Content uploaded by Daniel A. Newman
Author content
All content in this area was uploaded by Daniel A. Newman on Nov 24, 2014
Content may be subject to copyright.
Emotional Intelligence: An Integrative Meta-Analysis and
Cascading Model
Dana L. Joseph and Daniel A. Newman
University of Illinois at Urbana–Champaign
Research and valid practice in emotional intelligence (EI) have been impeded by lack of theoretical
clarity regarding (a) the relative roles of emotion perception, emotion understanding, and emotion
regulation facets in explaining job performance; (b) conceptual redundancy of EI with cognitive
intelligence and Big Five personality; and (c) application of the EI label to 2 distinct sets of constructs
(i.e., ability-based EI and mixed-based EI). In the current article, the authors propose and then test a
theoretical model that integrates these factors. They specify a progressive (cascading) pattern among
ability-based EI facets, in which emotion perception must causally precede emotion understanding,
which in turn precedes conscious emotion regulation and job performance. The sequential elements in
this progressive model are believed to selectively reflect Conscientiousness, cognitive ability, and
Neuroticism, respectively. “Mixed-based” measures of EI are expected to explain variance in job
performance beyond cognitive ability and personality. The cascading model of EI is empirically
confirmed via meta-analytic data, although relationships between ability-based EI and job performance
are shown to be inconsistent (i.e., EI positively predicts performance for high emotional labor jobs and
negatively predicts performance for low emotional labor jobs). Gender and race differences in EI are also
meta-analyzed. Implications for linking the EI fad in personnel selection to established psychological
theory are discussed.
Keywords: emotional intelligence, emotion regulation, job performance, personality, sex differences
During the last 20 years, emotional intelligence (EI) has become
an increasingly popular topic within the fields of psychology and
management (Grandey, 2000; Law, Wong, & Song, 2004; Mayer,
Roberts, & Barsade, 2008). The impressive growth of EI in schol-
arly work (for a review, see Matthews, Zeidner, & Roberts, 2002)
has been partially fueled by claims that EI is as strong a predictor
of job performance as is IQ (Goleman, 1995). This purported
relationship between EI and work performance has also stimulated
interest among human resource practitioners, who have made EI a
widely used tool for personnel hiring and training (Fineman,
2004). As evidence of this, a September 2008 count showed at
least 57 consulting firms devoted principally to EI, 90 organiza-
tions that specialize in training or assessment of EI, 30 EI certifi-
cation programs, and 5 EI “universities” (see www.eq.org).
Despite its commercial and academic expansion, many impor-
tant questions about the theoretical bases of EI remain. These
include issues of the respective roles of EI subfacets in the EI
model (Mayer & Salovey, 1997), as well as connections of EI to
more mainstream psychological theory on emotion regulation
(Grandey, Fisk, & Steiner, 2005; Gross, 1998b) and individual
differences (Law et al., 2004; McCrae, 2000).
Historically speaking, the study of social intelligence has a
surprisingly lengthy and empirically disappointing record (Landy,
2005, 2006; Matthews et al., 2002). Although EI researchers tend
to attribute the first mention of social intelligence to Thorndike
(1920; Bar-On, 2000; Mayer & Geher, 1996; Mayer & Salovey,
1993), Dewey introduced the concept in 1909 (see Landy, 2006).
After 50 years of research, Cronbach claimed that “social intelli-
gence remains undefined and unmeasured” (1960, p. 319). Nearly
100 years following the first mention of social intelligence, evi-
dence still seems weak (with a few notable exceptions; e.g.,
Olson-Buchanan et al., 1998). In regard to the latest incarnation of
the social intelligence concept—“emotional intelligence”—critics
remain dubious on the definition and measurement of EI and
whether EI has incremental validity in organizational contexts
beyond personality traits and cognitive ability (Landy, 2005;
Locke, 2005; Murphy, 2006; cf. Van Rooy & Viswesvaran, 2004).
Consequently, the current state of EI is somewhat paradoxical;
although EI is a wildly popular tool in organizations, organiza-
tional science has yet to answer many theoretical, measurement,
and validity questions surrounding the construct.
In order to begin addressing these issues, this paper makes five
contributions to theory on EI and work behavior. First, we propose
and test an original conceptual model of emotional intelligence
(called the “cascading model”) by laying out the theoretical causal
mechanisms among EI subfacets (Salovey & Mayer, 1990), job
performance, and other individual differences. This model high-
lights the extent to which emotion perception, emotion understand-
Dana L. Joseph and Daniel A. Newman, Department of Psychology,
University of Illinois at Urbana–Champaign.
An earlier version of this article was presented at the annual conference
of the Society for I/O Psychology, New York City, April 2007. We thank
Winfred Arthur, Jr., Murray Barrick, Gerianne Alexander, Steven Jarrett,
Mike McDaniel, and Rob Ployhart for their help and/or advice on previous
drafts of this article.
Correspondence concerning this article should be addressed to Dana L.
Joseph, Department of Psychology, University of Illinois at Urbana–
Champaign, 603 East Daniel Street, Champaign, IL 61820. E-mail:
dlrhodes@gmail.com
Journal of Applied Psychology © 2010 American Psychological Association
2010, Vol. 95, No. 1, 54–78 0021-9010/10/$12.00 DOI: 10.1037/a0017286
54
ing, and emotion regulation fit a progressive structure, in which
emotion perception causally precedes emotion understanding,
which in turn gives rise to conscious emotion regulation and job
performance. The cascading model also specifies the role of EI
vis-a`-vis relevant personality traits of Conscientiousness, Emo-
tional Stability (Barrick, Mount, & Judge, 2001), and cognitive
intelligence (F. L. Schmidt & Hunter, 2004); thus, it provides an
integrative empirical test of social psychological and personality
theory in the context of work organizations. Second, we distin-
guish three construct–method pairings that have dominated the
empirical work on EI (labeled performance-based ability tests of
EI, self-report ability tests of EI, and self-report mixed EI) to
reveal that these alternative conceptualizations and operations of
EI play three distinct roles in the emotional intelligence–work
performance relationship. Third, we examine sex- and race-based
subgroup differences in EI and how these differences vary sub-
stantially with the measure used. Fourth, in order to test our
integrative model and estimate incremental validity and subgroup
differences in EI, we conducted over a dozen original meta-
analyses on the relationships among EI, Conscientiousness, Emo-
tional Stability, cognitive ability, job performance, and demo-
graphics. We then combined these original meta-analyses with
published meta-analytic effects to estimate the final model. Fifth,
we coded the emotional labor content of the jobs included in our
meta-analyses to assess whether emotional labor moderates the
validity (and incremental validity) of EI. The empirical evidence
validates our theoretical view that EI facets tend to have a causal
ordering, EI tests can favor women, and EI is a better predictor of
job performance in high emotional labor jobs; but the evidence
also reveals that mixed-based EI is an empirically stronger (albeit
theoretically weaker) predictor of job performance than is ability-
based EI.
What Is EI? Ability Versus Mixed Models
Currently, there are two popular construct models available with
which to define EI: (a) an ability model and (b) a mixed (traits with
abilities) model (Mayer, Salovey, & Caruso, 2000). Ability mod-
els, originally conceptualized by Mayer et al. (2000), propose that
EI is a type of intelligence or aptitude and therefore should overlap
with cognitive ability. Ability models posit EI as “the ability to
carry out accurate reasoning about emotions and the ability to use
emotions and emotional knowledge to enhance thought” (Mayer et
al., 2008, p. 511).
In contrast to ability models, mixed EI models do not classify EI
as an intelligence but rather as a combination of intellect and
various measures of personality and affect (Petrides & Furnham,
2001). For example, Bar-On’s (1997) mixed model defines EI as
“an array of noncognitive capabilities, competencies, and skills
that influence one’s ability to succeed in coping with environmen-
tal demands and pressures” (p. 14). Mixed model definitions of EI
are the source of many EI criticisms because (a) they appear to
define EI by exclusion as any desirable characteristic not repre-
sented by cognitive ability (Elfenbein, 2008; Locke, 2005; Mat-
thews et al., 2002; Murphy, 2006; Zeidner, Matthews, & Roberts,
2004) and (b) they are too redundant with personality traits to
justify a distinct construct (Conte, 2005; Daus & Ashkanasy, 2003;
Van Rooy, Dilchert, Viswesvaran, & Ones, 2006). As a result,
some have concluded that only ability EI models are worth study-
ing (Daus & Ashkanasy, 2005) or at least that mixed models are
profoundly flawed (through lack of empirical bases and overly
broad conceptualization; Murphy, 2006).
In summary, there are two senses in which the term emotional
intelligence has been used: (a) as a narrow, theoretically specified
set of constructs pertaining to the recognition and control of
personal emotion (called ability-based EI) and (b) as an umbrella
term for a broad array of constructs that are connected only by
their nonredundancy with cognitive intelligence (called mixed-
based EI). Due to the lack of scientific rigor often associated with
mixed-based models of EI, the current paper begins by developing
a theoretical model of EI and job performance that focuses on the
ability-based EI model.
A Cascading Model of EI
According to the ability-based model, EI can be broken down
into four subdimensions: emotion perception, emotion understand-
ing, emotion facilitation, and emotion regulation (Mayer &
Salovey, 1997). We use this conceptualization of EI to propose a
cascading model of EI, in which three of the four EI subdimen-
sions are related to job performance in a sequential fashion, as
shown in Figure 1. We note that the third dimension of EI, emotion
facilitation, is not included in our cascading model of EI. The
choice to exclude the emotion facilitation facet from our model
was made a priori, due to its increasingly well-known conceptual
redundancy with other EI dimensions and its lack of empirical
support. Gignac (2005); Palmer, Gignac, Manocha, and Stough
(2005); and Rossen, Kranzler, and Algina (2008) have demon-
strated the poor fit of EI factor analytic models that included the
dimension of emotion facilitation and the superior fit for EI models
from which this dimension was removed. Poor construct validity
evidence for the emotion facilitation facet is due in part to theo-
retical ambiguity over how emotion facilitation differs at its core
from emotion regulation, the fourth dimension of EI. For example,
Salovey and Mayer (1990) posit that emotion facilitation involves
using emotion in a variety of contexts to facilitate the attainment of
goals. For goal attainment, using emotion must essentially involve
the induction of an emotion, such as the induction of a positive
(e.g., joy) or negative (e.g., anger) emotion, which is conceptually
redundant with regulating positive or negative emotion (Cole,
Martin, & Dennis, 2004; Gross, 1998b). Due to this conceptual
redundancy and because empirical research has shown a lack of
construct validity for the emotion facilitation facet, our theoretical
cascading model does not include this aspect of EI.
Emotion Regulation and Job Performance
In developing the cascading model of EI, we draw on theories of
emotion, emotion regulation, and self-regulation (Gailliott, Mead,
& Baumeister, 2008; Gross, 2008) in order to answer calls for a
theoretical elaboration of EI and its purported relationship with job
performance (Zeidner et al., 2004). In so doing, we focus on the
role of emotion regulation as the key dimension of EI that influ-
ences job performance. Emotion regulation has been defined as
“the processes by which individuals influence which emotions
they have, when they have them, and how they experience and
express these emotions” (Gross, 1998b, p. 275). Although emo-
tions (e.g., surprise, joy, anger, sadness) can be distinguished from
55
EMOTIONAL INTELLIGENCE: AN INTEGRATIVE REVIEW
moods (e.g., pleasant and unpleasant state affect)—in that emo-
tions arise from a particular cause and correspond to specific
action tendencies (e.g., escape, attack; Frijda, 1986; Lazarus,
1991)—in the current paper we consider both types of state affect
to be subject to emotion regulation. Within an organizational
setting, emotion regulation is theoretically related to job perfor-
mance through the induction of affective states that are beneficial
to job performance. That is, emotion regulation is the tool through
which we create and maintain positive affective states, which have
been suggested to benefit work behavior (George, 1991). In sup-
port of the advantages of positive affective states, Fredrickson’s
broaden-and-build theory (2001) proposes that positive emotions
broaden behavioral repertoires, improve behavioral flexibility, and
increase attentional scope, all of which may enhance job perfor-
mance. Further developing the link between positive affective
states and job performance, Tsai, Chen, and Liu (2007) have
developed a model of the relationship between positive moods and
job performance in which moods predict task performance indi-
rectly through interpersonal processes (helping other coworkers
and being helped by coworkers) and motivational processes (self-
efficacy and task persistence). Initial evidence appears to support
this model (Tsai et al., 2007), and previous work on the relation-
ship between positive moods and task performance also suggests
the two are positively related (Eisenberger, Armeli, Rexwinkel,
Lynch, & Rhoades, 2001; Erez & Isen, 2002; Hirt, Melton,
McDonald, & Harackiewicz, 1996; Totterdell, 1999, 2000). Thus,
we might expect that emotion regulation processes allow an indi-
vidual to induce and sustain a positive affective state, which
subsequently promotes helping behavior and motivation, and ulti-
mately job performance.
Literature on emotion regulation involves not only the induction
of moods but also the suppression of moods (Gross, 1998b).
Emotion suppression involves inhibiting an emotion-expressive
behavior (Gross, 1998a), and it is generally thought of as having
negative consequences (Butler et al., 2003; Gross & John, 2003).
Individuals with high ability to regulate emotion would likely less
often engage in the strategy of suppression and instead engage in
a more effective (i.e., less cognitively taxing) strategy, such as
cognitive reappraisal (Butler et al., 2003). By so doing, individuals
with high emotion regulation competence would retain more cog-
nitive resources to devote to task performance. For all the reasons
stated above, in our cascading model of EI we propose a positive
relationship between the ability to regulate emotion and job per-
formance.
Hypothesis 1: Emotion regulation ability is positively related
to job performance.
