Rowatt
a,
⇑
a
Baylor University, United States
b
University of Mary Hardin-Baylor, United States
article info
Article history:
Received 12 January 2015
Revised 2 July 2015
Accepted 3 July 2015
Available online 6 July 2015
Keywords:
Intellectual humility
Intellectual arrogance
Person perception
Virtue
Positive psychology
abstract
Despite a growing interest in intellectual humility (IH) and intellectual arrogance (IA), adequate measure-
ment remains a challenging issue. This paper presents a pair of studies that compare two strategies:
self-assessments and relational measures of group consensus. In Study 1, unacquainted participants
provided round-robin judgments following a set of collaborative tasks. A social relations analysis
revealed no consensus for either construct, making the relational measure untenable. However, a
round-robin design following months of cooperative course work (Study 2) produced consensus for both
constructs. Self-reported IH in both studies was positively associated with self-enhancement, despite the
construct’s definitional association with accurate self-appraisals, whereas relational IH was not. These
studies reveal key ways in which personal and relational assessments can differ.
Ó2015 Elsevier Inc. All rights reserved.
‘‘Self-seeking, self-glory, that is not me. No. Many people say I
embarrass them with my humility.’’ Archbishop Peter Akinola, as
quoted in Polgreen and Goodstein (2006, December 25).
The study of human virtues has a long intellectual history, par-
ticularly in the fields of philosophy and theology. However, it has
only been within the past few decades that the positive psychology
movement led psychological researchers to begin to seriously con-
sider virtues and their role in human life (Peterson & Seligman,
2004). The result has been a wealth of recent scholarship on a vari-
ety of relevant topics, such as gratitude and forgiveness (Carlisle &
Tsang, 2013), love (Fehr, 2013), and self-control (Baumeister &
Vohs, 2012). Despite this broad, growing interest in positive
human attributes, humility, on the other hand, has been referred
to as the ‘‘most overlooked and underappreciated virtue’’
(Chancellor & Lyubomirsky, 2013, p. 819), as it has yet to produce
a comparably large body of work within empirical psychology.
This dearth has recently generated a great deal of consideration,
motivated in large part by extended reflection on the critical
importance of humility as a virtue specifically within the intellec-
tual and academic domain (Thrive Center for Human Development,
2014). For example, the advancement of scientific knowledge
seems to fully depend on practitioners possessing some degree of
intellectual humility. That is, researchers must be motivated to
pursue the truth, wherever that may lead them, instead of being
focused on status within the field, defending a ‘‘pet theory’’ regard-
less of its adequacy, or refusing to question one’s own initial
assumptions and positions in light of new, conflicting evidence
(Roberts & Wood, 2003). Even among non-scientists, learning from
others first requires an acknowledgment and admission of
ignorance (Hodges, Meagher, Norton, McBain, & Kimball, 2014),
so education itself is largely dependent upon these open expres-
sions of intellectual humility.
In light of the wide-reaching influence humility has on critical
aspects of human social functioning, empirical efforts to better
evaluate and understand this construct are well overdue. In this
paper, we begin with a brief description of the two primary
challenges responsible for curtailing empirical research on the
topic of humility generally: conceptual issues, in terms of defining
humility, and measurement issues, regarding how one can accu-
rately assess individual differences. Following this discussion, we
consider the relevance of these theoretical and methodological
issues for humility within the intellectual domain specifically.
1. Conceptual issues in the study of humility
The first stumbling block for an empirical approach to studying
humility has been a basic conceptual question: What is humility?
As is true for many terms in the psychological literature, concep-
tual definitions of humility often differ dramatically among lay
http://dx.doi.org/10.1016/j.jrp.2015.07.002
0092-6566/Ó2015 Elsevier Inc. All rights reserved.
⇑
Corresponding authors at: Department of Psychology and Neuroscience, Baylor
University, Waco, TX 76798, United States.
E-mail addresses: benjamin_meagher@baylor.edu (B.R. Meagher), wade_rowatt
@baylor.edu (W.C. Rowatt).
Journal of Research in Personality 58 (2015) 35–45
Contents lists available at ScienceDirect
Journal of Research in Personality
journal homepage: www.elsevier.com/locate/jrp

persons, theoreticians, and researchers. As noted by Tangney
(2009), dictionary definitions typically describe humility as merely
holding oneself in low regard, a trait entailing meekness,
self-abasement, and low self-esteem. However, despite this fairly
negative portrayal, lay theories of humility are actually quite pos-
itive, associating humility with good psychological adjustment and
positive emotions (Exline & Geyer, 2004). This finding is consistent
with the historical, English-language lexical research that led to
the development of the Five Factor Model of personality structure
(Costa & McCrae, 1985; McCrae & Costa, 1997), which places
humility-related items within the facet of modesty under the
higher-order factor of agreeableness. Thus, this framework views
humility as one component of having a prosocial and communal
orientation toward others. More recently, Lee and Ashton (2004),
developing their own model of personality structure, have argued
that lexical studies across multiple languages indicate that humil-
ity (with honesty) represents its own unique factor independent of
agreeableness, which is characterized by facets of sincerity, fair-
ness, greed-avoidance, and modesty. Notably, distinguishing
between agreeableness and Honesty–Humility is generally done
by researchers interested in different forms of social morality:
agreeableness predicting receptive forms (e.g., tolerance, forgive-
ness) and Honesty–Humility predicting more agentic forms (e.g.,
altruism, pro-sociality).
Outside of these lexical and lay theories of humility, a number
of philosophers, theologians, and psychologists have sought to
develop more nuanced definitions of the construct. Although
several different conceptions have been proposed (Chancellor &
Lyubomirsky, 2013; Davis, Worthington, & Hook, 2010; Exline
et al., 2004; Tangney, 2009), humility within these frameworks is
consistently characterized as a multidimensional construct, most
commonly including an accurate or moderate assessment of one’s
own abilities, being open to new ideas, having a low self-focus, and
being able to acknowledge one’s own mistakes. Notably, a large
portion of this theoretical work has involved distinguishing humil-
ity from a number of closely related constructs. For example,
although measures of modesty (e.g., self-reporting lower values
on desirable traits than do knowledgeable others) have often been
used as proxies for humility, Tangney (2009) argues that modesty
is a narrower construct, involving a moderate estimate of one’s
abilities but lacking the openness and low self-focus characteristic
of humility. Exline et al. (2004) make a slightly different distinc-
tion, suggesting that modesty is an exclusively social trait, entail-
ing a particular type of self-presentation that may or may not be
consistent with internal humility. In both cases, humility is con-
ceptualized as a more expansive construct than modesty.
Researchers have also argued that humility differs from certain
conceptually-related negative attributes, such as high self-esteem
or narcissism, which is characterized by feelings of grandiosity,
an overestimation of self-importance, and a sense of entitlement.
Tangney (2009) points out that although narcissistic people neces-
sarily lack humility, it is less evident that people low on narcissism
must also be high in terms of humility. For example, a person with
low self-esteem will likely be low in narcissism, but also
potentially low in humility, as they may be engaging in
self-deprecation as a means of eliciting a positive social response
and therefore still demonstrating a high self-focus.
The philosophical and theoretical literature on humility has
dramatically outpaced the empirical work attempting to scientifi-
cally evaluate these theories. As evidenced in this short review,
theoretical work has primarily sought to clarify many subtle defi-
nitional distinctions between humility and related constructs.
However, more empirical testing is needed to evaluate the ade-
quacy and robustness of these nuanced conceptual frameworks
when observing how individuals actually describe themselves
and others. As a result, the extent to which humility can be
empirically disentangled from numerous related traits, such as
agreeableness, modesty, narcissism, and arrogance, is still an open
question. The primary cause of this difficulty—concerns over mea-
surement—will be described next.
2. Measurement issues in the study of humility
The second chief challenge for empirical work on the topic of
humility has been the question of how to accurately measure the
construct (Davis et al., 2010). As with most psychological traits,
researchers have generally relied on self-report measures, such
as the Honesty–Humility subscale of the HEXACO Personality
Inventory (Lee & Ashton, 2004) or the Modesty–Humility subscale
of the Values in Action Strengths Inventory (Peterson & Seligman,
2004). However, as the opening quote of this paper hints, it is
unclear whether humility can be accurately self-reported. Would
a humble person be likely to brag to a newspaper that he embar-
rasses others with his humility? The large-scale distribution of
Rev. Akinola’s quote by his many detractors would suggest that
most people believe not.
This concern over self-report measurement stems from the very
characteristics of the attribute itself. Because humility is conceived
of as entailing an accurate or moderate view of oneself (i.e., not
self-enhancing), as well as a low self-focus, it is a construct that
is inherently linked to self-assessment. As a result, it is perhaps
not surprising that many question the internal validity of a
self-report measure. For example, people with low humility may
self-enhance and report high levels, and people with high humility
may express modesty and report lower levels (Davis et al., 2010).
This challenge has led a number of researchers to develop and
consider alternative measurement strategies (Chancellor &
Lyubomirsky, 2013), such as implicit assessments (Rowatt et al.,
2006). However, the most prominent alternative methodology
employed in the past several years is the use of personality
judgments from raters (Davis et al., 2010, 2011, 2013; Kruse,
Chancellor, Ruberton, & Lyubomirsky, 2014). Davis and colleagues
(Davis et al., 2010, 2011) have framed this approach as a measure-
ment of relational humility, defined as a social judgment, rather
than an intrinsic individual attribute. They argue that humility is
easier and more reliably assessed in others, as it avoids problems
related to self-enhancement and socially desirable responding.
Moreover, this perspective draws on Vazire’s (2010) self-other
knowledge asymmetry model, which proposes that ratings by
others tend to be more accurate than self-assessments on traits
that are highly evaluative, in that they are closely tied to motiva-
tional and ego-defensive processes. Humility, being a highly
valenced construct, falls into this evaluative classification.
Nevertheless, two important challenges exist for the quantita-
tive assessment of relational humility. First, Davis and colleagues’
(Davis et al., 2010, 2011, 2013) relational model proposes that rela-
tional humility is best measured in terms of inter-judge agreement
across a number of raters. However, these peer ratings will only be
meaningful if there actually is consensus among raters (Kenny,
1994). Groups are most likely to reach consensus for trait
judgments when perceivers witness the same or similar behaviors
that reflect a particular trait (Kenny, Albright, Malloy, & Kashy,
1994), viz., actions that provide good information (Funder, 1995).
A consensus assessment of humility may therefore be limited only
to a very specific set of contexts or relationships that are capable of
revealing this virtue. Several authors have argued that humility
will be most evident behaviorally in situations where it is directly
challenged, such as during interpersonal conflict, when receiving
recognition or praise, when interacting with someone of a lower
social status, or when describing past success (Chancellor &
Lyubomirsky, 2013; Davis et al., 2010, 2011; Kruse et al., 2014).
36 B.R. Meagher et al. / Journal of Research in Personality 58 (2015) 35–45

For example, Davis et al. (2013) had participants in groups take
part in a number of activities designed to reveal humility, such
as describing their respective strengths and weaknesses as a lea-
der. However, the effectiveness of this method in its ability to
reveal humility has not been robustly tested across different situ-
ations or time periods. If judgments of a particular individual’s
humility are primarily idiosyncratic to each perceiver, a measure
of inter-judge agreement will not be particularly meaningful.
Another potential issue for a relational model of humility is the
fact that it is much more time intensive than self-report measures.
It is therefore important to assess how this method relates to
self-reports of the same construct to determine if it actually pro-
vides different, let alone better, information. Research on related,
though typically negatively valenced, constructs have generally
revealed modest positive correlations between self- and
other-ratings. For example, several studies have tested self-other
agreement for impressions of narcissism, finding small but statisti-
cally significant, positive correlations of r< .20 (Cooper, Balsis, &
Oltmanns, 2012; Lukowitsky & Pincus, 2013). Carlson, Vazire, and
Oltmanns (2011) found no correlation between self- and
other-ratings of narcissistic traits between minimally acquainted
individuals, but found significant values when rated by a close
other (r= .15–.27). Similarly, in a workplace context, Johnson
et al. (2010) found modest correlations between employees’
self-ratings of arrogance and those provided by peers (r= .19)
and supervisors (r= .13). While clearly demonstrating a relation-
ship, these results also indicate substantial differences between
self- and other-assessments on self-evaluative constructs.
Despite the recent influx of alternative measures of humility,
fewer studies have actually provided direct comparisons between
self-ratings and other-ratings on humility specifically. Those that
have reveal fairly inconsistent results. For example, de Vries, Lee,
and Ashton (2008) found positive correlations between self and
other-ratings on Honesty–Humility using the HEXACO Personality
Inventory, but here too the magnitude of the correlation varied dra-
matically as a product of the dyad’s relationship (r= .22 for casual
acquaintances to r= .60 for romantic partners). Rowatt et al.
(2006) also found a positive correlation between self-reported
humility and ratings given by close others (e.g., friend, partner, or
family member), although the strength of this association varied
across different scale measures. In stark contrast, Davis et al.
(2013) found a statistically significant negative self-other correla-
tion (r=.31) among minimally acquainted participants, suggest-
ing the possibility of a ‘‘modesty effect,’’ with truly humble people
self-reporting lower humility. This variability in the relationship
between self and other-ratings poses a number of significant ques-
tions. At the very least these results indicate that, in certain social
contexts, self and other-ratings can be quite distinct and will likely
predict very different outcomes.
3. Intellectual humility
In the research reviewed thus far, humility has been conceptu-
alized as a general, global virtue. However, an additional
conceptual debate is the extent to which the construct is better
understood as context- or ability-specific, implying that there
may be multiple forms of humility relevant in different psycholog-
ical domains. That is, an individual may show high levels of humil-
ity in one facet of life (e.g., academic ability) but not necessarily in
another (e.g., social relationships). The subdomain of humility that
has garnered the most attention from psychologists (Gregg &
Mahadevan, 2014; McElroy et al., 2014; Samuelson et al., 2014)
and philosophers (Roberts & Wood, 2003) is intellectual humility
(IH), an epistemic virtue tied to the realm of knowledge and ideas.
Conceptually, IH has been proposed to entail an accurate or
moderate assessment of one’s own intelligence, being receptive
to the contributions and ideas of others, and being able to accept
criticism about one’s own ideas (e.g., McElroy et al., 2014;
Samuelson et al., 2014). High IH therefore involves having a low
self-focus and little concern for status or ownership over particular
ideas, instead caring more about the intrinsic value of knowledge
and truth (Roberts & Wood, 2003).
As discussed at the beginning of this paper, IH has been
proposed to represent a singularly important epistemic virtue, nec-
essary for science, education, and learning generally (Thrive Center
for Human Development, 2014). Whether one is observing a lab of
scientists or a college study group, the extent to which each mem-
ber is receptive to others’ ideas, capable of accepting criticism, and
grounded in their assessment of their own knowledge will presum-
ably have important implications for the group’s capacity to learn,
improve, and cooperate successfully. However, even within this
specifically intellectual domain, similar challenges regarding the
conceptual definition and measurement of humility persist. For
example, there is considerable debate in terms of how intellectual
humility relates to or is distinct from intellectual arrogance (IA).
One perspective argues that they merely represent the opposing
ends of the same construct (Gregg & Mahadevan, 2014): ‘‘[IA is]
the inclination to regard a belief as true because it is one’s own.
IH, conversely, would then be the inclination not to, or the disincli-
nation to do so’’ (p. 8). In contrast, Samuelson et al. (2014) have
argued that IH represents a ‘‘virtuous mean,’’ located between
the vice of IA (claiming to know more than is merited) and the
opposing vice of intellectual diffidence (claiming to know less than
is merited). In support of this conceptual distinction, Samuelson
et al. (2014) found that folk descriptions of prototypical IH and
IA persons, though primarily inversely related, nevertheless
demonstrated unique dimensions for epistemic motivations: ‘‘An
IA person uses education in a prideful way to confer social status,
while an IH person pursues education out of curiosity and love of
learning’’ (p. 14).
An additional conceptual question is the extent to which IH is
distinguishable from a more general, global measure of humility.
Recent empirical work provides some initial evidence that IH
represents a distinct subdomain capable of predicting unique vari-
ance above and beyond general humility on a number of related
constructs (Davis, Rice, McElroy, DeBlaere, & Choe, 2015). Of
course, the value of testing for unique statistical variance depends
wholly on accurate measurement, and several recent attempts
have been made to develop scale measures specifically for IH
(Krumrei-Mancuso & Rouse, in press; McElroy et al., 2014). For
example, McElroy et al. (2014) developed a 16-item scale,
consisting of items such as: ‘‘Is open to others’ ideas,’’ ‘‘Is good at
considering the limitations of their perspective,’’ and ‘‘Seeks out
alternative viewpoints.’’ However, similar questions regarding the
adequacy of self-ratings also apply to these assessments of IH, just
as they do to general humility. To what extent do individuals accu-
rately report their own IH? A relational measure of IH derived from
peer assessments (Davis et al., 2010, 2011) may provide a valuable
alternative method, but this approach has not yet been evaluated
within this intellectual domain. Importantly, the adequacy of this
measurement strategy will again depend on whether groups actu-
ally do reach consensus in their evaluations.
The goal of the present studies was to compare and contrast
self-ratings of IH with the ratings of peers at both minimal
acquaintance (Study 1) and following several months of interaction
within an academic context (Study 2). Specifically, we sought to
address three primary questions. First, under what circumstances
are peer ratings useful in providing a reliable measure of IH?
That is, under what conditions will significant levels of consensus
be reached for this construct? In Study 1, unacquainted individuals
took part in a group task designed to require creative thinking,
B.R. Meagher et al. / Journal of Research in Personality 58 (2015) 35–45 37

intellectual skills (i.e., math and verbal abilities), and consideration
of one’s own strengths and weaknesses—activities that should in
theory be most likely to demonstrate high or low levels of
intellectual humility. In Study 2, consensus for IH was assessed
among groups of students enrolled in a college course using a
team-based, classroom learning environment (Michaelsen,
Knight, & Fink, 2002). This intensive amount of joint activity specif-
ically within an academic context provides a test of whether an
extended period of time working in groups on intellectual exer-
cises is capable of providing sufficient information for groups to
reach consensus about members. Although the relationship
between self-ratings and informant ratings has been shown to dif-
fer by level of acquaintance (de Vries et al., 2008), it is unclear if the
amount of inter-judge agreement also changes over time.