Although emotion regulation may enhance job performance
through the management of affective states, we must address a
prominent alternative theoretical viewpoint. Resource allocation
theory (Kahneman, 1973; Kanfer & Ackerman, 1989; D. A. Nor-
man & Bobrow, 1975) suggests there may actually be a negative
relationship between emotion regulation and job performance,
because emotion regulation demands our attentional resources and
can draw attention from the task at hand (Beal, Weiss, Barros, &
MacDermid, 2005). The work of Kanfer, Ackerman, and col-
leagues supports this idea with evidence that emotion regulation
strategies have the strongest relationship with performance when
attentional demands of the task are low (Kanfer & Ackerman,
1990, 1996; Kanfer, Ackerman, & Heggestad, 1996; i.e., when
emotion regulation and job performance are not competing for
resources). Similarly, self-regulation theory supports the idea that
we have a finite resource pool (Baumeister, 2002; Muraven, Tice,
& Baumeister, 1998; Vohs & Heatherton, 2000) from which
emotion regulation draws. This makes subsequent regulation, in-
cluding basic decision making (Vohs et al., 2008) and persistence
(Baumeister, Bratslavsky, Muraven, & Tice, 1998), more difficult.
Therefore, although our previous discussions have focused on the
positive relationship between emotion regulation and job perfor-
Job
Performance
Emotion
Regulation
Emotion
Understanding
Emotion
Perception
Emotional Stabilit
y
Co
g
nitive Abilit
y
Conscientiousness
Figure 1. Cascading model of emotional intelligence (EI). The cascading model is based on the ability EI
concept (Mayer & Salovey, 1997) and incorporates three subfacets of performance-based EI: emotion percep-
tion, emotion understanding, and emotion regulation.
56 JOSEPH AND NEWMAN
mance, a limited-resource perspective alternatively proposes a
negative relationship between the two.
The emotional labor literature has discussed the extent to which
different emotion regulation processes “drain” resources. In par-
ticular, surface acting, or the modification of facial expression,
requires more attention and effort than does deep acting,orthe
modification of inner feelings (Brotheridge & Grandey, 2002;
Brotheridge & Lee, 2002; Coˆte´, 2005; Diefendorff & Gosserand,
2003; Goldberg & Grandey, 2007; Grandey, 2003; Grandey et al.,
2005). Surface acting also results in more stress than does deep
acting (Beal, Trougakos, Weiss, & Green, 2006; Grandey, 2003;
Grandey, Dickter, & Sin, 2004; Totterdell & Holmann, 2003).
Thus, we would expect surface acting to elicit a greater drain on
resources, which would also drain the resources available for job
performance, than would deep acting. Emotion regulation litera-
ture has a similar taxonomy in antecedent-focused regulation and
response-focused regulation, of which the former is proposed to be
a greater drain on several types of resources (Gross, 1998a; Gross
& John, 2003). In sum, it appears that the extent to which emotion
regulation drains the resources available for job performance is
dependent on the type of emotion regulation process employed.
With regard to Hypothesis 1, it is our belief that emotion
regulation ability (a facet of EI) includes the ability to select
emotion regulation strategies that are relatively less draining of
personal resources. As such, implicit in Hypothesis 1 is the idea
that individuals high in emotion regulation intelligence will match
their chosen regulation strategy (surface vs. deep acting;
antecedent- vs. response-focused) to the demands of the task and
to the momentary store of personal resources at hand, so as to
maintain overall job performance.
Perceiving and Understanding Emotions
Although we have focused on the ability to regulate emotion, we
also propose that emotion regulation is preceded by two other
dimensions of EI: the ability to perceive emotion and the ability to
understand emotion (see Figure 1). To explain these relationships,
we draw on Gross and Thompson’s (2007) “modal” model of
emotion, which proposes a sequence of events: a stimulus requires
attention, then appraisal, and ultimately a response. Deriving from
this basic model of emotion, our cascading model of EI is designed
to capture the ability to complete each of these steps, in order, and
thus is an attempt to align EI theory with more traditional theory
of emotion.
The first step in models of emotion (Elfenbein, 2008; Gross &
Thompson, 2007), attention, corresponds to the EI facet of ability
to perceive emotion. As defined by Mayer and Salovey’s (1997)
ability model, emotion perception refers to “the ability to identify
emotions in oneself and others, as well as in other stimuli, includ-
ing voices, stories, music, and works of art” (Brackett, Rivers,
Shiffman, Lerner, & Salovey, 2006, p. 781). We note that Mayer
and Salovey’s measure of EI taps only perception of emotion in
others and not emotional self-perception. Data from self-report EI
measures suggest there may be sizable overlap between the ability
to perceive self-emotion and to perceive others’ emotion (Joseph
& Newman, in press; Wong & Law, 2002). Therefore, the current
paper treats the ability to perceive emotion in the self and others as
part of the same construct, emotion perception. Ekman and col-
leagues’ research has shown there are considerable individual
differences in the ability to perceive emotion (Ekman &
O’Sullivan, 1991; Matsumoto et al., 2000). Recognizing that in-
dividual differences in emotion perception exist, we expect that
individuals who are more aware of the verbal and nonverbal cues
in their environments, as well as their own emotional states,
subsequently have a larger base of emotional information. The
accrual of a larger and more accurate base of emotional informa-
tion then enables more accurate appraisal (i.e., Step 2 in basic
models of emotion; Gross & Thompson, 2007) and more appro-
priate response formation (i.e., Step 3 in basic models of emotion).
However, the reverse is not true; an individual who appraises and
responds appropriately to emotion may not subsequently perceive
emotion more accurately. Therefore, we place ability to perceive
emotion as the first step in the cascading model of EI.
The second step of basic models of emotion (Elfenbein, 2008;
Gross & Thompson, 2007), appraisal, is captured in the cascading
model of EI by the ability to understand emotion. The ability to
understand emotion entails understanding how emotions evolve
over time, how emotions differ from each other, and which emo-
tion is most appropriate for a given context (Mayer & Salovey,
1997). As such, the ability to understand emotion essentially refers
to a set of knowledge structures involving the generic origins and
consequences of emotional states. To place this definition in the
context of general cognitive ability (as is appropriate, given our
focus on the ability model of EI), our representation of the ability
to understand emotion as knowledge about emotion is analogous to
Humphreys’(1979) definition of cognitive ability as a individual’s
repertoire of knowledge and skills at any given point in time (cf.
Jensen, 1994). That is, abilities such as the ability to understand
emotion can be conceptualized as accumulated knowledge struc-
tures.
It is these knowledge structures that determine how an emotion
is appraised. A large body of literature on the cognitive appraisal
of emotion has helped clarify the dimensions on which we appraise
emotion and the processes that we engage in during cognitive
appraisal of emotion (Frijda 1986; Lazarus 1968, 1991; Roseman,
1979, 1984; C. A. Smith & Ellsworth, 1985). For example, Laza-
rus has proposed two appraisal processes: primary appraisal, which
answers the question “Does this situation affect me personally?”
and secondary appraisal, which answers the question “What if
anything can be done about this situation?” (Lazarus & Folkman,
1984, p. 31). How an individual answers these questions reflects
his or her emotion understanding. Models of knowledge activation
(Kintsch, 2000) show that attention, or in this case, emotion
perception, can activate relevant knowledge structures, and the
extent to which these knowledge structures are accurate embodies
the ability to understand emotion. For example, an individual may
have an inaccurate schema of anger as inflexible. Thus, when this
person perceives anger, and subsequently engages in appraisal, his
or her answer to the question “What if anything can be done about
this situation?” will be driven by the understanding of anger as
fixed and uncontrollable. This simple example illustrates how the
ability to understand emotion is affected by our ability to accu-
rately perceive emotion and in turn influences how we can respond
to/regulate that emotion.
As such, emotion understanding is expected to mediate the
relationship between emotion perception and emotion regulation
abilities (cf. James, Mulaik, & Brett, 2006). But is this full or
partial mediation? It is possible to imagine contexts in which the
57
EMOTIONAL INTELLIGENCE: AN INTEGRATIVE REVIEW
relationship between emotion perception and one’s ability to reg-
ulate emotion does not rely on accurate emotion understanding
(e.g., the perception of fear can lead to automatic down-regulation
of fear, even without knowledge of how the fear developed or of
the nature of the fear itself). That is, emotion perception can
directly affect emotion regulation if this process occurs automat-
ically or without voluntary control.
However, Mayer and Salovey defined the concept of emotion
regulation as the “conscious [emphasis added] regulation of emo-
tions to enhance emotional and intellectual growth” (1997, p. 14).
This suggests that unconscious regulation similar to that of
the automatic down-regulation of fear should not be included
in the cascading model of EI. The exclusion of unconscious
emotion regulation is consistent with the literature on general
self-regulation abilities, which separates effortful or conscious
self-regulation from automatic or unconscious self-regulation due
to their distinct neurological origins, antecedents, and outcomes
(for a review, see Eisenberg, Smith, Sadovsky, & Spinrad, 2004).
Because EI was originally conceptualized as a model of conscious
regulation (Mayer & Salovey, 1997) and self-regulation theories
suggest that voluntary and involuntary emotion regulation are
dissimilar enough not to be described with one model, our cascad-
ing model focuses solely on conscious processes. Therefore, the
automatic processes that allow the perception of emotion to di-
rectly influence the regulation of emotion are not included in the
cascading model. As a result, we expect the ability to understand
emotion to completely mediate the relationship between the ability
to perceive emotion and the ability to regulate emotion, because
we are dealing with a conscious emotion regulation process.
As an analogy, the EI cascading model is similar to a skill
development model by which schoolchildren learn responsive
writing skills. That is, children first learn to read words, then learn
reading comprehension, and finally develop the capability for
responsive writing. Similarly, emotion perception ability provides
an opportunity to develop emotion understanding, which in turn
provides an opportunity for the development of emotion regulation
skill. According to this perspective, emotion understanding is an
essential intermediate step.
Hypothesis 2: Emotion understanding will fully mediate the
effect of emotion perception on emotion regulation.
In summary, we have drawn on theory of emotion, emotion
regulation, and self-regulation in order to develop the cascading
model of EI, shown in Figure 1. We place particular importance on
the pattern of relationships among emotion perception, emotion
understanding, and emotion regulation, with emotion regulation
specified as the final step in enhancing job performance through
emotional competence. The ability to regulate emotion is preceded
by the ability to understand emotion and perceive emotion.
Incorporating Personality and Cognitive Ability Into
the Cascading Model
Having presented a causal rationale for the relationships be-
tween EI facets and job performance, we now acknowledge other
important constructs with which EI might potentially overlap in
the explanation of performance: personality traits and cognitive
ability. Cognitive ability robustly contributes to job performance,
due to its connection with accrued job knowledge and skills
(Hunter & Hunter, 1984; F. L. Schmidt, Hunter, & Outerbridge,
1986). Strong performance in most jobs also requires the individ-
ual to be responsible and achievement driven (Conscientious) as
well as low in anxiety, insecurity, and depression (Emotionally
Stable; Barrick & Mount, 2000).
According to Barrick, Mitchell, and Stewart’s (2003) full mo-
tivational mediator model, the personality traits of Conscientious-
ness and Emotional Stability are the only two Big Five traits that
universally predict overall job performance. The reasoning is that
these two traits uniquely give rise to a motivational mediator
labeled accomplishment strivings, which is a midlevel goal struc-
ture (Emmons, 1989) implicated in task-oriented behavior. Prior
empirical research has been consistent with this viewpoint (see
review of findings on Conscientiousness and Emotional Stability
by Hogan & Holland, 2003). Thus both theoretical and empirical
evidence pinpoint the unique connections of these two Big Five
traits with job performance. Our conceptual model of EI (see
Figure 1) does not incorporate other personality traits (e.g., Extra-
version, Agreeableness), as these traits have no theoretical con-
nection with overall job performance (Barrick & Mount, 2000);
thus, their inclusion is unnecessary for a fully specified structural
model. We nevertheless continue to include all five of the Big Five
traits in our regression-based estimation of operational incremental
validity, presented later.
We propose here that the dimensions of EI may serve as partial
mechanisms by which cognitive ability and the personality traits of
Conscientiousness and Emotional Stability influence job perfor-
mance. Conscientious individuals have been described as thor-
ough, organized, methodical, cautious, and careful (McCrae &
Costa, 1992) and have been shown to pay much greater attention
to detail (Nigg et al., 2002). Although these adjectives describe
Conscientiousness as a behaviorally oriented trait, Conscientious-
ness has been described as an emotionally oriented trait as well.
Conscientious individuals have shown above-average levels of
interpersonal functioning (Jensen-Campbell & Malcolm, 2007)
and increased capacity for self-conscious emotions (Tracy & Rob-
ins, 2004), such as guilt and shame (Abe, 2003; Einstein &
Lanning, 1998). It has been suggested that the impulse control
facet of Conscientiousness relies on the experience of guilt and
shame to guide socially appropriate behavior (Roberts, Jackson,
Fayard, Edmonds, & Meints, in press). Thus, conscientious indi-
viduals may develop a heightened perception of self-conscious
emotions as a sort of radar to detect when they have lost control of
their behavior. We therefore expect Conscientiousness to be pos-
itively related to emotion perception in the self. In a similar
manner, conscientious individuals may use the emotional cues
from others to guide their need for controlled behavior. That is, a
conscientious person would likely develop the ability to read
emotional cues in his or her environment in order to determine
when a behavior is appropriate or inappropriate (Matsumoto et al.,
2000). Given these theoretical links between Conscientiousness
and emotion perception in the self and others, we expect Consci-
entiousness to be positively related to emotion perception ability.
Hypothesis 3: Conscientiousness is positively related to emo-
tion perception ability.
58 JOSEPH AND NEWMAN
In addition, we propose that the relationship between cognitive
ability and performance involves the EI dimension of emotion
understanding. We endorse two definitions of cognitive ability: (a)
Humphreys’ (1979, p. 106) definition of cognitive ability as an
individual’s “entire repertoire of acquired skills, knowledge, learn-
ing sets, and generalization tendencies considered intellectual in
nature that is available at any one period of time” and (b) Acker-
man’s (1996) conceptualization of intelligence (i.e., intelligence-
as-process, personality, interests, and intelligence-as-knowledge).
Both definitions recognize that knowledge is a component of
cognitive ability. Moreover, it has been shown that the knowledge-
related component of cognitive ability is the primary avenue by
which cognitive ability influences job performance (F. L. Schmidt
& Hunter, 2004). That is, individuals with higher cognitive ability
acquire more job-related knowledge, which increases job perfor-
mance. As previously mentioned, the ability to understand emotion
represents a body of knowledge concerning which emotions are
appropriate in a given context (Mayer & Salovey, 1997). Thus, we
propose that individuals with high cognitive ability would acquire
a stronger knowledge base associated with understanding one’s
emotions. Empirical support for this hypothesis has been demon-
strated in the domain of emotion knowledge, where verbal ability
has been related to children’s understanding of emotion (Fine,
Izard, & Trentacosta, 2006).