The second goal of these studies was to evaluate the extent to
which ratings of IH within these contexts would be statistically
distinct from judgments of other constructs known to be related,
but proposed to be theoretically distinguishable. Specifically, we
collected both self- and other-ratings of items assessing both IH
and IA, in order to assess the extent to which they are distinguish-
able as traits, as proposed by Samuelson et al. (2014). In contrast,
finding large overlap between these two constructs would lend
greater support to theories placing these terms on either end of a
single spectrum (e.g., Gregg & Mahadevan, 2014; Roberts &
Wood, 2003). Also measured were a number of additional con-
structs and items associated with a variety of frameworks for
understanding general humility, such as agreeableness, interper-
sonal dominance (i.e., meekness), and self-esteem.
The final goal of the present studies was to compare self-ratings
and other-ratings of IH within these group contexts. This analysis
involved assessing the magnitude and direction of the self-other
correlation for each construct, as well as observing differences in
terms of their respective relationships with two particular out-
come variables. The first, self-enhancement, represents the extent
to which participants rate themselves highly on a variety of posi-
tive attributes, controlling for how positively they generally view
others and how positively other people view them (Kwan, John,
Kenny, Bond, & Robins, 2004). Although humility is a more expan-
sive construct than is modesty (Tangney, 2009), they nevertheless
partially overlap in terms of being related to an unexaggerated and
accurate view of oneself. Thus, a measure of intellectual humility
should not be associated with a general bias to self-enhance on
positive attributes. The second outcome, academic achievement,
was assessed in Study 2 with each participant’s individual course
grade given by the instructor, as well as their peer evaluation grade
provided by fellow group members. Previous work has revealed a
positive correlation between students’ course grades and an impli-
cit measure of humility (Rowatt et al., 2006), as well as between
judgments of team contribution and group ratings (Owens,
Johnson, & Mitchell, 2013). The current study expands on these
finding by providing a test of the association between academic
performance and intellectual humility specifically, in terms of both
explicit self-report and group consensus.
4. Study 1
4.1. Method
4.1.1. Participants
One hundred thirty-five undergraduate students (92 women)
participated in this experiment in exchange for course credit.
Ages ranged from 18 to 23 years old (M= 19.04), and the sample
was 61% White/Caucasian, 13% Asian/Pacific Islander, 12%
Black/African American, 11% Hispanic, and 3% other race or
ethnicity.
4.1.2. Measures and procedure
Unacquainted participants arrived to the study in groups rang-
ing in size from three to five people. Each participant was given a
name tag with a letter (A–E), and they were told that they would
be completing several activities together as a group. For the first
activity, participants were asked to reflect on their own strengths
and weaknesses and then share briefly (less than 30 s) about one
of each to the group. In the second task, the group completed a
brainstorming activity: the experimenter described a scenario in
which everyone in the world spontaneously developed two extra
thumbs that appear on the other side of their pinky fingers
(Bouchard & Hare, 1970). Participants were instructed to list as
many changes as possible that might occur as a result, and each
participant then shared their top two most creative changes with
the rest of the group. Finally, for the last activity, participants were
given five minutes to work together to complete three GRE prob-
lems as a group (two math questions and one verbal question).
After five minutes, the experimenter reviewed the correct answers
with the group, discussing what they had gotten right and wrong.
Together, these tasks took approximately 20 min for each group to
complete.
Immediately after the third activity, each participant then com-
pleted the experimental questionnaire, which required them to
provide personality judgments of every member of the group,
including themselves, on a set of 6-point Likert-type scales. The
two constructs of primary interest were: (1) intellectual humility,
which was measured with four items (
a
= .77; ‘‘Open to criticism
of ideas,’’ ‘‘Knows what he/she is not good at,’’ ‘‘Can learn from
others,’’ and ‘‘Is intellectually humble’’), and (2) intellectual
arrogance, which was measured with three items (
a
=.71;
‘‘Arrogant,’’ ‘‘Is closed-minded,’’ and ‘‘Believes own ideas superior
to other’s ideas’’). Consistent with previous theorizing (Gregg &
Mahadevan, 2014; Samuelson et al., 2014), we treated IH and IA
as distinct constructs, rather than a single bipolar attribute.
Several other additional constructs hypothesized to be relevant
to group functioning were also collected. These included interper-
sonal dominance (
a
= .75; ‘‘Assertive’’ and ‘‘Tends to dominate dis-
cussion’’) and competence (
a
= .73; ‘‘Intelligent,’’ ‘‘Good at public
speaking,’’ ‘‘Has strong math skills,’’ and ‘‘Has strong verbal skills’’),
as well as the Big-5 personality traits, which were adapted from
the Ten Item Personality Inventory (Gosling, Rentfrow, & Swann,
2003): extraversion (
a
= .71; ‘‘Extroverted, enthusiastic’’ and
‘‘Reserved, quiet’’), agreeableness (
a
= .56; ‘‘Critical, quarrelsome,’’
‘‘Sympathetic, warm,’’ and ‘‘Likeable’’), conscientiousness (
a
= .30;
‘‘Dependable, self-disciplined’’ and ‘‘Disorganized, careless’’), neu-
roticism (
a
= .59; ‘‘Anxious, easily upset,’’ ‘‘Calm, emotionally
stable,’’ and ‘‘Depressed’’), and openness (
a
= .60; ‘‘Open to new
experiences, complex,’’ ‘‘Conventional, uncreative,’’ and
‘‘Creative’’). Finally, several additional, exploratory measures were
measured with single-items: ‘‘Likes to be center of attention,’’
‘‘Honest,’’ ‘‘Funny,’’ ‘‘Has high self-esteem,’’ and ‘‘Is a good leader.’’
4.1.3. Data preparation and analysis
The analysis was limited to participants who were
unacquainted prior to the study session. As a result, groups of three
people that contained a pair of acquainted individuals were
dropped entirely from the analysis. This was true of four groups.
For groups that contained four or more people, one of the
acquainted participants was randomly selected to be removed
from the dataset. This occurred in three groups. In total, this left
a sample size of 120 participants divided across 32 groups.
Data was analyzed using the Social Relations Model (SRM;
Kenny, 1994), a statistical method that allows for the partitioning
of interpersonal judgments into three primary components: (1)
the perceiver effect, or how a perceiver generally rates everyone,
(2) the target effect, or how a target is generally rated by everyone,
38 B.R. Meagher et al. / Journal of Research in Personality 58 (2015) 35–45

and (3) the relationship effect, or the idiosyncratic way in which a
particular perceiver uniquely sees a particular target. By partition-
ing judgments into these three sources of variance, one can assess
of the relative influence of each these factors. For example, a large
amount of variance attributable to the perceiver effect reflects the
tendency for participants to generally rate all other members of
their group in a similar way (i.e., assimilation). Alternatively, a
large amount of variance attributable to the target effect would
reflect consensus within the group: there is agreement about
which people are seen as intellectually humble and which people
are not. In this study, variance partitioning was done using the
software program SOREMO (Kenny, 1998). Because SOREMO is
unable to control for missing data, the mode response was imputed
into all empty fields prior to conducting the variance partitioning
procedure. Imputed data accounted for just 0.3% of all responses.
4.2. Results
4.2.1. Relative variance partitioning
The relative variance partitioning for all measured construct
ratings are shown in Table 1. Significance tests for these values
involve computing variances for each group in the sample, and
then employing a one-way t-test to evaluate whether the means
of the variances differ from zero. Groups did not achieve statisti-
cally significant levels of consensus for IH. In fact, the largest
source of variance came from the perceiver, which was greater
than the target variance, t(31) = 4.06, p< .001, and marginally
greater than the relationship variance, t(31) = 1.73, p=.093.
Similarly, no target variance was found for IA. Instead, significantly
greater variance was attributable to the perceiver, t(31) = 2.87,
p= .007, and to the relationship, t(31) = 2.08, p= .045. The magni-
tude of IA perceiver variance did not differ from relationship
variance, t(31) = 0.589, p= .560. Constructs that did achieve statis-
tically significant levels of consensus included extraversion,
t(31) = 4.33, p< .001, interpersonal dominance, t(31) = 2.76,
p= .009, leadership, t(31) = 3.68, p= .001, and being funny,
t(31) = 3.70, p= .001.
4.2.2. Correlates of self-ratings
We examined associations between self-reported IH, IA, and
other measured traits. These correlations with the self-ratings on
the other measured constructs are shown in Table 1, under
self-self correlations (disattenuated for reliability). Notably, while
ratings of IH and IA were correlated, the magnitude of the correla-
tion indicates that participants did differentiate between the
constructs. Those rating themselves as intellectually humble also
tended to rate themselves positively on most other attributes,
including, most strongly, being agreeable, funny, honest, and
emotionally stable. Interestingly, self-reporting as high on IA was
negatively correlated with these items, and it was particularly
associated lower levels of emotional stability and lack of conscien-
tiousness. Both IH and IA were positively associated with interper-
sonal dominance. Because no consensus was reached in groups
regarding IH or IA, equivalent correlations with other-ratings were
not conducted.
4.3. Discussion
In this study, unacquainted groups of individuals engaged in a
series of interpersonal tasks prior to providing their impressions
of their fellow group members. These interpersonal judgments
were then partitioned into distinct sources of variance in order to
assess relative amounts of consensus, assimilation, or relational
idiosyncrasy. Our results indicated that the social tasks used in this
procedure were unable to reveal IH or IA, as essentially no consen-
sus was reached by groups for these constructs. A very similar
procedure was employed by Davis et al. (2013) to measure general
humility relationally, but they did not report the relative magni-
tudes of perceiver and target variance found in their study. One
substantial difference between the two studies was the length of
time participants spent together: whereas the tasks of the current
study took no more than half an hour to complete, participants in
Davis et al. (2013) spent approximately three hours together.
Kenny (2004) has noted that consensus can increase over this per-
iod of time following initial exposure, suggesting the need for a
more concerted effort to evaluate what particular situations and
tasks reliably elicit consensus on these judgments, as well as deter-
mining the dynamics of how consensus may change over time.
Because there was no consensus reached by groups regarding
IH and IA, other-ratings could not be directly compared to
self-ratings, as was initially planned. However, observing the
correlations between the various self-ratings still raised several
questions regarding validity. Conceptually, although IH and IA
showed an expected negative correlation, the magnitude of this
relationship suggests that participants did appear to distinguish
between them when evaluating themselves. However,
Table 1
Relative variance partitioning and self-rating correlations in Study 1.
Relative variance partitioning Self–Self correlations
a
Perceiver Target Relationship Error IH IA
Intellectual humility .316
**
.025 .154
**
.505 – .352
**
Intellectual arrogance .177
*
.000 .271
*
.552 .352
**
–
Interpersonal dominance .152
*
.190
*
.223
**
.435 .241
*
.339
**
Competence .180
**
.086 .133
**
.601 .337
**
.301
**
Extraversion .000 .271
**
.306
**
.423 .260
*
.040
Agreeableness .127
*
.033 .160
**
.679 .694
**
.443
**
Conscientiousness .000 .049 .176
**
.775 .101 .602
**
Emotional stability .234
**
.000 .078
*
.688 .449
**
.644
**
Openness .053 .089
*
.186
**
.672 .065 .001
Leader .279
**
.203
**
– .518 .295
**
.055
Center of attention .303
*
.178 – .520 .021 .361
**
Honest .541
**
.000 – .459 .474
**
.368
**
Funny .249
**
.227
**
– .524 .501
**
.186
*
High self-esteem .300
**
.066 – .635 .325
**
.086
Note: Variance attributable to relationship is indistinguishable from error on single-item measures.
*
p< .05.
**
p< .01.
a
Correlations are disattenuated based on scale reliability. Significance tests for self-self correlations are based on df = 87.
B.R. Meagher et al. / Journal of Research in Personality 58 (2015) 35–45 39

self-reported IH showed strong positive associations with a
number of qualities one can regard as high in social value, such
as competence, honest, and being funny. Its correlation with agree-
ableness was particularly substantial and consistent with a Five
Factor Model framework (Costa & McCrae, 1985; McCrae & Costa,
1997), which understands humility as subsumed within agreeable-
ness. In contrast, self-reported IA showed a relationship with
generally low self-appraisals, with particularly substantial nega-
tive correlations with conscientiousness and emotional stability.
Moreover, both self-reported intellectual humility and arrogance
were found to be positively correlated with interpersonal domi-
nance, a trait one would not expect to be related to being
other-oriented (Davis et al., 2011; Tangney, 2009).
The direction of these associations seem to reflect the concerns
raised by researchers regarding the validity of self-reporting in
terms of humility: People with low humility may self-enhance
broadly and report high levels of humility, whereas those with high
humility may express modesty and report lower levels of humility
(Davis et al., 2010). In this study, self-ratings appeared to reflect
either a general positive evaluation, which included viewing one-
self as high in IH, or a general negative evaluation, which including
describing oneself as high in IA. The association between high
self-reported IA and negative evaluations of oneself is particularly
surprising in light of previous work on related, global constructs,
which have found expected associations between self-report
assessments on narcissism and excessively positive views of one-
self (e.g., Carlson et al., 2011). However, it should be noted that
the brief items used here to assess IA differ substantially from
the items typically used to measure narcissism (e.g., NPI-16;
Ames, Rose, & Anderson, 2006), in focusing on willingness to
acknowledge others’ ideas, rather than on more general feelings
of importance and grandiosity.
Thus, the results of this initial study highlight a pair of chal-
lenges for measuring IH: (1) other-ratings at minimal acquaintance
do not appear to be reliable, even following tasks designed to
reveal these traits, and (2) self-ratings of IH were correlated with
a number of unrelated, socially valued attributes, despite the fact
that humility is theoretically defined in part as having less motiva-
tion to self-enhance (Davis et al., 2011). It is possible that the
positive self-ratings associated with IH in this study were not
due to a bias, but instead reflective of either accuracy (e.g., people
high in IH genuinely are more agreeable, honest, and competent) or
a tendency to see everyone more positively, including themselves.
In Study 2, we evaluate self-enhancement by controlling for these
alternative possibilities.
5. Study 2
The procedure employed in Study 1 failed to produce a
significant amount of consensus within the tested groups of unac-
quainted individuals. Although previous research has found fairly
stable levels of consensus over time for certain traits, at first
acquaintance such agreement depends largely on having shared
stereotypes regarding the given attribute (Kenny, 2004). It is possi-
ble that people simply do not share initial assumptions about high
levels of IH. As a result, a longer period of interaction may be nec-
essary to reveal this trait. In Study 2, we test this possibility by
assessing personality judgments within groups of students who
have worked together over several months. It was hypothesized
that in this context – where participants have multiple shared
experiences related to learning and social cooperation – groups
would be more likely to form reliable, shared opinions about
who is high or low in IH.
Secondly, because these assessments were collected within a
classroom context among well-acquainted individuals, the
relationship between IH and a pair of important outcomes could
also be assessed. First, we sought to evaluate the relationship
between self-ratings and other-ratings of IH with a tendency to
self-enhance on other positive attributes, employing the procedure
developed by Kwan et al. (2004). This approach assesses the corre-
lation between IH and a given construct, after controlling for the
ratings of other, well-acquainted individuals, as well as the per-
ceiver’s tendency to rate everyone a particular way. If a positive
association persists even after including these controls, it would
reflect a positive self-bias. Secondly, the relationship between
academic performance and IH could also be measured, both for
self-ratings and other-ratings. Previous work has found a positive
relationship between general humility and academic success
(Rowatt et al., 2006), as well as positive team evaluation (Owens
et al., 2013). In fact, Owens et al. (2013) found that perceived
humility can compensate for low general intelligence when being
evaluated by others, and groups whose leaders express humility
show greater engagement and satisfaction. The present study
offers a further test of this relationship by comparing self-report
and others’ ratings of intellectual humility and arrogance, both
for individual academic performance and peer evaluations of
performance.
5.1. Method
5.1.1. Participants
All participants were undergraduate students enrolled in one of
four upper level psychology courses (Health Psychology, Industri
al–Organizational Psychology, Lifespan Development, or Theories
of Personality) that utilized a team-based learning approach
(TBL; Michaelsen et al., 2002). The overarching goal of this peda-
gogy is to encourage critical thinking, engagement, and the devel-
opment of interpersonal and communication skills. TBL involves
assigning students to groups at the beginning of the semester, with
whom they complete a variety of tasks, both individual and collec-
tively, over the duration of an academic semester. Initially,
out-of-class reading assignments (e.g., textbook chapters, articles)
are assessed using ‘‘Readiness Assurance Tests,’’ which are first
completed individually and then retaken collaboratively with fel-
low group members, who provide feedback on each member’s
work. The majority of time in-class is devoted to application
exercises that require each team to form a joint decision about a
specific problem based on their content knowledge, which they
then report to and discuss with the rest of the class. Students earn
credit for their individual performance and for the performance of
their group, which are weighted by peer evaluations of their
contributions to group productivity.
In the current study, participants were assigned to groups of
four to six people. At the end of the semester, the experimenter
explained the research to the students, asked for volunteer partic-
ipation, and acquired informed consent. Of the 108 students (83
women) enrolled, 103 agreed to participate in the study divided
across 23 team groups. Ages ranged from 18 to 30 years old
(M= 21.08), and the sample was 49% White/Caucasian, 24%
Hispanic, 13% Black/African American, 9% Asian/Pacific Islander,
and 5% other race or ethnicity.