Hypothesis 4: Cognitive ability is positively related to
emotion-understanding ability.
Finally, we propose that Emotional Stability is a disposition
underlying the EI dimension of emotion regulation. Emotional
Stability has been described as a lack of emotionality (W. T.
Norman, 1963), Neuroticism (Eysenck, 1970), and anxiety
(Cattell, 1957). A large body of work has demonstrated that
neurotic individuals experience higher levels of trait negative
affect (Gross, Sutton, & Ketelaar, 1998) and hyperreactivity to
daily stressors in the form of negative mood states even after
controlling for prior mood (Marco & Suls, 1993; Suls, Green, &
Hillis, 1998). Neurotic individuals also show more frequent use
of emotion-focused coping, or coping focused at regulating
emotional reactions (Connor-Smith & Flachsbart, 2007; Shew-
chuk, Elliott, MacNair-Semands, & Harkins, 1999). These em-
pirical findings present an interesting paradox: Even though
neurotic individuals attempt to regulate their affect more fre-
quently than do emotionally stable individuals, they report
tonically high levels of negative affect. This leads us to ask why
neurotic individuals’ frequent attempts to regulate emotion fail.
We believe that neurotic individuals lack the ability to regulate
emotion effectively. To put it differently, we propose that
neurotic individuals are using ineffective emotion regulation
strategies. For example, neurotic individuals may be engaging
in surface acting, or the modification of facial expression
(Grandey, 2003), which does not modify the inner emotional
state or eliminate a negative mood. Recent work supports our
suspicion that neurotic individuals do not engage in effective
emotion regulation strategies (i.e., reappraisal) as often as emo-
tionally stable individuals do (Gross & John, 2003). Thus, we
expect Emotional Stability to be positively related to the ability
to regulate emotion.
Hypothesis 5: Emotional Stability is positively related to
emotion regulation ability.
In sum, the cascading model of EI (see Figure 1) presents
emotion perception, emotion understanding, and emotion regula-
tion as partial mediators of the effects of Conscientiousness, cog-
nitive ability, and Emotional Stability on performance, respec-
tively.
Convergent, Discriminant, and Incremental Validity
of Various EIs
The cascading model of EI (see Figure 1) incorporates our
fundamental theoretical ideas and predictions about EI dimensions,
job performance, cognitive ability, and personality. As previously
mentioned, Figure 1 is based upon an ability model of EI (Daus,
2006; Mayer et al., 2000).
At this point, we recognize that mixed models of EI, despite
their criticisms, are very common in the science and practice of EI.
This alternative formulation (i.e., mixed EI) involves inclusive,
compound conceptualizations of EI that combine the EI compe-
tencies reviewed above with many other tendencies, such as mo-
tivation (e.g., need for achievement), social styles (e.g., assertive-
ness), self-esteem, impulse control (Mayer et al., 2008), and a
“grab bag” of other concepts that are only loosely connected. Due
to the popularity of measures based on mixed models of EI and the
broad array of constructs they assess, we feel it is necessary to
compare these models to the more narrowly defined ability models
in an attempt to assess whether it is appropriate to label both of
these models as models of emotional intelligence.
In order to compare the EI literatures regarding ability and
mixed EI models, we first point out a construct–method distinction
(Arthur & Villado, 2008). That is, ability and mixed models of EI
can each be measured via self-report or via performance-based
tests (e.g., multiple-choice measures where answers are scored as
correct or incorrect). Previous meta-analytic work (Van Rooy,
Viswesvaran, & Pluta, 2005) has not distinguished the construct
(ability vs. mixed EI) from the method (self-report vs. perfor-
mance-based). That is, Van Rooy et al. coded all self-report
measures as mixed EI and all performance-based measures as
ability EI—despite the fact that some self-report measures are
based on the ability EI model (e.g., the Wong and Law Emotional
Intelligence Scale; Wong & Law, 2002). The current study draws
distinctions among the various construct–method pairings of EI by
crossing the construct distinction (ability vs. mixed) with the
method distinction (self-report vs. performance-based). This re-
sults in three distinct construct–method pairings of EI: perfor-
mance-based ability EI, self-report ability EI, and self-report
mixed EI (for the fourth construct–method pairing, performance-
based mixed EI, no measures currently exist). Given these
construct–method pairings, we can now investigate the discrimi-
nant, convergent, and incremental validity of ability and mixed
models of EI.
Discriminant, Convergent, and Incremental Validity of
EI Construct–Method Pairings
Ability models of EI conceptualize EI as an intelligence, and
mixed models define EI as a combination of intelligence, person-
59
EMOTIONAL INTELLIGENCE: AN INTEGRATIVE REVIEW
ality, and affect (Mayer et al. 2000). Although many of the
underlying premises of the ability and mixed models may be
similar (Ciarrochi, Chan, & Caputi, 2000), empirical evidence
suggests “they [ability and mixed EI models] diverge more than
they converge (ˆability,mixed ⫽.24), indicating that two different
constructs are being tapped” (Van Rooy, Viswesvaran, & Pluta,
2005, p. 453). If ability EI and mixed EI are in fact distinct
constructs, we would expect to find evidence for discriminant
validity, or evidence showing these constructs are less-than-
perfectly related (Campbell & Fiske, 1959).
Convergent validity, in contrast, exists when two measures of
the same construct are strongly related (Campbell & Fiske, 1959).
Given the construct–method pairings of EI, we expect to find
evidence of convergent validity between self-report measures of
ability EI and performance-based measures of ability EI because
they are both assessing the same construct (ability EI). That is,
self-reports of ability EI should show a strong relationship with
performance-based measures of ability EI.
Finally, because we expect a strong relationship between mea-
sures of the same construct and a weak relationship between
measures of different constructs, we can make predictions regard-
ing the incremental validity of these measures over each other. In
theory, measures of the same construct should overlap and thus
should capture no additional variance over each other in the
prediction of relevant outcomes. We therefore suspect that self-
report ability EI measures and performance-based ability EI mea-
sures will exhibit little (if any) incremental validity over each other
in predicting job performance. In contrast, we should find that
mixed EI and ability EI measures exhibit significant incremental
validity over each other in the prediction of job performance.
Hypothesis 6: EI construct–method pairings will exhibit ev-
idence of (a) convergent validity (strong relationship of self-
report ability EI with performance-based ability EI), (b) dis-
criminant validity (weak relationship of ability EI with mixed
EI), and (c) incremental validity (of ability EI and mixed EI
over each other) in predicting job performance.
Incremental Validity
Aside from the conceptual models of EI and job performance
presented above, there exists another important question from a
practitioner perspective: “How large is the incremental validity of
EI for predicting job performance, over and above cognitive ability
and personality?” We answer this question by examining the
incremental validity of the three construct–method EI pairings.
Because ability measures of EI are based on a conceptualization of
EI as an intelligence (Mayer et al., 2000), it is expected that these
measures will show a strong relationship with cognitive ability. As
such, we do not expect ability measures of EI to exhibit meaning-
ful incremental validity over cognitive ability. However, it is
expected that ability measures of EI will exhibit significant incre-
mental validity over Big Five personality traits, because the five-
factor model of personality does not include elements of ability.
Hypothesis 7: Measures of ability EI (self-report and perfor-
mance-based) will exhibit significant incremental validity
over measures of Big Five personality.
In contrast, measures of mixed EI are based on a definition of EI
that emphasizes aspects of personality, intelligence, and affect
(Mayer & Salovey, 1997; Petrides & Furnham, 2001). Given this
definition, we expect measures of mixed EI to overlap somewhat
with Big Five personality and cognitive ability measures in their
prediction of job performance. However, because measures of
mixed EI include not only content related to personality and
intelligence but also content covering a host of other individual
differences, including affect, self-efficacy, and motivation (e.g.,
Schutte et al., 1998), mixed measures are expected to contribute
significant unique variance to the prediction of job performance.
This hypothesis contrasts with previous authors’ suggestions of
very little (if any) incremental validity of mixed EI over well-
established personality and ability constructs (e.g., Landy, 2005).
Hypothesis 8: Measures of mixed EI will exhibit incremental
validity over cognitive ability and Big Five personality.
Subgroup Differences
In addition to concerns regarding the questionable incremental
validity of EI, concerns also exist regarding subgroup differences
on EI (Conte & Dean, 2006). For example, there is a growing
tendency among EI researchers to conclude that women score
higher than men on measures of emotional intelligence (Van Rooy
et al., 2006). Our own explanation for this finding stems from the
large body of literature on sex differences in emotion. In particular,
women have been shown to be better at perceiving nonverbal
emotion cues (Hall, 1978, 1984; McClure, 2000) and to have more
complex emotion knowledge (Ciarrochi, Hynes, & Crittenden,
2005), which could contribute to higher EI scores in women.
Differences in female–male EI can also be explained through the
extreme male brain theory of autism (Baron-Cohen, 2002), which
suggests that men tend to “systemize” (i.e., to analyze the world in
a series of “if-then” rules) more than women and that women tend
to “empathize” (i.e., to attribute mental states to others and respond
with appropriate affective responses) more than men. This expla-
nation points to differences in female and male cognition, with
women using emotion more often and more appropriately than do
men. Ultimately, we would expect these sex differences to result in
higher EI scores for women.
Hypothesis 9: Women will score higher on EI measures than
will men.
Regarding race differences, previous work suggests near-zero
racial subgroup differences (Van Rooy, Alonso, & Viswesvaran,
2005) in measures of EI. However, because we have no theoretical
reason to expect race differences on measures of EI, we investigate
race differences in an exploratory manner in the current paper.
Method
In order to test the cascading model of EI, we constructed a
correlation matrix based on meta-analytic estimates (as recom-
mended by Viswesvaran & Ones, 1995). These estimates include
21 published meta-analytic correlations plus 66 original meta-
analyses (see Tables 1, 2, and 6). Existing correlation estimates
(see Table 3) based on the 21 published meta-analyses were
corrected for attenuation in the predictor and criterion. Correla-
60 JOSEPH AND NEWMAN
tions with job performance were also corrected for direct range
restriction (Sackett & Yang, 2000; Thorndike, 1949).
In correcting for range restriction in the observed EI validities,
we used the unrestricted standard deviations reported in EI inven-
tory manuals (cf. Salgado, 1997). When this method was used, the
sample-weighted average ratio of restricted to unrestricted stan-
dard deviation was .99 for performance-based EI and .83 for
self-reported EI. To correct Big Five personality validities for
range restriction, we used a .92 standard deviation ratio (Hurtz &
Donovan, 2000, p. 873). Cognitive ability validities were corrected
for range restriction by using the average standard deviation ratio
of .67 (Hunter & Hunter, 1984; Hunter, Schmidt, & Le, 2006,
p. 601).
The minimum sample size for personality intercorrelations is
135,529 (Ones, 1993), and the minimum Nfor correlations be-
tween personality factors and cognitive ability is 11,190 (Judge,
Jackson, Shaw, Scott, & Rich, 2007). Job performance validities
come from Hunter and Hunter (1984; N⫽32,124, for cognitive
tests) and from an updated version of Hurtz and Donovan (2000;
minimum N⫽7,797 for Big Five personality inventories). That is,
we updated Hurtz and Donovan’s (2000) meta-analyses of Big
Five personality traits and job performance through 2008.
Table 1
Meta-Analytic Results for Performance-Based Ability Emotional Intelligence Dimensions With Personality Traits, Cognitive Ability,
and Job Performance
95% CI 80% CI
kN r ˆSD LL UL LL UL % variance
Conscientiousness
Emotion perception 23 3,582 .25 .28 .34 .13 .38 ⫺.16 .71 6
Emotion understanding 22 3,374 .07 .09 .10 .02 .12 ⫺.04 .21 51
Emotion facilitation 23 3,582 .09 .11 .11 .04 .14 ⫺.02 .24 48
Emotion regulation 22 3,374 .13 .16 .09 .08 .18 .04 .28 53
Emotional Stability
Emotion perception 24 3,696 .11 .12 .02 .07 .14 .10 .14 94
Emotion understanding 22 3,374 .08 .09 .08 .03 .12 ⫺.01 .19 63
Emotion facilitation 23 3,582 .09 .11 .03 .05 .13 .07 .15 88
Emotion regulation 22 3,374 .14 .17 .16 .07 .20 ⫺.04 .38 27
Agreeableness
Emotion perception 23 3,582 .13 .15 .07 .09 .17 .06 .24 64
Emotion understanding 22 3,374 .09 .12 .04 .05 .13 .06 .17 84
Emotion facilitation 23 3,582 .13 .17 .02 .10 .16 .13 .20 91
Emotion regulation 22 3,374 .23 .30 .03 .19 .27 .26 .34 75
Extraversion
Emotion perception 24 3,696 .08 .09 .04 .04 .11 .03 .14 82
Emotion understanding 22 3,374 .06 .07 .11 .01 .11 ⫺.07 .22 45
Emotion facilitation 23 3,582 .08 .10 .06 .04 .12 .02 .18 70
Emotion regulation 22 3,374 .14 .18 .09 .10 .19 .06 .29 53
Openness
Emotion perception 23 3,582 .06 .07 .10 .01 .11 ⫺.05 .20 49
Emotion understanding 22 3,374 .14 .18 .14 .07 .19 .00 .36 36
Emotion facilitation 23 3,582 .08 .10 .14 .02 .13 ⫺.08 .27 37
Emotion regulation 22 3,374 .12 .16 .13 .07 .18 ⫺.01 .33 37
Cognitive ability
Emotion perception 21 4,710 .09 .10 .05 .05 .12 .04 .16 72
Emotion understanding 20 4,581 .31 .39 .15 .25 .37 .20 .58 17
Emotion facilitation 18 3,971 .15 .18 .15 .08 .21 ⫺.01 .36 22
Emotion regulation 19 4,277 .13 .16 .06 .09 .17 .09 .24 59
Job performance
ⴱ
Emotion perception 8 562 .08 .10 .01 ⫺.02 .18 ⫺.02 .22 67
High emotional labor 4 220 .18 .21 .00 .06 .29 .21 .21 100
Low emotional labor 3 223 .01 .01 .01 ⫺.14 .16 ⫺.09 .11 76
Emotion understanding 8 562 .13 .15 .01 .03 .23 .02 .28 67
High emotional labor 4 220 .19 .22 .00 .08 .29 .22 .22 100
Low emotional labor 3 223 .02 .02 .00 ⫺.10 .15 .02 .02 100
Emotion facilitation 8 562 .05 .07 .00 ⫺.04 .14 .07 .07 86
High emotional labor 4 220 .10 .12 .00 .00 .21 .12 .12 100
Low emotional labor 3 223 ⫺.04 ⫺.04 .00 ⫺.13 .05 ⫺.04 ⫺.04 100
Emotion regulation 8 562 .15 .18 .02 .03 .27 ⫺.01 .38 45
High emotional labor 4 220 .22 .26 .03 .03 .41 .04 .48 44
Low emotional labor 3 223 .01 .01 .00 ⫺.04 .06 .01 .01 100
Note.