5.1.2. Measures and procedure
Each participant completed the same questionnaire employed
in Study 1, which asked them to evaluate the other members of
the group on 6-point Likert-type scales. Again, the constructs
measured were: (1) intellectual humility,
a
= .77, (2) intellectual
arrogance,
a
= .74, (3) interpersonal dominance,
a
= .75, (4) compe-
tence,
a
= .71, and (5) Big-5 personality traits: extraversion,
a
= .75, agreeableness,
a
= .61, conscientiousness,
a
= .64, neuroticism,
a
= .66, and openness,
a
= .65. Finally, several additional constructs
40 B.R. Meagher et al. / Journal of Research in Personality 58 (2015) 35–45

were measured with a single-item: ‘‘Likes to be center of atten-
tion,’’ ‘‘Honest,’’ ‘‘Funny,’’ ‘‘Has high self-esteem,’’ and ‘‘Is a good
leader.’’
At the completion of the semester, students’ final grades con-
sisted of two scores: the total number of points they had earned
for individual performance (e.g., tests, essays) and group perfor-
mance (e.g., exercises). To promote personal accountability and
decrease social loafing, peer evaluations from those in their
group determines what percentage of the full group performance
points students ultimately receive. This latter outcome involved
asking students to distribute 100 points among the other mem-
bers of the group, based on their contribution during group work
over the semester. The instructor then scaled each student’s
average peer rating relative to a hypothetical equal distribution
of the points (e.g., if 4 people were rated in the group, each
person’s average would be divided by 25), so that a 1 would
mean they contributed exactly their fair share, a larger number
means the contributed more than their fair share, and a number
less than one means they contributed less than their fair share.
5.1.3. Data preparation and analysis
Variance partitioning was again done using SOREMO (Kenny,
1998). Empty fields, accounting for just 0.07% of all responses,
were imputed with the mode response for that item.
5.2. Results
5.2.1. Relative variance partitioning
The relative variance partitioning for all measured construct
ratings are shown in Table 2. Unlike in the previous study, groups
showed statistically significant levels of consensus for both IH,
t(22) = 3.29, p= .003, and for IA, t(22) = 3.38, p= .003. For IH, target
variance was statistically smaller than perceiver variance,
t(22) = 2.23, p= .036, but not smaller than the relationship vari-
ance, t(22) = 1.06, p= .302. For IA, target variance did not signifi-
cantly differ in magnitude from perceiver variance, t(22) = 0.62,
p= .541, but was marginally greater than the relationship variance,
t(22) = 1.97, p= .061. Comparing these results to Study 1, consen-
sus for IA was significantly higher in Study 2 with greater acquain-
tance, t(53) = 3.69, p< .001, but consensus for judgments of IH
were not statistically significantly greater, t(53) = 1.47, p= .148.
5.2.2. Correlates of self and other-ratings
Correlations between self-reported IH and IA and the other
measured constructs are shown in Table 3, under the ‘‘Self–Self’’
column. As in Study 1, ratings of IH and IA were strongly correlated
but not indistinguishable, with a disattenuated correlation of
.471. Also consistent with Study 1, those rating themselves as
intellectually humble also tended to rate themselves positively
on most traits, including competence, agreeableness, emotional
stability, leadership, honesty, and being funny. This study also
found a strong positive correlation between IH and openness,
which was not observed in Study 1. High self-ratings on IA also
predicted higher levels of reported competence, as well as higher
dominance, higher desire to be the center of attention, and lower
levels of emotional stability.
Because consensus was found in this study for both IH and IA,
correlations could also be calculated for the target effects of IH
and IA with the other measured judgments. These values indicate
whether those rated highly by the group on one trait are also likely
to be rated highly by the group on another trait. These values are
shown in the ‘‘Other–Other’’ column of Table 3. The magnitude
of the disattenuated correlations between IH and IA, as well as
for Agreeableness, indicate that there was essentially no distinc-
tion between these constructs in terms of their target variance.
That is, the consensus in other-ratings for IH was largely equivalent
to their judgments of IA and agreeableness.
Finally, correlations between other-ratings for IH and IA with
self-ratings on the other constructs are shown in the ‘‘Other–Sel
f’’ column. These values indicate which self-ratings were associ-
ated with being viewed as high in IH or IA by the group. Self–other
agreement for ratings of IH was positive but non-significant,
disattenuated r= .214, p= .162, whereas self–other agreement for
ratings of IA was statistically significant, disattenuated r= .394,
p= .005. This indicates a modest level of agreement between an
individual’s appraisal of him or herself and the impressions held
by others in the group. Groups tended to view people as intellectu-
ally humble who reported being low in dominance and high in
agreeableness. Groups tended to view people as intellectually
arrogant who reported being high in dominance, extraversion,
and wanting to be the center of attention, but low in agreeableness
and conscientiousness.
5.2.3. Relationship with self-enhancement
Although people who rated themselves as intellectually humble
also tended to rate themselves more positively in general, it is
possible that this relationship is the product of either: (a) their
tendency to rate everyone positively (i.e., their perceiver effect),
or (b) everyone viewing them positively (i.e., their target effect).
In order to disentangle self-ratings from other-ratings,
self-enhancement indices were calculated based on the procedure
developed by Kwan et al. (2004). This index proposes that
self-ratings are the additive result of how one generally sees
others, how they are generally seen by others, and their unique
view of themselves. Thus, unique self-ratings (i.e., their self-
enhancement) can be computed by partialing out group-
mean-deviated perceiver and target effects. The columns of
Table 4 provide the partial correlations for these analyses, measur-
ing the relationship between the self-rating for each construct with
IH and IA self-ratings, controlling for the other SRM components of
the outcome variable.
1
Reporting oneself as high in IH was generally
associated with a more positive view, showing self-enhancement in
terms of competence, agreeableness, openness, being a leader, and
having high self-esteem. IA self-reports were associated with
idiosyncratically higher ratings of oneself in competence.
Table 2
Relative variance partitioning in Study 2.
Perceiver Target Relationship Error
Intellectual humility .238
**
.105
**
.158
**
.499
Intellectual arrogance .206
**
.261
**
.111
**
.422
Interpersonal dominance .040 .446
**
.169
**
.345
Competence .135
**
.125
**
.162
**
.578
Extraversion .000 .438
**
.226
**
.336
Agreeableness .068 .181
**
.157
**
.594
Conscientiousness .159
**
.254
**
.124
*
.463
Emotional stability .250
**
.016 .140
**
.594
Openness .141
*
.072 .183
**
.604
Leader .149
**
.387
**
– .464
Center of attention .057 .268 – .675
Honest .382
**
.099 – .519
Funny .281
**
.212
**
– .508
High self-esteem .245
*
.175
**
– .581
Note: Variance attributable to relationship is indistinguishable from error on single-
item measures.
*
p< .05.
**
p< .01.
1
Self-enhancement can also be measured by creating a discrepancy score by
subtracting the perceiver effect and target effect from self-ratings (e.g., Davis et al.,
2013). However, discrepancy scores can potentially confound the effects of
self-ratings and perceiver/target effects (e.g., humility can correlate with a discrep-
ancy score only because it is strongly related to the self-rating, regardless of its
relationship with the perceiver or target effects).
B.R. Meagher et al. / Journal of Research in Personality 58 (2015) 35–45 41

Moreover, the relationship between one’s target effect on these
constructs and the tendency to self-enhance on the various
self-report items was also calculated. Because the target effects for
IH and IA were indistinguishable, these scales were combined for
this analysis to form a single IH value with IA items reverse coded.
To assess self-enhancement, self-ratings for each construct were
correlated with this IH target effect, controlling for the other SRM
components of the variable. Unlike self-reported humility, group
consensus was not found to be related to self-enhancement on any
construct. Interestingly, only for self-esteem was there trend toward
self-enhancement among those identified by the group as high in
intellectual humility, p= .098.
5.2.4. Relationship with academic performance and peer ratings
To evaluate how IH and IA related to academic performance,
self and other-ratings of the two key constructs were correlated
with both individual grades (i.e., evaluation on assignments
completed individually) and peer evaluation grades (i.e., evalua-
tion given by group members for group assignments).
Self-ratings of IH showed a non-significant, negative relationship
with performance, with a disattenuated r=.223, p= .161, for
individual grade and disattenuated r=.148, p= .265, for peer
evaluation. Self-ratings of IA, on the other hand, had a statistically
significant, positive association with individual grades, disattenu-
ated r= .369, p= .029, and a marginally significant association with
peer evaluations, disattenuated r= .201, p= .066. In contrast,
other-ratings of IH showed a marginally significant, positive asso-
ciation with peer evaluations, disattenuated r= .380, p= .055.
However, there was no relationship between other-ratings of IH
and individual grades, disattenuated r= .017, p= .535.
5.3. Discussion
This study expanded on Study 1 by evaluating judgments of IH
among students who had been interacting with one another over
several months within an academic setting. Unlike in the initial
study, this social context proved sufficient to elicit consensus
within groups regarding who among them were intellectually
humble and arrogant. How did they accomplish this? Successful
judgments of others’ personality traits depend on having access
to observable behaviors (Funder, 1995). It is for this reason that
traits like extraversion, which are manifested in overt behavior
(e.g., speaking frequently), tend to elicit the largest amounts of
agreement. In this study, it appears that groups also used this
interpersonal behavioral information to inform their humility
judgments. Group judgments of IH were strongly correlated with
self-reported dominance (negatively) and agreeableness (posi-
tively), and impressions of IA were related to greater extraversion,
dominance, and desire for attention. Thus, in the present context of
cooperative group coursework, IH was inferred largely from evi-
dence of positive interaction and deference to others, whereas
those rated highly in terms of IA were those who spoke frequently
and dominated the direction of the group. This result provides
some initial evidence for what types of behavior are used by obser-
vers to form impressions of this intellectual virtue.