ⴱ
Emotional labor is a moderator of the emotional intelligence-job performance relationships; k⫽number of effect sizes in the meta-analysis; N⫽
total sample size in the meta-analysis; r⫽sample-size weighted mean correlation; ˆ⫽correlation corrected for attenuation and range restriction;
SD ⫽standard deviation of corrected correlation; CI ⫽confidence interval; LL ⫽lower limit; UL ⫽upper limit; % variance ⫽percent of variance
accounted for by sampling error.
61
EMOTIONAL INTELLIGENCE: AN INTEGRATIVE REVIEW
To update the meta-analyses of Big Five validities, we ob-
tained studies from the reference list of Hurtz and Donovan
(2000). Of the 26 studies included in Hurtz and Donovan’s
original meta-analysis, four could not be located. All of the
studies that could not be located were conference papers for
which the corresponding author did not respond to a request for
the paper. Of the 22 studies that could be located, four were
removed from the analysis because the authors used training
performance as the criterion rather than job performance, which
is the focus of the current paper (in all, the remaining 18 studies
yielded k⫽35 independent samples).
To update this collection of studies, a keyword search in
PsycINFO (1887-1008), Google Scholar, and Dissertation Ab-
stracts International (1861–2008) was conducted for keywords
personality, Big Five personality, five factor model, job perfor-
mance, and several variations of these keywords for the years
between 1996 [the last year of studies included in the original
Hurtz and Donovan (2000) meta-analysis] and 2008. The keyword
search found 675 articles, which were each coded for the following
inclusion criteria: (a) the study examined actual workers; (b) the
study measured personality with an inventory designed to assess
Big Five personality traits (personality measures included Hogan
Personality Inventory, Goldberg’s Big 5 markers, NEO (PI/PI–R/
FFI), 16PF, Global Personality Inventory, and Personal Character-
istics Inventory); (c) the study reported enough information to
calculate a correlation between at least one Big Five personality
trait and job performance; and (d) the study used supervisor ratings
of job performance. As a result of the inclusion criteria, 33 inde-
pendent samples were added to the original collection of studies
reported in Hurtz and Donovan (2000), for a final collection of 68
independent samples.
Original Meta-Analyses
We conducted several original meta-analyses to estimate the
correlations involving emotional intelligence. To identify studies
for inclusion, we conducted searches of the American Psycholog-
ical Association’s PsycINFO (1887–2008), Google Scholar, and
Table 2
Meta-Analytic Results for Construct–Method Pairings of Emotional Intelligence, Big Five Personality Traits, Cognitive Ability, and
Job Performance
95% CI 80% CI
kN rˆSD LL UL LL UL % variance
Conscientiousness 60 18,462 .28 .32 .17 .24 .32 .11 .54 11
Self-report mixed EI 31 5,591 .33 .38 .17 .27 .39 .16 .60 16
Self-report ability EI 27 8,566 .32 .38 .10 .28 .36 .26 .51 23
Performance-based EI 21 4,155 .12 .13 .10 .07 .16 .00 .26 39
Emotional Stability 60 18,416 .33 .39 .24 .28 .38 .08 .69 6
Self-report mixed EI 30 5,386 .45 .53 .22 .38 .53 .25 .81 8
Self-report ability EI 26 8,479 .34 .40 .14 .28 .39 .22 .59 13
Performance-based EI 22 4,401 .17 .20 .26 .08 .27 ⫺.13 .54 9
Agreeableness 59 18,302 .28 .34 .16 .25 .32 .14 .54 13
Self-report mixed EI 30 5,386 .36 .43 .13 .31 .41 .27 .59 21
Self-report ability EI 26 8,479 .26 .31 .13 .21 .30 .14 .48 17
Performance-based EI 23 4,287 .25 .29 .15 .18 .31 .09 .48 22
Extraversion 60 18,450 .28 .33 .27 .22 .34 ⫺.01 .67 5
Self-report mixed EI 30 5,552 .40 .46 .13 .35 .45 .29 .63 18
Self-report ability EI 26 8,479 .27 .32 .28 .18 .37 ⫺.04 .69 5
Performance-based EI 23 4,269 .15 .18 .26 .05 .24 ⫺.15 .51 9
Openness 58 18,170 .23 .27 .20 .19 .28 .02 .53 10
Self-report mixed EI 30 5,386 .26 .29 .20 .19 .32 .04 .55 15
Self-report ability EI 26 8,479 .24 .29 .19 .18 .31 .05 .54 10
Performance-based EI 21 4,155 .18 .21 .18 .11 .25 ⫺.01 .44 18
Cognitive ability 54 10,519 .13 .16 .18 .09 .17 ⫺.07 .39 21
Self-report mixed EI 19 2,880 .09 .11 .17 .03 .15 ⫺.10 .33 26
Self-report ability EI 16 2,158 .00 .00 .08 ⫺.05 .05 ⫺.11 .10 56
Performance-based EI 28 5,538 .22 .25 .13 .17 .27 .08 .42 28
Job performance 22 2,593 .24 .32 .04 .16 .31 .08 .56 23
Self-report mixed EI 9 1,110 .32 .47 .00 .25 .39 .47 .47 51
High emotional labor 4 270 .37 .59 .01 .25 .48 .46 .72 72
Low emotional labor 3 300 .27 .43 .00 .18 .35 .43 .43 100
Self-report ability EI 7 835 .17 .23 .00 .10 .25 .15 .31 73
High emotional labor 7 516 .20 .28 .02 .10 .35 .09 .47 41
Low emotional labor 4 390 .14 .20 .00 .002 .05 .20 .20 100
Performance-based EI 10 887 .16 .18 .01 .08 .24 .06 .30 61
High emotional labor 4 220 .22 .24 .00 .13 .31 .24 .24 100
Low emotional labor 3 223 .00 .01 .00 ⫺.12 .14 .01 .01 100
Note. k ⫽number of effect sizes in the meta-analysis; N⫽total sample size in the meta-analysis; r⫽sample-size weighted mean correlation; ˆ⫽
correlation corrected for attenuation and range restriction; SD ⫽standard deviation of corrected correlation; CI ⫽confidence interval; LL ⫽lower limit;
UL ⫽upper limit; % variance ⫽percent of variance accounted for by sampling error.
62 JOSEPH AND NEWMAN
Table 3
Meta-Analytic Correlation Matrix of Emotional Intelligence (EI), Performance, Personality, Cognitive Ability, and Emotional Labor
1234567 8 9 10 11 12 13
1. Performance-based ability EI —
2. Emotion perception .76
a
—
13/2,679
3. Emotion understanding .82
a
.46
a
—
13/2,679 13/2,679
4. Emotion facilitation .87
a
.63
a
.62
a
—
13/2,679 13/2,679 13/2,679
5. Emotion regulation .76
a
.34
a
.55
a
.53
a
—
13/2,679 13/2,679 13/2,679 13/2,679
6. Self-report ability EI .12
a
.09
a
.10
a
.25
a
.37
a
—
14/2,426 6/969 6/969 6/969 6/969
7. Self-report mixed EI .26
a
.14
a
.17
a
.19
a
.33
a
.59
a
—
10/1,572 10/1,572 10/1,572 10/1,572 10/1,572 9/2,945
8. Conscientiousness .13
a
.28
a
.09
a
.11
a
.16
a
.38
a
.38
a
—
21/4,155 23/3,582 22/3,374 23/3,582 22/3,374 27/8,566 31/5,591
9. Emotional Stability .20
a
.12
a
.09
a
.11
a
.17
a
.40
a
.53
a
.26
b
—
22/4,401 24/3,696 22/3,374 23/3,582 22/3,374 26/8,479 30/5,386 587/490,296
10. Agreeableness .29
a
.15
a
.12
a
.17
a
.30
a
.31
a
.43
a
.27
b
.25
b
—
23/4,287 23/3,582 22/3,374 23/3,582 22/3,374 26/8,479 30/5,386 344/162,975 561/415,679
11. Extraversion .18
a
.09
a
.07
a
.10
a
.18
a
.32
a
.46
a
.00
b
.19
b
.17
b
—
23/4,269 24/3,696 22/3,374 23/3,582 22/3,374 26/8,479 30/5,552 632/683,001 710/440,440 234/135,529
12. Openness .21
a
.07
a
.18
a
.10
a
.16
a
.29
a
.29
a
⫺.06
b
.16
b
.11
b
.17
b
—
21/4,155 23/3,582 22/3,374 23/3,582 22/3,374 26/8,479 30/5,386 338/356,680 423/254,937 236/144,205 418/252,004
13. Cognitive ability .25
a
.10
a
.39
a
.18
a
.16
a
.00
a
.11
a
⫺.04
c
.09
c
.00
c
.02
c
.22
c
—
28/5,538 21/4,710 20/4,581 18/3,971 19/4,277 16/2,158 19/2,880 56/15,429 61/21,404 38/11,190 61/21,602 46/13,182
14. Job performance .18
a
.10
a
.15
a
.07
a
.18
a
.23
a
.47
a
.21
d
.11
d
.07
d
.09
d
.06
d
.44
e
10/887 8/562 8/562 8/562 8/562 7/835 9/1,110 64/12,434 53/9,184 56/9,702 56/9,664 48/7,797 425/32,124
High emotional labor .24
a
.21
a
.22
a
.12
a
.26
a
.28
a
.59
a
.20
a
.10
a
.09
a
.09
a
.10
a
.37
a
4/220 4/220 4/220 4/220 4/220 7/516 4/270 31/4,673 26/3,846 28/4,172 29/4,364 21/3,225 122/4,454
Low emotional labor .01
a
.01
a
.02
a
⫺.04
a
.01
a
.20
a
.43
a
.26
a
.15
a
.12
a
.10
a
.04
a
.39
a
3/223 3/223 3/223 3/223 3/223 4/390 3/300 26/4,967 21/3,638 23/3,935 22/3,705 21/2,872 361/30,160
Note. Each cell contains the correlation corrected for attenuation in the predictor and criterion, as well as range restriction, followed by knumber of effect sizes and Nsample size.
a
Original meta-analysis.
b
Ones (1993).
c
Judge et al. (2007).
d
Hurtz & Donovan, 2000 (updated through 2008), for Big Five correlations with job performance (uncorrected r,SD [95% CI]) ⫽
Conscientiousness, .15, .01 [.13, .18]; Emotional Stability, .08, .00 [.06, .10]; Agreeableness, .05, .02 [.02, .08]; Extraversion, .06, .01 [.04, .09]; Openness, .04, .01 [.01,.07].
e
Hunter & Hunter (1984;
uncorrected r⫽.26).
63
EMOTIONAL INTELLIGENCE: AN INTEGRATIVE REVIEW
Dissertation Abstracts International (1861–2008) for the following
keywords (and several variations thereof): emotional intelligence,
cognitive ability, personality, job performance, race, and sex.
Studies used in meta-analyses by Van Rooy and colleagues (Van
Rooy & Viswesvaran, 2004; Van Rooy, Viswesvaran, & Pluta,
2005) were also obtained from their reference lists. References of
all available studies and relevant reviews were searched for studies
that were missed in previous searches, and several authors were
contacted for unpublished work relating to emotional intelligence.
This search identified 171 studies that were then examined for
congruence with several inclusion criteria. Following Van Rooy
and Viswesvaran (2004), a study was included only if it used a
measure that was specifically referred to as a measure of emotional
intelligence. The remaining studies were then examined for rela-
tionships with job performance, cognitive ability, and personality
and intercorrelations of EI measures and EI subdimensions.
When collecting estimates of the relationship between EI and
job performance, we invoked especially high standards for what
would count as job performance. Although this decision yields
smaller meta-analytic sample sizes, it provides higher quality data
for drawing inferences to EI personnel selection scenarios. Studies
of the relationship between EI and job performance were included
if (a) enough information to calculate a correlation between EI and
job performance was provided, (b) ratings of job performance were
provided by a supervisor (not self-reported), and (c) the study
involved employed individuals (this does not include students
acting as if they were managers who provide performance ratings
of students acting as subordinates; e.g., Day & Carroll, 2004).
Studies were also excluded if (a) job performance was manipulated
or (b) academic performance was considered job performance
(e.g., Holbrook, 1997).
Primary studies of the relationship between EI and cognitive
ability were excluded from the analysis if they used student GPA
as a measure of cognitive ability. Studies of relationships between
EI and personality were included if a measure of Big Five person-
ality traits was administered and enough information was provided
to calculate a correlation with any one or more of the Big Five
personality traits. In addition, all studies being considered for the
current meta-analysis were required to provide the sample size and
to consist primarily of adult participants (over age 16). These
inclusion criteria resulted in a final database of 118 usable studies,
with a total sample size of 30,077.
Data Coding
Studies that passed the inclusion criteria were coded on several
attributes. Each study was coded for an effect size between EI and
job performance, personality, cognitive ability, sex, or race, as well
as for measures used to assess the relevant variables, reliability of
these measures, sample size, and participant characteristics. All
measures of EI were also coded for the construct measured (ability
or mixed EI) and method employed (self-report or performance-
based). Any EI measure purported to be based on an ability model
was subsequently classified as an ability measure. This group
included the following EI measures: MSCEIT (Mayer, Salovey,
Caruso, & Sitarenios, 2003), MEIS (Mayer, Caruso, & Salovey,
1999), WLEIS (Wong & Law, 2002), EIS (Schutte et al., 1998),
and WEIP (Jordan, Ashkanasy, Hartel, & Hooper, 2002). All
other EI measures were classified as mixed measures. Regard-
ing the method of each EI measure, any measure scored by
marking questions as correct or incorrect based on expert or
consensus scores was considered a performance-based measure (i.e.,
MSCEIT and MEIS), and all other measures were coded as self-report
EI measures. As a result of the construct and method coding, all EI
measures were classified as one of the following construct–method
pairings: self-report ability EI, self-report mixed EI, performance-
based ability EI, or performance-based mixed EI. Because no studies
involved performance-based mixed measures of EI, we hereafter
refer to performance-based ability EI measures as performance-
based EI measures. Finally, all performance-based measures of EI
were coded for relationships between dimensions of EI and any
relevant variable (see Table 3). In order to determine the accuracy
of the coding process, two researchers coded all articles identified
in the original search. The agreement between the two coders at the
item level was 98%, and any disagreements were discussed and
resolved between the coders.
Data Analysis
The current meta-analysis followed procedures outlined by
Hunter and Schmidt (2004). Because the current meta-analysis is
an attempt to determine the theoretical relationship between EI and
various variables, all effect sizes were corrected for range restric-
tion and attenuation due to measurement error in both predictor
and criterion. In an effort to use independent sample effects, we
included only one effect size per sample in each meta-analysis. If
a sample provided multiple, facet-level effect sizes for one rela-
tionship, a composite correlation was constructed (Nunnally, 1978;
see Judge, Thoresen, Bono, & Patton, 2001).
In order to test the fit of the cascading model of EI (see Figure
1), we used the meta-matrix shown in Table 3 (see Shadish, 1996).
For the cascading model of EI (see Figure 1), correlations involv-
ing the subdimensions of EI are based only on the MEIS and
MSCEIT measures of EI. The MEIS and MSCEIT are the only two
performance-based EI measures on which facet-level data were
available.
To assess incremental validity, we ran a series of multiple
regression models based upon the meta-matrix in Table 4. Table
4 contains correlations that are corrected only for range restric-
tion and measurement error in the criterion (i.e., practical,
operational validities). Table 4 correlations between Big Five
personality traits and cognitive ability found in Judge et al.
(2007) were uncorrected (i.e., we attenuated them, to calculate
operational validities). We assumed reliability of .90 for cog-
nitive ability and used the unit-weighted internal consistency
reliabilities found in Viswesvaran and Ones (2000, p. 231) for
the Big Five personality measures. Finally, we calculated
Black–White and female–male meta-analytic subgroup dvalues
for each construct–method EI pairing.
Results
Meta-Analytic Results
Results of the meta-analyses are presented in Tables 1, 2, 3, and
6. The intercorrelations among EI construct–method pairings (see
Table 3) are not what one would expect from different measures of
the same construct. In particular, the attenuation-corrected corre-
64 JOSEPH AND NEWMAN
lation between self-report ability EI and performance-based ability
EI is only .12, which suggests that these measures may not both be
measuring ability EI as purported. A similarly low correlation is
found between performance-based EI and self-report mixed EI
(ˆ⫽.26; cf. Van Rooy & Viswesvaran, 2004). The correlation
between self-report ability EI and self-report mixed EI (ˆ⫽.59)
is substantial, suggesting that these measures may tap into a similar
construct (affirming the decision of Van Rooy, Viswesvaran, and
colleagues to lump these two EI self-reports together in their
analyses). Hence, although we report correlations of overall EI
with personality, cognitive ability, and job performance in Table 2,
we warn that the concept of overall EI (averaged across the three
construct–method pairings) is of limited conceptual value due to
inconsistent and low correlations among some types of EI mea-
sures.
Correlations between ability-based EI dimensions and relevant
variables are reported in Table 1. As predicted, the EI facet of
emotion regulation shows a positive relationship with job perfor-
mance, ˆ⫽.18, 95% CI [.03, .27]. However, the effect size has a
credibility interval that includes zero and the percentage of vari-
ance accounted for by sampling error is 45%, suggesting there are
substantive moderators of the relationship between emotion regu-
lation and job performance. We also found positive hypothesized
relationships between Conscientiousness and emotion perception,
ˆ⫽.28, 95% CI [.13, .38]; cognitive ability and emotion under-
standing, ˆ⫽.39, 95% CI [.25, .37]; and Emotional Stability and
emotion regulation, ˆ⫽.17, 95% CI [.07, .20].
When we look at self-reported EI (see Table 2), the mixed EI
correlations with Big Five personality traits are relatively large
共ˆAgreeableness ⫽.43, ˆConscientiousness ⫽.38, ˆEmotionalStability ⫽.53,
ˆExtraversion ⫽.46, ˆOpenness ⫽.29), which is consistent with the
conceptualization of mixed EI as a mixture of personality traits,
motivation, and affect (Bar-On, 1997; Goleman, 1995). Also con-
sistent with this conceptualization, there is a small relationship
between mixed EI and cognitive ability (ˆ⫽.11).
Findings involving self-reported mixed EI were echoed by find-
ings for self-reported ability-based EI measures (see Table 2).
Self-reported ability EI (like self-reported mixed EI) exhibited
moderate associations with Big Five personality (ˆAgreeableness ⫽.31,
ˆConscientiousness ⫽.38, ˆEmotionalStability ⫽.40, ˆExtraversion ⫽.32, ˆOpenness ⫽
.29) but a nil relationship with cognitive ability (ˆ⫽.00). The
latter correlation is a bit surprising, because it brings into question
the labeling of these measurements as ability-based measurements
of emotional intelligence.
In contrast to self-reported EI measures, performance-based
ability EI showed uniformly weaker correlations with personality
(ˆAgreeableness ⫽.29, ˆConscientiousness ⫽.13, ˆEmotionalStability ⫽.20,
ˆExtraversion ⫽.18, ˆOpenness ⫽.21) and a stronger relationship with
cognitive ability (ˆ⫽.25). Finally, we note that mixed-based EI
measures showed a considerably stronger relationship with job
performance (ˆ⫽.47) in comparison to self-report ability EI (ˆ⫽
.23) and performance-based EI (ˆ⫽.18).
The Cascading Model of EI
The meta-matrix used to test the fit of the cascading model of EI
is presented in Table 3. Results testing the fit of the cascading
model of EI are presented in Figure 2. The fit statistics for the full
model with nine degrees of freedom are root mean square error of
approximation (RMSEA) ⫽.071, comparative fit index (CFI) ⫽
.97, Tucker–Lewis index (TLI; nonnormed fit index) ⫽.92, and
standardized root mean square residual (SRMR) ⫽.040. All of the
fit statistics were judged to be in acceptable ranges (Hu & Bentler,
1999). All hypothesized structural paths were statistically signifi-
cant and in the expected directions (Hypothesis 1: ⫽.08, p⬍
.05; Hypothesis 3: ⫽.28, p⬍.05; Hypothesis 4: ⫽.35, p⬍
.05; Hypothesis 5: ⫽.12, p⬍.05), except for the direct path
from Emotional Stability to job performance (⫽⫺.00, ns). As a
test of Hypothesis 2, we estimated the direct (unmediated) path
from emotion perception to emotion regulation (Perc,Re g⫽.10;
p⬍.05). This result suggests that the relationship between emo-
tion perception and emotion regulation is not completely mediated
(James et al., 2006; contrary to Hypothesis 2). On the other hand,
this direct effect is much smaller in magnitude than the indirect
effect (Perc,UnderUnder,Re g⫽共.43兲共.53兲⫽.23), and the impact of
specifying the direct (unmediated) effect has little relative impact
on overall model fit (⌬CFI ⫽.008). In other words, much of the
relationship between emotion perception and emotion regulation is
mediated by emotion understanding.
Table 4
Meta-Analytic Correlation Matrix of Operational Validities (Corrected for Criterion Unreliability and Range Restriction Only)
123 4 56789
1. Performance-based ability EI —
2. Self-report ability EI .12 —
3. Self-report mixed EI .23 .52 —
4. Conscientiousness .12 .32 .33 —
5. Emotional Stability .17 .34 .45 .19
a
—
6. Agreeableness .25 .26 .36 .19
a
.18
a
—
7. Extraversion .15 .27 .40 .00
a
.14
a
.12
a
—
8. Openness .18 .24 .26 ⫺.04
a
.12
a
.08
a
.12
a
—
9. Cognitive ability .22 .00 .09 ⫺.03
b
.08
b
.00
b
.02
b
.18
b
—
10. Job performance .17 .22 .42 .19
c
.10
c
.07
c
.09
c
.06
c
.40
d
High emotional labor .23 .26 .47 .19 .10 .07 .09 .09 .33
Low emotional labor .01 .19 .37 .23 .14 .10 .09 .04 .35
Note. Correlations with job performance are corrected for attenuation in the criterion and range restriction. All other correlations are observed correlations.
a
Ones (1993).
b
Judge et al. (2007).
c
Hurtz & Donovan (2000; updated through 2008).
d
Hunter & Hunter (1984).
65
EMOTIONAL INTELLIGENCE: AN INTEGRATIVE REVIEW
Discriminant, Convergent, and Incremental Validity of
EI Construct–Method Pairings
To examine convergent validity, we assessed the correlation be-
tween two measures of the same construct (ability EI) and found the
correlation to be quite weak (ˆ
Self-Report Ability.Performance-Based Ability
⫽
.12; see Table 3). This suggests that self-report ability EI measures
and performance-based ability EI measures are not assessing a
singular ability EI construct, as they are purported to (failing to
support Hypothesis 6a). When discriminant validity was assessed,
the correlation between performance-based measures of ability EI
and self-reported mixed EI was low enough to suggest that ability
EI and mixed EI are in fact distinct constructs (discriminant
validity; ˆ
Self-Report Mixed.Performance-Based Ability
⫽.26; supporting
Hypothesis 6b). It is interesting to note that the two self-report
measures were correlated at ˆ
Self-Report Mixed.Self-Report Ability
⫽.59,
signaling a strong method effect (monomethod correlation ⫽.59,
heteromethod correlation ⫽.26). Finally, an examination of the
incremental validity of these measures over each other reaffirmed
the lack of convergent validity between the two ability EI mea-
sures: that is, self-reported ability EI predicted job performance
beyond performance-based ability EI measures (⌬R2⫽4.4%, p⬍
.05), and performance-based ability EI also showed incremental
validity over self-reported ability EI (⌬R2⫽2.4%, p⬍.05).
Regarding mixed EI, neither self-report ability EI nor perfor-
mance-based ability EI measures predicted job performance after
controlling for self-report mixed EI measures. The fact that self-
reports of ability EI offered almost no incremental validity over
self-report mixed EI (⌬R2⫽0.3%, p⬍.05) supports the inter-
pretation that self-reported ability EI greatly reflects the self-report
method factor and does a poor job of reflecting the ability EI trait
factor. Performance-based ability measures also offered little
unique variance in predicting job performance over self-reported
mixed EI (⌬R2⫽0.4%, p⬍.05). In contrast, self-report mixed EI
showed substantial incremental validity over self-report ability EI
(⌬R2⫽17.1%, p⬍.05), performance-based ability EI (⌬R2⫽
19.2%, p⬍.05), and a combination of self-report ability EI with
performance-based ability EI (⌬R2⫽15.1%, p⬍.05). This
incremental validity is likely attributable to the nonability (moti-
vational) content of mixed EI measures (Mayer et al., 2000) that is
not captured by ability EI measures.
In sum, although these results show evidence for the discrimi-
nant validity of ability EI from mixed EI, there is weak evidence
for similarity among ability EI measures (convergent validity) and
weak evidence for incremental validity of self-report ability EI.
These poor construct validity results for self-reported ability EI
affirm our decision to omit this concept from the integrated theo-
retical model (see the cascading model in Figure 1). Our further
decision to omit self-reported mixed EI from the theoretical model
is based on the fact that mixed EI is a muddled construct—an
ill-defined composite of ability, personality, affect, and possibly
other poorly specified content (Murphy, 2006).
Incremental Validity of EI Over Big Five Personality
and Cognitive Ability
The meta-matrix used to test the incremental validity of EI is
presented in Table 4, and incremental validity results are presented
in Tables 5a to 5c. Regarding the incremental validity of EI
construct–method pairings over and above Big Five personality
traits, all three construct–method pairings demonstrated significant
(greater than zero) incremental validity (⌬RPerformance⫺Based EI
2⫽
1.5%, ⌬RSelf⫺report Ability EI
2⫽1.7%, ⌬RSelf⫺report Mixed EI
2⫽15.7%; ps⬍
.05) and thus provided support for Hypothesis 7. Similarly, all
three construct–method pairings showed some incremental validity
over and above cognitive ability (⌬RPerformance⫺Based EI
2⫽0.7%,
⌬RSelf⫺report Ability EI
2⫽4.8%, ⌬RSelf⫺report Mixed EI
2⫽14.9%; ps⬍.05).
These results defied our expectation that ability EI would offer no
incremental prediction above cognitive ability but were consistent
with our expectation that mixed EI would exhibit significant
incremental validity over Big Five personality and cognitive abil-
ity (supporting Hypothesis 8), due to mixed EI’s inclusion of
.09*
.26*
-.04
.00
.12*
.44*
.35*
.08*
.53* .43*
.22*
.28*
Job
Performance
Emotion
Regulation
Emotion
Understanding
Emotion
Perception
Emotional Stabilit
y
Co
g
nitive Abilit
y
Conscientiousness
Figure 2. Cascading model of emotion intelligence (EI; parameter estimates). The above model was tested on
data from performance-based EI measures (RMSEA ⫽.071, CFI ⫽.97, TLI ⫽.92, SRMR ⫽.040, harmonic
mean N⫽2,285). RMSEA ⫽root mean square error of approximation; CFI ⫽comparative fit index; TLI ⫽
Tucker–Lewis index; SRMR ⫽standardized root mean square residual.
66 JOSEPH AND NEWMAN
surplus motivational constructs and other sundry content that
might be performance relevant (see Bar-On, 1997).