Critically, relational judgments by the group showed a modest
correlation with self-ratings on the two critical constructs.
Although self-reports did have a positive directional relationship
with group consensus for both IH and IA, only for IA was this
Table 3
Disattenuated correlations of IH and IA for self- and other-ratings in Study 2.
Intellectual humility Intellectual arrogance
Self–Self Other–Other Other–Self Self–Self Other–Other Other–Self
Intellectual humility – – .214 .471
**
.979
**
.102
Intellectual arrogance .471
**
.979
**
.388
*
– – .394
**
Interpersonal dominance .202 .735
**
.446
*
.470
**
.865
**
.467
**
Competence .599
**
.458 .030 .410
*
.600
**
.173
Extraversion .104 .442
**
.344 .010 .589
**
.378
*
Agreeableness .422
**
.977
**
.524
**
.302 .896
**
.435
**
Conscientiousness .078 .412
**
.282 .210 .314
*
.323
*
Emotional stability .332
*
– .328 .332
*
–.190
Openness .611
**
– .043 .093 – .012
Leader .446
**
.218 .115 .130 .359
*
.067
Center of attention .086 – .277 .450
**
– .285
*
Honest .253
*
–.055 .232 – .019
Funny .310
*
.074 .032 .178 .077 .113
High self-esteem .264
*
.676
**
.051 .218 .738
**
.028
Note: Self-rating correlations are disattenuated based on scale reliability. Other-ratings are disattenuated based on the reliability of the target variance estimate. Significance
tests are based on the uncorrected correlations, df = 79.
*
p< .05.
**
p< .01.
Table 4
Partial correlations of IH and IA on self-enhancement scores in Study 2.
IH self-rating IA self-rating IH (w/IA items)
other-rating
Intellectual humility – .122 –
Intellectual arrogance .105 – –
Interpersonal dominance
b
.212 .212 .026
Competence .422
**
.245
*
.019
Extraversion
b
.156 .079 .008
Agreeableness
b
.290
**
.162 .116
Conscientiousness .134 .028 .003
Emotional stability
a
.012 .020 .090
Openness
a
.311
**
.127 .015
Leader .382
**
.092 .067
Center of attention – –
Honest
a
.084 .135 .021
Funny .220
*
.190 .069
High self-esteem .250
*
.117 .187
Note: Self-enhancement values represent partial correlations between self-ratings
of constructs with arrogance/intellectual humility after controlling for perceiver
and target effects. Significance tests are based on df = 77.
a
Enhancement measure partials out perceiver effect only.
b
Enhancement measure partials out target effect only.
*
p< .05.
**
p< .01.
42 B.R. Meagher et al. / Journal of Research in Personality 58 (2015) 35–45

association statistically significant, indicating a clear distinction
between the two methods. Notably, self-ratings for the tested
items used in this study appeared to show greater conceptual
nuance than peer judgments. That is, when evaluating themselves,
correlations between judgments of IH, IA, and agreeableness,
though substantial, were not indicative of conceptual overlap. In
contrast, the correlations between the target effects of these
constructs approached 1.0, particularly after correcting for attenu-
ation. This indicates that at the group level, perceivers did not
differentiate between these variables when assessing members.
According to Vazire’s (2010) Self–Other Congruency Model,
self-ratings more accurately assess traits that are low in observ-
ability, whereas other-ratings are more accurate for judgments of
traits that are high in evaluativeness. However, it is also possible
that self- and other- trait judgments may differ not just in terms
of their accuracy, but also in terms of their complexity and clarity.
With less available information than people evaluating themselves,
other-ratings may be more likely to exaggerate the overlap
between constructs. As a result, self-ratings may be able to reflect
the theoretical distinction between IH and IA argued by Samuelson
et al. (2014), but observations by others may not. Alternatively, one
could argue that IH and IA are in fact overlapping constructs (e.g.,
Gregg & Mahadevan, 2014), and the results of the other-ratings in
this study merely demonstrate this fact.
The distinction between self-ratings and other-ratings on IH
and IA was further highlighted by their respective relationships
with the other measured constructs. Self-ratings of IH were
positively associated with the majority of positively valenced
attributes, including being competent, agreeable, open, a leader,
and funny. Although a portion of this relationship may be attribu-
table to the fact that these individuals tended to rate everyone
more positively, these positive correlations remained even after
controlling for participants’ respective perceiver and target effects
(Kwan et al., 2004). Thus, rating oneself as higher in IH actually
predicted a tendency to self-enhance. Although it is theoretically
possible that this could reflect accuracy (e.g., intellectually humble
people really are more competent), the range of items showing a
self-enhancement bias makes this proposal seem unlikely. For
example, there is little reason to believe that intellectual humility
predicts being truly funnier than your group actually appreciates.
Moreover, both self-reported intellectual humility and arrogance
were positively related to self-enhancing in terms of interpersonal
dominance, a trait one would not expect to be related to being
other-oriented (Davis et al., 2011; Tangney, 2009).
Relational ratings of IH, on the other hand, did not show a
statistically significant correlation with self-enhancement on any
construct. In other words, people identified by their group as high
in IH tended to be those with moderate ratings of themselves.
Moreover, correlations with academic performance differed dra-
matically between measures. Consistent with previous work
employing implicit (Rowatt et al., 2006) and peer evaluations
(Owens et al., 2013) of general humility, group consensus on IH
was related to receiving a better grade in the class, but only for
assignments evaluated by peers. In contrast, self-ratings of IH did
not correlate with better grades, instead trending in a negative
direction, whereas self-ratings of IA were related to superior
grades. This result is quite surprising, as IA is theoretically defined
as having an exaggerated view of one’s intellectual ability and
knowledge. Nevertheless, the observed relationship found here
again appears to highlight a challenge inherent in assessing intel-
lectual virtues and vices with self-assessments. People who did
well in the course were those who generally rated themselves
highly as thinking their ideas are superior to others’. In fact, their
grades suggest that they were correct in this assessment. If one
is genuinely more competent and has better ideas than others in
the group, they may be justified in being close-minded, if their
ultimate goal is to secure a better grade for themselves. If, on the
other hand, one has more prosocial motivations, such as helping
others in the group get a better grade, sensitivity to a more
complex array of values may be necessary for successful group
functioning (Hodges et al., 2014). Thus, assessing group outcomes
may provide a way to determine in what ways IA can hinder
performance, rather than predict success.
6. General discussion
In this pair of studies, we sought to evaluate two predominant
ways of assessing intellectual humility and intellectual arrogance
within a group context: self-assessments and group consensus rat-
ings. Both methods proved to have limitations. For self-ratings of
IH, the primary issue remains validity: in both studies, high
self-reported IH had a strong, positive association with
self-enhancement on a number of socially valued attributes,
despite the fact that definitions of humility entail having accurate
or moderate views of one’s knowledge and ability (Davis et al.,
2010; McElroy et al., 2014; Tangney, 2009). Although it remains
possible that people high in IH are also truly competent, better
leaders, and funnier, the uniquely high self-appraisals across such
a range of attributes makes this hypothesis seem rather unlikely.
The alternative explanation, that high IH self-reports are driven
in large part by socially desirable responding, is consistent with
more general models of personality ratings that challenge the
accuracy of self-reports on highly evaluative traits (Vazire, 2010).
Having said this, one limitation of the current studies is a reli-
ance on a brief measure of IH that has not been empirically vali-
dated in prior work. Because the empirical study of intellectual
humility is a new field of inquiry, there is yet to be an established
self-report scale in use from which we could have drawn. McElroy
et al. (2014) has only recently developed a 16-item informant scale
of IH, which could conceivably be adapted to a self-report measure.
A measure such as this, which also has multiple sub-scales, may
provide a better test of how well a self-report measure of IH can
predict expected outcomes. Having said this, there was in fact sub-
stantial overlap between the brief items used in the present studies
and the larger scale developed by McElroy et al. (2014). The latter
contains items that closely parallel those used here: e.g., ‘‘Is good at
considering the limitations of their perspective,’’ ‘‘Is open to
competing ideas,’’ and ‘‘Is open to others’ ideas.’’ Moreover, it is
important to note that employing a longer, multi-item scale within
a round-robin context can prove challenging for participants in
larger groups, as this exponentially increases the time required
to complete the study, thus making its application potentially
limited. Nevertheless, future scale testing of measures such as
these would benefit from making direct comparisons between
how people evaluate themselves and how they are evaluated by
others on these items.
In light of the concerns for self-report assessments, relational
measures of humility have been proposed to provide a viable alter-
native method for assessing the construct (Davis et al., 2010, 2011,
2013). The current studies were an initial step toward testing this
claim, specifically in the domain of IH. Supporting its adequacy,
relational IH in Study 2 showed no correlation with
self-enhancement and a positive relationship with peer evalua-
tions of academic performance, consistent with previous work
(Owens et al., 2013; Rowatt et al., 2006). However, a possible
weakness of this approach was the high overlap between judg-
ments of IH and other, conceptually related constructs (McCrae &
Costa, 1997; Roberts & Wood, 2003; Samuelson et al., 2014).
Although it is possible that IH and IA are genuinely two ends of a
single construct, the high overlap also observed for evaluations of
agreeableness do indicate that one limitation of the relational
B.R. Meagher et al. / Journal of Research in Personality 58 (2015) 35–45 43

measure may be reduced discriminant validity. Scales with a larger
number of items may help to address this issue (McElroy et al.,
2014), but the time required to complete long scales increases
exponentially when conducting research with round-robin groups.
An additional concern for a relational measure of IH was the
finding that the magnitude of consensus within these tested
groups was fairly small. There was essentially no agreement at
minimal acquaintance in Study 1, and consensus accounted for
only about 11% of the variance after several months of collabora-
tive group work. The viability of this method may therefore be lim-
ited, as a relational estimate can function usefully as an IH
measurement only in the specific contexts where the group shows
agreement (Kenny, 1994). As a result, a primary goal for future
work must be to better theorize and assess how IH is revealed to
others. Answering this question will involve two empirical direc-
tions. First, researchers should seek to determine the types of
behaviors and other individual features people associate with IH.
In the current work, the consensus found for IH and IA was related
to self-ratings of extraversion, dominance, and a desire to be the
center of attention. However, these impressions provide only a
general sense of how IH manifests in behavior. A more
fine-grained analysis or coding of interpersonal interactions and
communication within groups, such as perspective taking (Epley,
Keysar, Van Boven, & Gilovich, 2004; Moll & Kadipasaoglu, 2013),
speech convergence (Pickering & Garrod, 2004), and behavioral
synchrony (Lakin, 2013), may reveal the more subtle interpersonal
information that reflects a focus on and openness to others’ ideas.
Secondly, researchers should propose and test the features of social
situations expected to elicit and reveal IH to observers. Several
authors have already proposed several situationally defined vari-
ables believed to elicit humility, such as during interpersonal con-
flict, interacting along across established social hierarchies, or
when asked to reminisce about past achievements (Chancellor &
Lyubomirsky, 2013; Davis et al., 2010, 2011; Kruse et al., 2014).
However, these claims have not been empirically tested.
Importantly, although Study 1 adopted certain features of the pro-
cedure used by Davis et al. (2013), it failed to reveal IH reliably to
participating groups. What types of procedures, settings, and time
periods are necessary to reliably measure IH interpersonally there-
fore remains an open question. This lack of research at the level of
situations reflects a general dearth within social psychology and a
failure to fully develop taxonomic understanding of the settings
and interactions that characterize human life (Funder, 2009; Reis,
2008), despite their demonstrated importance in eliciting and
revealing particular traits (Fleeson, 2007). Nevertheless, a handful
of existing frameworks do exist to provide a starting point for this
work; for example, in terms of dyadic motivations (Kelley et al.,
2003), domains of sociality (Fiske, 1992), values that constrain
behavior (Hodges, 2007; Hodges et al., 2014), or ecological units
(Barker, 1968). Moreover, the recent development of relevant mea-
surement tools, such as the Riverside Situational Q-Sort
(Rauthmann et al., 2014; Sherman, Nave, & Funder, 2010), provide
a means of systematically assessing and classifying situations for
this purpose. These approaches can provide researchers with the
language needed to begin hypothesizing the different scenarios
in which IH is potentially necessary and likely to be detected reli-
ably by perceivers.
7. Conclusion
The scientific study of IH is a field still in its infancy. As with the
study of humility generally, a key initial focus of inquiry must
therefore be to determine what is actually being measured when
individuals report high or low levels of IH about themselves or
others. The two studies described here indicate that
self-assessment and other-assessment appear to tap into fairly dis-
tinct qualities and outcomes. Self-ratings of IH were predictive of
self-enhancement for a variety of traits in both of these studies,
highlighting the fact that socially desirable reporting is a clear
issue in the self-report of this construct. Relational IH ratings, on
the other hand, were more consistent with expectations for the
construct, showing no relationship with self-enhancement and a
positive directional association with academic performance.
Nevertheless, this approach is limited to social contexts where
groups can actually reach consensus in their judgments, and the
extent to which observers can disentangle impressions of IH with
other, related constructs, such as IA and agreeableness, remains
unclear. Thus, future work is needed to investigate just what
behavioral or situational information specifies peer assessments
of IH, in order to better understand how it is expressed and
detected within group contexts.
Acknowledgments
This research was supported by a generous Grant from The John
Templeton Foundation (#29630). We thank Skylar Brannon,
Pauline Minnaar, and Kristin Smart for their assistance in data col-
lection in Study 1.
References
Ames, D. R., Rose, P., & Anderson, C. P. (2006). The NPI-16 as a short measure of
narcissism. Journal of Research in Personality, 40(4), 440–450. http://dx.doi.org/
10.1016/j.jrp.2005.03.002.
Barker, R. G. (1968). Ecological psychology: Concepts and methods for studying the
environment of human behavior. Stanford, CA: Stanford University Press.
Baumeister, R. F., & Vohs, K. D. (2012). Self-regulation and the executive function of
the self. In M. R. Leary & J. P. Tangney (Eds.), Handbook of self and identity (2nd
ed., pp. 180–197). New York, NY, US: Guilford Press.
Bouchard, T. J., Jr., & Hare, M. (1970). Size, performance, and potential in
brainstorming groups. Journal of Applied Psychology, 54(1), 51–55.
Carlisle, R. D., & Tsang, J.-A. (2013). The virtues: Gratitude and forgiveness. In K. I.
Pargament, J. J. Exline, & J. W. Jones (Eds.), APA handbook of psychology, religion,
and spirituality (vol. 1): Context, theory, and research (pp. 423–437). Washington,
DC, US: American Psychological Association.
Carlson, E. N., Vazire, S., & Oltmanns, T. F. (2011). You probably think this paper’s
about you: Narcissists’ perceptions of their personality and reputation. Journal
of Personality and Social Psychology, 101(1), 185–201. http://dx.doi.org/10.1037/
a0023781.
Chancellor, J., & Lyubomirsky, S. (2013). Humble beginnings: Current trends, state
perspectives, and hallmarks of humility. Social and Personality Psychology
Compass, 7(11), 819–833. http://dx.doi.org/10.1111/spc3.12069.
Cooper, L. D., Balsis, S., & Oltmanns, T. F. (2012). Self- and informant-reported
perspectives on symptoms of narcissistic personality disorder. Personality
Disorders: Theory, Research, and Treatment, 3(2), 140–154. http://dx.doi.org/
10.1037/a0026576.
Costa, P. T., Jr., & McCrae, R. R. (1985). The NEO personality inventory manual. Odessa,
FL: Psychological Assessment Resources.
Davis, D. E., Hook, J. N., Worthington, E. L., Jr., Van Tongeren, D. R., Gartner, A. L.,
Jennings, D. J., et al. (2011). Relational humility: Conceptualizing and measuring
humility as a personality judgment. Journal of Personality Assessment, 93(3),
225–234. http://dx.doi.org/10.1080/00223891.2011.558871.
Davis, D. E., Rice, K. G., McElroy, S. E., DeBlaere, C., & Choe, E. (2015). Distinguishing
intellectual humility and general humility. The Journal of Positive Psychology.
http://dx.doi.org/10.1080/17439760.2015.1048818.
Davis, D. E., Worthington, E. L., Jr., & Hook, J. N. (2010). Humility: Review of
measurement strategies and conceptualization as personality judgment. The
Journal of Positive Psychology, 5(4), 243–252. http://dx.doi.org/10.1080/
17439761003791672.
Davis, D. E., Worthington, E. L., Jr., Hook, J. N., Emmons, R. A., Hill, P. C., Bollinger, R.
A., et al. (2013). Humility and the development and repair of social bonds: Two
longitudinal studies. Self and Identity, 12(1), 58–77. http://dx.doi.org/10.1080/
15298868.2011.636509.
de Vries, R. E., Lee, K., & Ashton, M. C. (2008). The Dutch HEXACO Personality
Inventory: Psychometric properties, self-other agreement, and relations with
psychopathy among low and high acquaintanceship dyads. Journal of Personality
Assessment, 90(2), 142–151. http://dx.doi.org/10.1080/00223890701845195.
Epley, N., Keysar, B., Van Boven, L., & Gilovich, T. (2004). Perspective taking as
egocentric anchoring and adjustment. Journal of Personality and Social
Psychology, 87(3), 327–339. http://dx.doi.org/10.1037/0022-3514.87.3.327.
Exline, J. J., Campbell, W. K., Baumeister, R. F., Joiner, T., Krueger, J., & Kachorek, L. V.