Finally, only two construct–method pairings exhibited incre-
mental validity over and above both Big Five personality traits and
cognitive ability (⌬RSelf⫺report Ability EI
2⫽2.3%, ⌬RSelf⫺report Mixed EI
2⫽
14.2%; ps⬍.05), suggesting that performance-based EI measures
are redundant with Big Five personality traits and cognitive ability
when predicting overall job performance, on average across all
jobs (⌬RPerformance⫺Based EI
2⫽0.2%). It is important to note that
although many of the incremental validity analyses produced sta-
tistically significant ⌬R2, this does not warrant the conclusion of
practical significance. For example, performance-based ability EI
contributes an additional 0.7% of variance in job performance
above cognitive ability ( p⬍.05; statistically significant). Whether
we evaluate this effect size as practically significant depends on
the given context and the likely utility of this tiny increment in R2
due to EI.
Subgroup Differences and Emotional Intelligence
Table 6 presents results on female–male and Black–White sub-
group differences in emotional intelligence. Regarding sex differ-
ences in EI, we find differences favoring women for performance-
based EI tests (d
uncorrected
⫽0.47), supporting the common
assumption that women have higher EI scores than do men (sup-
porting Hypothesis 9). With a sex-based subgroup difference dof
this magnitude, adverse impact against the lower scoring group
(against men) is mathematically very likely (Newman, Jacobs, &
Bartram, 2007, p. 1404; Equal Opportunity Employment Commis-
sion, Civil Service Commission, U.S. Department of Labor, &
U.S. Department of Justice, 1978). In contrast, we find no average
sex-related differences for self-report ability EI and mixed mea-
sures of EI (d
uncorrected
⫽0.01 for both self-report measures) and
therefore no systematic potential for adverse impact against either
sex.
Table 5
Incremental Validity for Emotional Intelligence Construct–Method Pairings Over Big Five Personality
Variable
Models
P I II III
Overall High EL Low EL Overall High EL Low EL Overall High EL Low EL Overall High EL Low EL
Big Five personality
Agreeableness .01 .01 .04 ⫺.01 ⫺.03 .05 ⫺.01 ⫺.01 .02 ⴚ.10 ⴚ.12 ⫺.05
Conscientiousness .18 .18 .21 .17 .17 .21 .14 .13 .19 .05 ⫺.03 .11
Emotional Stability .05 .04 .08 .03 .03 .09 .01 .00 .06 ⴚ.12 ⴚ.14 ⫺.04
Extraversion .08 .07 .07 .06 .05 .08 .04 .03 .05 ⴚ.09 ⴚ.12 ⫺.05
Openness .05 .08 .03 .03 .05 .04 .02 .04 .01 ⫺.05 ⫺.03 ⫺.04
Emotional intelligence
Performance-based .13 .19 ⫺.06
Self-report ability .16 .20 .09
Self-report mixed .54 .62 .40
R
2
.050 .054 .070 .065 .087 .074 .067 .083 .076 .207 .263 .157
Adjusted R
2
.045 .036 .051 .059 .062 .044 .060 .072 .065 .203 .246 .136
Change in R
2
.015 .033 .004 .017 .029 .006 .157 .209 .087
Note. Bold values are significant ( p⬍.05). Standardized regression coefficients. Model P ⫽personality; Overall ⫽all jobs; High EL ⫽high emotional
labor jobs; Low EL ⫽low emotional labor jobs.
Table 5b
Incremental Validity for Emotional Intelligence Construct–Method Pairings Over Cognitive Ability
Variable
Models
IV V VI
Overall High EL Low EL Overall High EL Low EL Overall High EL Low EL
Cognitive ability .38 .29 .37 .40 .33 .35 .37 .29 .32
Emotional intelligence
Performance-based .09 .17 ⫺.07
Self-report ability .22 .26 .19
Self-report mixed .39 .44 .34
R
2
.167 .135 .127 .208 .177 .159 .309 .304 .238
Adjusted R
2
.165 .127 .118 .207 .173 .155 .307 .299 .232
Change in R
2
.007 .024 .004 .048 .068 .036 .149 .162 .115
Note. Bold values are significant ( p⬍.05). Standardized regression coefficients. Overall ⫽all jobs; High EL ⫽high emotional labor jobs; Low EL ⫽
low emotional labor jobs.
67
EMOTIONAL INTELLIGENCE: AN INTEGRATIVE REVIEW
Turning to our findings regarding race and emotional intelli-
gence, we first note a substantial lack of available data (see Table
6). Therefore, our race results should be interpreted with caution
and should serve as a call for future EI researchers to report race
differences in their measures. With this in mind, we find that
performance-based ability EI tests show the largest subgroup dif-
ferences favoring Whites (d
uncorrected
⫽⫺0.99), followed by self-
report mixed EI measures (d
uncorrected
⫽⫺0.22). In contrast, self-
report ability EI measures showed subgroup differences favoring
Blacks (d
uncorrected
⫽0.31). Comparing these results to subgroup
differences on cognitive ability, current estimates (within medium
complexity jobs) suggest that Whites’ average scores on cognitive
ability measures are .72 standard deviations above Blacks’ average
scores (Roth, Bevier, Bobko, Switzer, & Tyler, 2001). Because
Black–White subgroup dvalues for self-report ability EI and
self-report mixed EI measures appear to be smaller than those
for cognitive ability tests, a composite of these EI measures
with cognitive tests has the potential to reduce adverse impact
against Blacks. In contrast, performance-based measures may
actually increase the potential for adverse impact against Blacks
when used in combination with a cognitive ability test. In sum,
we have shown that race- and sex-based subgroup differences
vary substantially according to which type of measure one uses
to assess EI, with the most construct-valid (i.e., performance-
based) EI measures showing the largest sex- and race-based
differences.
Discussion
The current study sought to clarify the theoretical basis of
emotional intelligence and to address proponents’ claims that EI is
highly predictive of job performance above well-established con-
structs (Goleman, 1995). We extended the influential work of Van
Rooy, Viswesvaran, and colleagues (Van Rooy & Viswesvaran,
Table 5c
Incremental Validity for Emotional Intelligence Construct–Method Pairings Over Big Five Personality and Cognitive Ability
Variable
Models
P&C VII VIII IX
Overall High EL Low EL Overall High EL Low EL Overall High EL Low EL Overall High EL Low EL
Big Five personality
Agreeableness .02 .02 .04 .01 ⫺.01 .07 .00 ⫺.01 .03 ⴚ.09 ⴚ.11 ⫺.04
Conscientiousness .19 .19 .22 .19 .18 .23 .15 .13 .19 .07 .05 .13
Emotional Stability .02 .02 .06 .01 .01 .07 ⫺.02 ⫺.03 .03 ⴚ.14 ⴚ.16 ⫺.06
Extraversion .08 .08 .07 .07 .06 .09 .04 .03 .05 ⴚ.08 ⴚ.11 ⫺.04
Openness ⫺.02 .03 ⫺.03 ⫺.02 .01 ⫺.02 ⫺.06 ⫺.02 ⫺.06 ⴚ.11 ⫺.08 ⫺.10
Cognitive ability .41 .33 .36 .40 .30 .38 .42 .34 .36 .39 .31 .34
Emotional intelligence
Performance-based .05 .13 ⴚ.14
Self-report ability .18 .23 .11
Self-report mixed .51 .60 .38
R
2
.209 .157 .193 .211 .172 .210 .232 .194 .202 .351 .353 .271
Adjusted R
2
.204 .138 .173 .204 .145 .181 .226 .183 .191 .347 .336 .249
Change in R
2
.002 .015 .017 .023 .037 .009 .142 .196 .078
Note. Bold values are significant ( p⬍.05). Standardized regression coefficients. Model P&C ⫽personality and cognitive ability; Overall ⫽all jobs;
High EL ⫽high emotional labor jobs; Low EL ⫽low emotional labor jobs.
Table 6
Meta-Analytic Estimates of Female–Male and Black–White Subgroup d
Measure
Female–male subgroup dBlack–White subgroup d
kN% female d95% CI d
corrrected
kN% Black d95% CI d
corrrected
Overall EI 47 16,383 55 .07 [.004, .14] .08 7 1,991 38 ⫺.19 [⫺.40, .02] ⫺.17
Performance-based EI 14 2,216 54 .47 [.24, .72] .52 1 131 42 ⫺.99 — ⫺1.06
Emotion perception 8 1,065 55 .49 [.34, .64] .53 1 136 43 ⫺.71 — ⫺.75
Emotion understanding 6 861 56 .29 [.11, .47] .31 1 135 43 ⫺.93 — ⫺.99
Emotion facilitation 9 1,280 51 .38 [.25, .51] .41 1 135 43 ⫺1.06 — ⫺1.14
Emotion regulation 9 1,190 51 .43 [.31, .56] .47 1 132 43 ⫺.79 — ⫺.84
Self-report ability EI 20 5,542 56 .01 [⫺.06, .06] .01 2 305 20 .31 — .33
Self-report mixed EI 19 8,942 54 .01 [⫺.06, .08] .02 4 1,555 42 ⫺.22 [⫺.32, ⫺.12] ⫺.26
Note. Positive dvalues mean females scored higher and Blacks scored higher, respectively. EI ⫽emotional intelligence; k⫽number of effect sizes in
the meta-analysis; N⫽total sample size in the meta-analysis; % female/Black ⫽percent of total sample size that was reported as female or Black; d⫽
sample-size weighted mean standardized difference; 95% CI ⫽lower/upper bound of confidence interval; d
corrrected
⫽standardized mean difference
corrected for attenuation.
68 JOSEPH AND NEWMAN
2004; Van Rooy, Viswesvaran, & Pluta., 2005) by distinguishing
the subfacets of EI (emotion regulation, emotion perception, and
emotion understanding); testing an integrated theoretical model for
how these subfacets relate to cognitive ability, Conscientiousness,
Neuroticism, and job performance; using narrower definitions of
job performance (excluding student performance); distinguishing
self-reported ability-based EI from self-reported mixed-based EI;
and including an updated database with many newer studies. By
answering calls for a theoretical elaboration of EI (Murphy, 2006;
Zeidner et al., 2004), we provide a theoretically driven model of EI
and job performance, labeled the cascading model of EI (see
Figure 1). This model of EI specifies a sequential (causal chain)
relationship among the three subdimensions of EI and job perfor-
mance and includes personality traits and cognitive ability as
important antecedents of the EI processes. Meta-analytic data
showed good fit with the cascading model, although the emotion
regulation–job performance connection turned out to be inconsis-
tent (as we discuss below). This result is supported by a history of
emotion regulation literature that points to differences in the extent
to which certain emotion regulation processes are effective (Broth-
eridge & Grandey, 2002; Brotheridge & Lee, 2002; Coˆte´, 2005;
Diefendorff & Gosserand, 2003; Goldberg & Grandey, 2007;
Grandey, 2003; Grandey et al., 2005).
Finally, the current study contributed to EI research and person-
nel selection practice by evaluating the potential for EI measures to
incrementally predict job performance. What has come to be
known as the fadification of EI (Murphy & Sideman, 2006) has
been driven by claims related to the importance of EI in predicting
job performance. For example, Goleman has proposed, “for star
performance in all jobs, in every field, emotional competence is
twice as important as purely cognitive abilities” (1998, p. 34). Our
incremental validity results based on meta-analytic data show that
this is not true. At best, only mixed models of EI show substantial
incremental validity over cognitive ability and Big 5 personality
traits. At worst, measures of ability models of EI show only a
modicum of incremental validity over cognitive ability and per-
sonality traits, again providing evidence against Goleman’s (1998)
expansive claims.
Emotional Intelligence Facets
The current study addressed long-standing questions about the
overlaps of EI with cognitive ability, Neuroticism, Conscientious-
ness, and job performance; showing that these relationships are
critically dependent upon which facet of EI is under consideration.
By focusing on facet-level emotional intelligence, we showed that
cognitive ability plays a role in emotional understanding, Consci-
entiousness is involved in emotion perception, and Emotional
Stability is a basis for emotion regulation. We also highlighted the
key conceptual role of emotion regulation in connecting EI to job
performance (see Figure 1). This specification helps to tie the
faddish EI practitioner literature into a long-standing, theory-laden
tradition of psychological research on emotion regulation and
self-regulation (Eisenberg, 2000; Gross, 1998a; Izard, 1991;
Muraven & Baumeister, 2000).
Emotional Labor as a Moderator
On the topic of emotion regulation at work, results of the current
meta-analysis have created some serious questions regarding the
ubiquity of emotional intelligence as a precursor to job perfor-
mance (see Table 1; cf. Goleman, 1995). For example, although
the path coefficient connecting emotion regulation to job perfor-
mance is statistically significant (⫽.08; see Figure 2), the small
magnitude of this parameter questions the practical significance of
this relationship. We further note the presence of situational mod-
erators of the validity of emotion regulation, as suggested by the
wide credibility interval for the regulation–job performance rela-
tionship (see Table 1; 80% CI includes zero).
To investigate potential situational moderation for emotional
intelligence effects on performance, we conducted a post hoc
analysis of the relationship between EI and job performance by
splitting the EI–performance primary studies into two, theoret-
ically distinct subpopulations based on emotional labor require-
ments of the job. Emotional labor theory suggests that a job’s
demands for emotional labor, or “the process of regulating both
feelings and expressions for organizational goals” (Grandey,
2000, p. 97), may serve as a moderator of the relationship
between emotional intelligence and performance (Grandey,
2000; Wong & Law, 2002). For example, occupations in which
there is frequent customer/interpersonal interaction (i.e., high
emotional labor) require more emotion regulation. Emotion
regulation demands can drain resources from task performance
(Baumeister et al., 1998; Beal et al., 2005), unless the employee
possesses heightened ability to effectively regulate emotion.
Thus, we would expect individuals with high emotion regula-
tion ability to perform especially well in jobs that require high
emotional labor. On the other hand, occupations in which there
is infrequent interpersonal interaction rely less on the ability to
regulate emotions, and we would expect the relationship be-
tween emotion regulation ability and job performance to be
lower in these jobs.
In our classification, we asked nine PhD students in industrial/
organizational psychology to each rate the emotional labor de-
mands for 191 job titles taken from (a) each primary study in-
cluded in our original meta-analyses concerning EI and job
performance, (b) all the primary studies in our updated [through
2008] meta-analyses of Hurtz and Donovan’s (2000) Big Five
personality validities, and (c) the primary studies included in
Hunter and Hunter’s (1984) synthesis of research on cognitive
ability validity (in all, 476 of the 515 primary studies could be
located and provided enough information about job titles to enable
emotional labor coding, with 191 distinct job titles represented).
We limited our moderator analyses to those primary studies in-
volving correlations with job performance, because we had no
theoretical reason to expect emotional labor to moderate the rela-
tionships among predictors (i.e., between EI and personality or
between EI and cognitive ability). Each rater coded all 191 job
titles by answering each of the following four items about emo-
tional labor (items are adapted from p. 92 of Grandey’s, 2003,
three-item measure of emotional display rules, with the addition of
one item, and are based on Hochschild’s, 1983, criteria for emo-
tional labor jobs): “Workers are expected to express positive
emotion as part of this job,” “Part of this job is to make the
customer feel good,” “Part of the product to customers is friendly,
cheerful service,” and “Displaying positive emotion is an impor-
tant part of job performance” (␣⫽.98 for 4-item scale). Each item
was coded as “yes” or “no.” We note that Grandey’s (2003) scale
focuses on positive display requirements of the job (although it
69
EMOTIONAL INTELLIGENCE: AN INTEGRATIVE REVIEW
should be mentioned that emotional labor requirements could also
logically include negative displays of emotion). Interrater reliabil-
ity (Cronbach’s ␣across the nine raters) was .94, and rater agree-
ment was rWG共J兲⫽.97.
We next averaged across items and calculated the mean of
raters’ emotional labor scores for each job title. The distribution of
these mean emotional labor ratings was bimodal and showed a
natural breaking point at .5, making this an ideal cutoff for high
versus low emotional labor jobs. A large number of jobs fit into
the low emotional labor category (i.e., 141 of the 191 job titles
coded were below the .5 scale midpoint, and 50 of the 191 jobs
were at or above the midpoint). Examples of low emotional
labor jobs include military policeman, cigarette factory worker,
Air Force mechanic, and research and development scientist;
examples of high emotional labor jobs are retail salesperson,
real estate agent, call center employee, and residential coun-
selor. When a single primary study involved a mixed sample of
jobs, we included the mixed sample in our moderator analyses
only if (a) the study reported the types of jobs involved in the
mixed sample and (b) the mean ratings of emotional labor for
every job in the sample fell above or below the cutoff of .5.
Meta-analytic correlations within the high and low emotional
labor groups were calculated (see Tables 1, 2, 3, and 4), and
incremental validity of EI over personality and cognitive ability
was assessed within high and low emotional labor jobs sepa-
rately (see Table 5).
Results of our post hoc analyses (see Tables 1, 2, 3, 4, and 5)
support emotional labor theory (Hochschild, 1983; Grandey,
2000), with high emotional labor jobs showing a stronger emotion
regulation–performance relationship (ˆ⫽.26) than low emotional
labor jobs (ˆ⫽.01). It is important to note that these results are
based on small sample sizes (N⫽220 and N⫽223, respectively)
and should be interpreted accordingly. A similar pattern exists
within other dimensions of EI, where stronger relationships with
job performance are found in high emotional labor jobs
共ˆEmotionPerception ⫽.21, ˆEmotionUnders tan ding ⫽.22, ˆEmotionFacilitation ⫽
.12) than low emotional labor jobs (ˆEmotionPerception ⫽.01,
ˆEmotionUnders tan ding ⫽.02, ˆEmotionFacilitation ⫽⫺.04). This pattern also
exists across construct–method pairings of EI (high emotional
labor: ˆPerformance⫺based ⫽.24, ˆSelf⫺reportAbility ⫽.28, ˆSelf⫺reportMixed ⫽
.59; low emotional labor: ˆPerformance⫺based ⫽.01, ˆSelf⫺reportAbility ⫽
.20, ˆSelf⫺reportMixed ⫽.43). Our tests of incremental validity (see
Table 5) within strictly high emotional labor jobs also demon-
strated the value of EI in these contexts. For example, mixed EI
measures showed more incremental validity over personality and
cognitive ability in high emotional labor jobs (⌬RSelf⫺report Mixed EI
2⫽
19.6%, p⬍.05) than in low emotional labor jobs
共⌬RSelf⫺report Mixed EI
2⫽7.8%, p⬍.05), as did self-report ability EI
measures (high emotional labor: ⌬RSelf⫺reportAbilityEI
2⫽3.7%, p⬍
.05; low emotional labor: ⌬RSelf⫺reportAbilityEI
2⫽0.9%, p⬍.05), and
performance-based measures of EI (high emotional labor:
⌬RPerformance⫺basedEI
2⫽1.5%, p⬍.05; low emotional labor:
⌬RPerformance⫺basedEI
2⫽1.7%, p⬍.05—we note that the semipartial
correlation of performance-based EI with job performance was
negative for low emotional labor jobs see Table 5).
We also reestimated the cascading model of EI using the high
emotional labor and low emotional labor correlations separately
(analyses are reported in Figure 3). The cascading model showed
adequate empirical fit for both high and low emotional labor jobs
(RMSEA ⫽.067, .078; CFI ⫽.97, .95), although the theoretical
model does appear to fit slightly better for high emotional labor
jobs. The path coefficient connecting emotion regulation to job
performance was stronger for high emotional labor jobs (⫽
.17) than for low emotional labor jobs (⫽⫺.11). In sum, the
correlational and incremental validity results point to emotional
labor as an important moderator of the EI–performance rela-
tionship, across all EI construct–method pairings and in the
cascading model. Again, it should be mentioned that the emo-
tional labor moderator results are based on small sample sizes
for the emotional labor subgroups (Ns ranging from 220 to 516
per subgroup) and should therefore serve as preliminary evi-
dence for emotional labor as a moderator of the EI–performance
relationship. Future work in this area is certainly needed, given
these results.
.26*, .26*
.09*, .09*
-.04, -.04
-.01, .06*
.12*, .12*
.35*, .41*
.35*, .35*
.17*, -.11*
.53*, .53* .43*, .43*
.19*, .28*
.28*, .28*
Job
Performance
Emotion
Regulation
Emotion
Understanding
Emotion
Perception
Emotional Stabilit
y
Co
g
nitive Abilit
y
Conscientiousness
Figure 3. Emotional labor (EL) as a moderator in the cascading model of emotional intelligence. High EL
parameters presented first, followed by low EL parameters. The above model was tested on data from
performance-based emotional intelligence (EI) measures (RMSEA ⫽.067, .078; CFI ⫽.97, .95; TLI ⫽.93, .89;
SRMR ⫽.044, .042; harmonic mean N⫽1,201, 1,213). RMSEA ⫽root mean square error of approximation;
CFI ⫽comparative fit index; TLI ⫽Tucker–Lewis index; SRMR ⫽standardized root mean square residual.
70 JOSEPH AND NEWMAN
The Two Emotional Intelligences: Ability-Based
Versus Mixed-Based
As mentioned earlier, there are two senses in which the term
emotional intelligence has been used: (a) as a narrow, theoretically
specified set of constructs pertaining to the recognition and control
of personal emotion (i.e., ability-based EI) and (b) as an umbrella
term for a broad array of constructs that are connected only by
their nonredundancy with cognitive intelligence (i.e., mixed-based
EI). In the current study, we showed that EI measures derived
using the first sense of the EI term (as a set of ability concepts
rooted in social and personality psychology) are more theoretically
grounded but suffer nongeneralizable criterion validity (i.e., emo-
tional competence predicts performance only for high emotional
labor jobs) and show substantial sex- and race-based subgroup
differences. Measures derived under the second sense of EI (as an
umbrella; mixed EI) show robust empirical evidence of criterion
validity and smaller sex- and race-based subgroup differences,
albeit with questionable theoretical value.
Correlations among EI construct–method pairings (mixed, per-
formance-ability, and self-report ability) do not suggest these
measures are reflecting the same construct. Surprisingly, the low-
est correlation between construct–method pairings existed between
self-report ability and performance-based ability EI measures (ˆ⫽
.12), which are designed to measure the same construct (ability
EI). Potential explanations for this weak relationship are discussed
below.
Results appear to support critics’ claims that mixed EI exhibits
significant overlap with Big Five personality traits (Daus & Ash-
kanasy, 2003). Upon examining the items of a popular mixed-
based measure of EI, the EQ-i (Bar-On, 1997), we see that several
EQ-I items deal directly with nervousness and anxiety. It therefore
comes as no surprise that mixed EI shows a moderate relationship
with Emotional Stability. Self-reports of mixed EI were also
shown to have a weak relationship with cognitive ability, consis-
tent with the conceptualization of mixed EI models as measuring
a combination of intellect, personality, motivation, and affect.
Meanwhile, performance-based ability measures and self-report
ability measures of EI show similar patterns of relationships with
Big 5 personality traits but different relationships with cognitive
ability (ˆ
Performance-based
⫽.28, ˆ
Self-report Ability
⫽.00). A closer
look at items on self-report measures of ability EI casts doubt on
the extent to which an actual ability is being measured. For
example, one of the 16 items on the WLEIS (Wong & Law, 2002;
Emotion Facilitation facet) is “I always set goals for myself and
then try my best to achieve them.” This item—which is similar to
three other WLEIS items—appears to address motivation rather
than ability. We propose, due to items like this on self-reports of
ability EI, that self-reports of ability EI are similar to mixed-based
measures of EI in that research has yet to confirm exactly what set
of constructs are being measured with these scales (see Joseph &
Newman, in press). Notably, a self-report of ability may also be
susceptible to socially desirable responding (Paulhus, 1984), and
self-reports of ability have been criticized for the inherent paradox
in asking individuals to report their own level of intelligence
(Matthews, Emo, Roberts, & Zeidner, 2006). The only construct–
method pairing examined in the current study that appears to be an
appropriate use of the label emotional intelligence is the perfor-
mance-based EI model, which is theoretically based in emotion
and emotion regulation literature and has a relationship with gen-
eral cognitive ability, as the name intelligence implies.
Limitations and Directions for Future Research
The current study tested a theoretically driven model of EI and
job performance. Unfortunately, the sample sizes for some of our
key effects are smaller than we would have liked. Surprisingly,
although EI measures appear ubiquitous in practice, there is a
paucity of research on EI that uses actual job performance (e.g.,
supervisor ratings of job performance) as a criterion. Many poten-
tial primary studies speaking to the relationship of EI with job
performance were not included because they chose to use proxies
of job performance (e.g., academic performance, self-reported job
performance, teacher ratings of students acting as employees).
Although a PsycINFO keyword search of emotional intelligence
returned over 900 peer-reviewed journal articles, only 22 of these
studies measured real job performance. There remains a pressing
need for future work regarding the relationship between EI and
actual performance on the job.
A second recommendation to EI researchers involves measuring
and reporting sex-based and race-based subgroup differences on
EI. Our results show that although a substantial amount of evi-
dence exists regarding sex differences on overall EI, very little
empirical work has reported on sex differences in EI dimensions.
The evidence for race differences in EI is even more sparse. Only
one primary study has investigated race differences in perfor-
mance-based EI measures, and because the evidence points to
substantial differences favoring Whites on these measures (which
could produce adverse impact), the issue deserves additional at-
tention. Although there are more estimates of race differences on
self-report measures of EI than on performance-based EI mea-
sures, the number of effect sizes is not substantial and also war-
rants additional attention. In sum, researchers investigating EI
should measure and report both sex-based and race-based differ-
ences on overall EI and EI dimensions.
Third, although EI researchers have begun to examine what
self-reported EI isn’t (e.g., personality and cognitive ability), the
EI literature has yet to investigate exactly what self-reported EI is.
Though the results of the current meta-analysis show the incre-
mental validity of self-report measures of mixed EI over person-
ality and cognitive ability, the lack of conceptual clarity in mixed
EI models leads us to caution against their application in organi-
zations. That is, without further investigation of the constructs
that mixed EI measures assess, selection practitioners will be
relying upon inventories with limited construct validity (Landy,
1986). Self-report measures of ability EI suffer from problems
similar to those of mixed measures, in that not all of the items
on these measures appear to assess a true intelligence. There-
fore, we caution against applying the current results to organi-
zational settings without knowledge of the constructs assessed
by these self-report measures. Future work on the construct-
related validity of both self-report mixed and self-report ability
measures is necessary for the advancement of self-reported EI
as a viable construct (or set of constructs; for exemplary work
of this sort, see Law et al., 2004). This would involve an
in-depth examination of the relationships between these mea-
sures and established constructs other than personality and
cognitive ability, such as self-efficacy and achievement moti-
71
EMOTIONAL INTELLIGENCE: AN INTEGRATIVE REVIEW
vation (Mayer et al., 2008). Although we have established that
the mixed model of EI is not redundant with personality traits
or cognitive ability, its redundancy with other well-established
constructs merits investigation.
Another limitation of the current study is our lack of longitudi-
nal data on which to base the cascading model. That is, despite our
supportive test of a mediation relationship from emotion percep-
tion to emotion understanding to emotion regulation, we have not
actually observed the unfolding of this process in situ. To further
confirm our cross-sectional test of the theoretical model, we esti-
mated all possible alternative sequential orderings among the three
EI facets and found that all orderings other than the one shown in
Figure 1 exhibited clearly inferior empirical fit (⌬共⌬df ⫽0兲
2⬎39.0,
change in RMSEA ⬎.012). Nonetheless, we have provided only
a snapshot of the dynamic process.
Conclusion and Implications
These results have serious implications for the large number of
organizational practitioners who use EI measures in training and
selection. We recommend that practitioners use caution when
choosing a measure of EI. Our results show large differences in
predictive validity and subgroup differences between types of EI
measures, and although mixed EI measures appear to offer the
strongest predictive power, we warn against their use due to their
unknown content and theoretical value. A summary of these im-
plications is presented in Table 7.
In sum, the EI concept has been developed and understood in
two, distinct ways: (a) as a set of specific competencies for
recognizing and controlling individual emotions and (b) as a grab
bag of constructs that contribute to job performance but are not
redundant with cognitive ability. In the current study, we the-
oretically and empirically contrasted these two perspectives.