(2004). Humility and modesty. In C. Peterson & M. E. P. Seligman (Eds.),
44 B.R. Meagher et al. / Journal of Research in Personality 58 (2015) 35–45

Character strengths and virtues: A handbook and classification (pp. 461–475). New
York, NY: Oxford University Press.
Exline, J. J., & Geyer, A. L. (2004). Perceptions of humility: A preliminary study. Self
and Identity, 3(2), 95–114. http://dx.doi.org/10.1080/13576500342000077.
Fehr, B. (2013). The social psychology of love. In J. A. Simpson & L. Campbell (Eds.),
The Oxford handbook of close relationships (pp. 201–233). New York, NY, US:
Oxford University Press.
Fiske, A. P. (1992). The four elementary forms of sociality: Framework for a unified
theory of social relations. Psychological Review, 99(4), 689–723.
Fleeson, W. (2007). Situation-based contingencies underlying trait-content
manifestation in behavior. Journal of Personality, 75(4), 825–862. http://
dx.doi.org/10.1111/j.1467-6494.2007.00458.x.
Funder, D. C. (1995). On the accuracy of personality judgment: A realistic approach.
Psychological Review, 102(4), 652–670. http://dx.doi.org/10.1037/0033-
295x.102.4.652.
Funder, D. C. (2009). Persons, behaviors and situations: An agenda for personality
psychology in the postwar era. Journal of Research in Personality, 43(2), 120–126.
http://dx.doi.org/10.1016/j.jrp.2008.12.041.
Gosling, S. D., Rentfrow, P. J., & Swann, W. B. Jr., (2003). A very brief measure of the
Big-Five personality domains. Journal of Research in Personality, 37(6), 504–528.
http://dx.doi.org/10.1016/s0092-6566(03)00046-1.
Gregg, A. P., & Mahadevan, N. (2014). Intellectual arrogance and intellectual
humility: An evolutionary-epistemological account. Journal of Psychology and
Theology, 42(1), 7–18.
Hodges, B. H. (2007). Values define fields: The intentional dynamics of driving,
carrying, leading, negotiating, and conversing. Ecological Psychology, 19(2),
153–178. http://dx.doi.org/10.1080/10407410701332080.
Hodges, B. H., Meagher, B. R., Norton, D., McBain, R., & Kimball, A. (2014). Speaking
from ignorance: Not agreeing with others we believe are correct. Journal of
Personality and Social Psychology, 106(2), 218–234. http://dx.doi.org/10.1037/
a0034662.
Johnson, R. E., Silverman, S. B., Shyamsunder, A., Swee, H.-Y., Rodopman, O. B., Cho,
E., et al. (2010). Acting superior but actually inferior?: Correlates and
consequences of workplace arrogance. Human Performance, 23(5), 403–427.
http://dx.doi.org/10.1080/08959285.2010.515279.
Kelley, H. H., Holmes, J. G., Kerr, N. L., Reis, H. T., Rusbult, C. E., & Van Lange, P. A. M.
(2003). An atlas of interpersonal situations. New York, NY, US: Cambridge
University Press.
Kenny, D. A. (1994). Interpersonal perception: A social relations analysis. New York, NY
US: Guilford Press.
Kenny, D. A. (1998). Social relations model program: Computer software and
instructions downloaded from <http://davidakenny.net/srm/srmp.htm>.
Kenny, D. A. (2004). PERSON: A general model of interpersonal perception.
Personality and Social Psychology Review, 8(3), 265–280. http://dx.doi.org/
10.1207/s15327957pspr0803_3.
Kenny, D. A., Albright, L., Malloy, T. E., & Kashy, D. A. (1994). Consensus in
interpersonal perception: Acquaintance and the big five. Psychological Bulletin,
116(2), 245–258. http://dx.doi.org/10.1037/0033-2909.116.2.245.
Krumrei-Mancuso, E. J., & Rouse, S. V. (in press). The development and validation of
the comprehensive intellectual humility scale. Journal of Personality Assessment.
Kruse, E., Chancellor, J., Ruberton, P. M., & Lyubomirsky, S. (2014). An upward spiral
between gratitude and humility. Social Psychological and Personality Science,
5(7), 805–814. http://dx.doi.org/10.1177/1948550614534700.
Kwan, V. S. Y., John, O. P., Kenny, D. A., Bond, M. H., & Robins, R. W. (2004).
Reconceptualizing individual differences in self-enhancement bias: An
interpersonal approach. Psychological Review, 111(1), 94–110. http://
dx.doi.org/10.1037/0033-295x.111.1.94.
Lakin, J. L. (2013). Behavioral mimicry and interpersonal synchrony. In J. A. Hall & M.
L. Knapp (Eds.), Nonverbal communication (pp. 539–575). Boston, MA, US: De
Gruyter Mouton.
Lee, K., & Ashton, M. C. (2004). Psychometric properties of the HEXACO personality
inventory. Multivariate Behavioral Research, 39(2), 329–358. http://dx.doi.org/
10.1207/s15327906mbr3902_8.
Lukowitsky, M. R., & Pincus, A. L. (2013). Interpersonal perception of pathological
narcissism: A social relations analysis. Journal of Personality Assessment, 95(3),
261–273. http://dx.doi.org/10.1080/00223891.2013.765881.
McCrae, R. R., & Costa, P. T. Jr., (1997). Personality trait structure as a human
universal. American Psychologist, 52(5), 509–516. http://dx.doi.org/10.1037/
0003-066X.52.5.509.
McElroy, S. E., Rice, K. G., Davis, D. E., Hook, J. N., Hill, P. C., Worthington, E. L. Jr.,,
et al. (2014). Intellectual humility: Scale development and theoretical
elaborations in the context of religious leadership. Journal of Psychology and
Theology, 42(1), 19–30.
Michaelsen, L. K., Knight, A. B., & Fink, L. D. (2002). Team-based learning: A
transformative use of small groups. Westport, CT: Praeger.
Moll, H., & Kadipasaoglu, D. (2013). The primacy of social over visual perspective-
taking. Frontiers in Human Neuroscience, 7.http://dx.doi.org/10.3389/
fnhum.2013.00558.
Owens, B. P., Johnson, M. D., & Mitchell, T. R. (2013). Expressed humility in
organizations: Implications for performance, teams, and leadership.
Organization Science, 24(5), 1517–1538. http://dx.doi.org/10.1287/orsc.1120.
0795.
Peterson, C., & Seligman, M. E. P. (2004). Character strengths and virtues: A handbook
and classification. New York, NY: Oxford University Press.
Pickering, M. J., & Garrod, S. (2004). Toward a mechanistic psychology of dialogue.
Behavioral and Brain Sciences, 27(2), 169–226.
Polgreen, L., & Goodstein, L. (2006). At axis of Episcopal split, an anti-gay Nigerian.
The New York Times. Retrieved from <http://www.nytimes.com>.
Rauthmann, J. F., Gallardo-Pujol, D., Guillaume, E. M., Todd, E., Nave, C. S., Sherman,
R. A., et al. (2014). The Situational Eight DIAMONDS: A taxonomy of major
dimensions of situation characteristics. Journal of Personality and Social
Psychology, 107(4), 677–718. http://dx.doi.org/10.1037/a0037250.
Reis, H. T. (2008). Reinvigorating the concept of situation in social psychology.
Personality and Social Psychology Review, 12(4), 311–329. http://dx.doi.org/
10.1177/1088868308321721.
Roberts, C. R., & Wood, W. J. (2003). Humility and epistemic goods. In M. DePaul & L.
Zagzebski (Eds.), Intellectual virtue: Perspectives from ethics and epistemology
(pp. 257–279). New York, NY: Oxford University Press.
Rowatt, W. C., Powers, C., Targhetta, V., Comer, J., Kennedy, S., & Labouff, J. (2006).
Development and initial validation of an implicit measure of humility relative
to arrogance. The Journal of Positive Psychology, 1(4), 198–211. http://dx.doi.org/
10.1080/17439760600885671.
Samuelson, P. L., Jarvinen, M. J., Paulus, T. B., Church, I. M., Hardy, S. A., & Barrett, J. L.
(2014). Implicit theories of intellectual virtues and vices: A focus on intellectual
humility. The Journal of Positive Psychology, 1–18.http://dx.doi.org/10.1080/
17439760.2014.967802.
Sherman, R. A., Nave, C. S., & Funder, D. C. (2010). Situational similarity and
personality predict behavioral consistency. Journal of Personality and Social
Psychology, 99(2), 330–343. http://dx.doi.org/10.1037/a0019796.
Tangney, J. P. (2009). Humility. In S. J. Lopez & C. R. Snyder (Eds.), Oxford handbook of
positive psychology (2nd ed., pp. 483–490). New York, NY, US: Oxford University
Press.
Thrive Center for Human Development (2014). Intellectual Humility Grants.
Retrieved from <http://thethrivecenter.org/intellectualhumility/intellectual-
humilityawardees/>.
Vazire, S. (2010). Who knows what about a person? The self-other knowledge
asymmetry (SOKA) model. Journal of Personality and Social Psychology, 98(2),
281–300. http://dx.doi.org/10.1037/a0017908.
B.R. Meagher et al. / Journal of Research in Personality 58 (2015) 35–45 45