The former viewpoint (called the ability model of EI) was
specified with strong theoretical propositions drawing upon
decades of research in social and personality psychology, but it
exhibited criterion validity only in localized settings (i.e., it did
not predict job performance across all types of jobs). The latter
viewpoint (called the mixed model of EI) showed greater prom-
ise for generalizable prediction of job performance, but it
suffers extreme theoretical underdevelopment. As such, the
current status of emotional intelligence research presents the
scientist–practitioner with a trade-off between theory and da-
ta—an ugly state of affairs. More research should assess (a) the
relationship between EI and actual job performance, (b) sex and
race differences on EI, and (c) the overlap of self-report ability
and self-report mixed EI with many long-established constructs
other than Big Five personality and cognitive ability. Indeed,
one of the eventual advantages of the mixed EI notion may be
to shed greater light on noncognitive constructs (other than Big
Five personality) that predict job performance. In the current
paper, we have attempted to integrate the existing research
within a theoretical framework in hopes of directing future
investigations on emotional competence in the workplace.
References
References marked with an asterisk indicate studies included in one or
more of the current meta-analyses.
Abe, J. A. (2003). Shame, guilt, and personality judgment. Journal of
Research in Personality, 38, 85–104.
Ackerman, P. L. (1996). A correlational analysis of skill specificity:
Learning, abilities, and individual differences. Journal of Experimental
Psychology: Learning, Memory, and Cognition, 16, 883–901.
*Alfolabi, O. A., & Ehigie, B. O. (2005). Psychological diversity and team
interaction processes: A study of oil-drilling work teams in Nigeria.
Team Performance Management, 11, 280–301.
*Amelang, M., & Steinmayr, R. (2006). Is there a validity increment for
tests of emotional intelligence in explaining the variance of performance
criteria? Intelligence, 34, 459–468.
Arthur, W., Jr., & Villado, A. J. (2008). The importance of distinguishing
between constructs and methods when comparing predictors in person-
nel selection research and practice. Journal of Applied Psychology, 93,
435–442.
*Austin, E. J., Farrelly, D., Black, C., & Moore, H. (2007). Emotional
intelligence, Machiavellianism, and emotional manipulation: Does EI
have a dark side? Personality and Individual Differences, 43, 179–189.
*Austin, E. J., Saklofske, D. H., & Egan, V. (2005). Personality, well-being
and health correlates of trait emotional intelligence. Personality and
Individual Differences, 38, 547–558.
*Austin, E. J., Saklofske, D. H., Huang, S. H. S., & McKenney, D. (2004).
Measurement of trait emotional intelligence: Testing and cross-
validating a modified version of Schutte et al.’s (1998) measure. Per-
sonality and Individual Differences, 36, 555–562.
*Barchard, K. A. (2001). Emotional and social intelligence: Examining its
place in the nomological network. Unpublished doctoral dissertation,
University of British Columbia, Vancouver, British Columbia, Canada.
*Barchard, K. A., & Hakstian, A. R. (2004). The nature and measurement
of emotional intelligence abilities: Basic dimensions and their relation-
ships with other cognitive ability and personality variables. Educational
and Psychological Measurement, 64, 437–462.
Table 7
Practical Advice For Using Emotional Intelligence Measures In
Personnel Selection
1. Choose your EI measure carefully.
There are two, distinct definitions of the term “Emotional
Intelligence”: (a) ability to perform emotional tasks, and (b) a grab-
bag of everything that is not cognitive ability. It is critical to
distinguish these two, because measures based on the two EI
definitions do not have the same content, predictive validity, or
subgroup differences.
2. Exercise extreme caution when using mixed EI measures.
Grab-bag measures of EI (i.e., self-report mixed measures) appear to
exhibit some incremental validity over cognitive ability and
personality measures on average (based on nine studies), but it is
not clear why. As such, use of these measures for personnel
decisions may be difficult to defend, without extensive local
validation.
3. Know that ability EI measures may add little to the selection system.
Ability-based measures of EI (performance-based and self-report)
exhibit little incremental validity over cognitive ability and
personality, on average.
4. Base the decision to use an EI measure on the job type (i.e., consider
the emotional labor content of the job).
When dealing with high emotional labor jobs (jobs that require
positive emotional displays), all types of EI measures exhibit
meaningful validity and incremental validity over cognitive ability
and personality. In contrast, for low emotional labor jobs, EI
validities are weaker or can even be negative.
5. Be aware of subgroup differences on EI.
Although more data are needed, preliminary evidence suggests that
performance-based EI measures favor women and Whites, which
may produce adverse impact against men and African Americans.
72 JOSEPH AND NEWMAN
Bar-On, R. (1997). Bar-On Emotional Quotient Inventory: Technical man-
ual. Toronto, Canada: Multihealth Systems.
Bar-On, R. (2000). Emotional and social intelligence: Insights from the
Emotional Quotient Inventory (EQ-I). In R. Bar-On & J. D. Parker
(Eds.), Handbook of emotional intelligence (pp. 363–388). San Fran-
cisco, CA: Jossey-Bass.
Baron-Cohen, S. (2002). The extreme male brain theory of autism. Trends
in Cognitive Science, 6, 248–254.
Barrick, M. R., Mitchell, T. R., & Stewart, G. L. (2003). Situational and
motivational influences on trait–behavior relationships. In M. R. Barrick
& A. M. Ryan (Eds.), Personality and work: Reconsidering the role of
personality in organizations (pp. 60–82). San Francisco, CA: Jossey-
Bass.
Barrick, M. R., & Mount, M. K. (2000). Select on Conscientiousness and
Emotional Stability. In E. A. Locke (Ed.), Handbook of principles of
organizational behavior (pp. 15–28). Oxford, England: Blackwell.
Barrick, M. R., Mount, M. K., & Judge, T. A. (2001). Personality and
performance and the beginning of the new millennium: What do we
know and where do we go next? International Journal of Selection and
Assessment, 9, 9–30.
*Bastian, V. A., Burns, N. R., & Nettlebeck, T. (2005). Emotional intel-
ligence predicts life skills, but not as well as personality and cognitive
abilities. Personality and Individual Differences, 39, 1135–1145.
Baumeister, R. F. (2002). Yielding to temptation: Self-control failure,
impulsive purchasing, and consumer behavior. Journal of Consumer
Research, 28, 670–676.
Baumeister, R. F., Bratslavsky, E., Muraven, M., & Tice, D. M. (1998).
Ego depletion: Is the active self a limited resource? Journal of Person-
ality and Social Psychology, 74, 1252–1265.
Beal, D. J., Trougakos, J., Weiss, H. M., & Green, S. (2006). Episodic
processes in emotional labor: Perceptions of affective delivery and
regulation strategies. Journal of Applied Psychology, 91, 1053–1065.
Beal, D. J., Weiss, H. M., Barros, E., & MacDermid, S. M. (2005). An
episodic process model of affective influences on performance. Journal
of Applied Psychology, 90, 1054–1068.
*Beshears, R. S. (2004). The ability of emotional intelligence to predict
transformational leadership when personality, affect, and cognitive abil-
ity are controlled. Unpublished doctoral dissertation, Wayne State Uni-
versity.
*Boncher, M. K. (2003). The relationship between attachment styles and
emotional intelligence. Unpublished doctoral dissertation, New York
University.
*Boyce, D. A. (2001). The correlation of emotional intelligence, academic
success, and cognitive ability in master’s level physical therapy students.
Unpublished doctoral dissertation, Spalding University.
*Brackett, M. A., & Mayer, J. D. (2003). Convergent, discriminant, and
incremental validity of competing measures of emotional intelligence.
Personality and Social Psychology Bulletin, 29, 1147–1158.
*Brackett, M. A., Mayer, J. D., & Warner, R. M. (2004). Emotional
intelligence and its relation to everyday behavior. Personality and Indi-
vidual Differences, 36, 1387–1402.
*Brackett, M. A., Rivers, S. E., Shiffman, S., Lerner, N., & Salovey, P.
(2006). Relating emotional abilities to social functioning: A comparison
of self-report and performance measures of emotional intelligence. Jour-
nal of Personality and Social Psychology, 91, 780–795.
Brotheridge, C., & Grandey, A. (2002). Emotional labor and burnout:
Comparing two perspectives of “people work.” Journal of Vocational
Behavior, 60, 17–39.
Brotheridge, C., & Lee, R. T. (2002). Testing a conservation of resources
model of the dynamics of emotional labor. Journal of Occupational
Health Psychology, 7, 57–67.
*Bryant, D. (2005). The components of emotional intelligence and the
relationship to sales performance. Unpublished doctoral dissertation,
George Washington University.
Butler, E. A., Egloff, B., Wilhelm, F. H., Smith, N. C., Erickson, E. A., &
Gross, J. J. (2003). The social consequences of expressive suppression.
Emotion, 3, 48–67.
*Byrne, J. C., Dominick, P. G., Smither, J. W., & Reilly, R. R. (2007).
Examination of the discriminant, convergent, and criterion-related va-
lidity of self-ratings on the Emotional Competence Inventory. Interna-
tional Journal of Selection and Assessment, 15, 341–353.
Coˆte´, S. (2005). A social interaction model of the effects of emotion
regulation on work strain. Academy of Management Review, 30, 509–
530.
*Coˆte´, S., & Miners, C. T. H. (2006). Emotional intelligence, cognitive
intelligence, and job performance. Administrative Science Quarterly, 51,
1–28.
Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant
validation by the multitrait–multimethod matrix. Psychological Bulletin,
56, 81–105.
*Carmeli, A., & Josman, Z. E. (2006). The relationship among emotional
intelligence, task performance, and organizational citizenship behaviors.
Human Performance, 19, 403–419.
*Carrothers, R. M., Gregory, S. W., & Gallagher, T. J. (2000). Measuring
emotional intelligence of medical school applicants. Academic Medi-
cine, 75, 456–463.
Cattell, R. B. (1957). Personality and motivation structure and measure-
ment. New York, NY: World Book.
*Cavins, B. J. (2005). The relationship between emotional–social intelli-
gence and leadership practices among college student leaders. Unpub-
lished doctoral dissertation, Bowling Green State University.
*Chamorro-Premuzic, T., Bennett, E., & Furnham, A. (2007). The happy
personality: Mediational role of trait emotional intelligence. Personality
and Individual Differences, 42, 1633–1639.
*Chapman, B. P. (2005). Emotional intelligence at mid life: A cross
sectional investigation of structural variance, social correlates, and
relationship to established personality and ability taxonomies. Unpub-
lished doctoral dissertation, University of North Texas.
*Ciarrochi, J., Chan, A. Y. C., & Caputi, P. (2000). A critical evaluation of
the emotional intelligence construct. Personality and Individual Differ-
ences, 28, 539–561.
Ciarrochi, J., Hynes, K., & Crittenden, N. (2005). Can men do better if they
try harder? Sex and motivational effects on emotional awareness. Cog-
nition and Emotion, 19, 133–141.
*Click, H. S. (2002). An exploration of emotional intelligence scores
among students in educational administration endorsement programs.
Unpublished doctoral dissertation, East Tennessee State University.
*Cobeˆro, C., Primi, R., & Muniz, M. (2006). Inteligeˆncia emocional e
desempenho no trabalho: Um estudo com MSCEIT, BPR-5 e 16PF
[Emotional intelligence and performance at work: A study of MSCEIT,
BPR-5 and 16PF]. Paideia, 16, 337–348.
Cole, P. M., Martin, S. E., & Dennis, T. A. (2004). Emotion regulation as
a scientific construct: Methodological challenges and directions for child
development research. Child Development, 75, 317–333.
Connor-Smith, J. K., & Flachsbart, C. (2007). Relations between person-
ality and coping: A meta-analysis. Journal of Personality and Social
Psychology, 6, 1080–1107.
Conte, J. M. (2005). A review and critique of emotional intelligence
measures. Journal of Organizational Behavior, 26, 433–440.
Conte, J. M., & Dean, M. A. (2006). Can emotional intelligence be
measured? In K. R. Murphy (Ed.), A critique of emotional intelligence
(pp. 59–80). Mahwah, NJ: Erlbaum.
Cronbach, L. J. (1960). Essentials of psychological testing (2nd ed.). New
York, NY: Harper & Row.
Daus, C. S. (2006). The case for an ability-based model. In K. R. Murphy
(Ed.), A critique of emotional intelligence (pp. 301–324). Mahwah, NJ:
Erlbaum.
Daus, C. S., & Ashkanasy, N. M. (2003). Will the real emotional intelli-
73
EMOTIONAL INTELLIGENCE: AN INTEGRATIVE REVIEW
gence please stand up? On deconstructing the emotional intelligence
“debate.” The Industrial–Organizational Psychologist, 41, 69–72.
Daus, C. S., & Ashkanasy, N. M. (2005). The case for an ability-based
model of emotional intelligence in organizational behavior. Journal of
Organizational Behavior, 26, 453–466.
*Davies, M. D., Stankov, L., & Roberts, R. D. (1998). Emotional intelli-
gence: In search of an elusive construct. Journal of Personality and
Social Psychology, 75, 989–1015.
*Dawda, D., & Hart, S. D. (2000). Assessing emotional intelligence:
Reliability and validity of the Bar-On Emotional Quotient Inventory
(EQ-i) in university students. Personality and Individual Differences, 28,
797–812.
*Day, A. L., & Carroll, S. A. (2004). Using an ability-based measure of
emotional intelligence to predict individual performance, group perfor-
mance, and group citizenship behaviors. Personality and Individual
Differences, 36, 1443–1458.
*Day, A. L., Therrien, D. L., & Carroll, S. A. (2005). Predicting psycho-
logical health: Assessing the incremental validity of emotional intelli-
gence beyond personality, type A behaviour and daily hassles. European
Journal of Personality, 19, 519–536.
*Derksen, J., Kramer, I., & Katzo, M. (2002). Does a self-report measure
for emotional intelligence assess something different than general intel-
ligence? Personality and Individual Differences, 32, 37–48.
Diefendorff, J. M., & Gosserand, R. (2003). Understanding the emotional
labor process: A control theory perspective. Journal of Organizational
Behavior, 24, 945–959.
*Douglas, C., Frink, D. D., & Ferris, G. R. (2003). Emotional intelligence
as a moderator of the relationship between conscientiousness and per-
formance. Journal of Leadership and Organizational Studies, 10, 2–13.
*Dulewicz, V., Higgs, M., & Slaski, M. (2003). Measuring emotional
intelligence: Content, construct and criterion-related validity. Journal of
Managerial Psychology, 5, 405–420.
*Dunn, E. W., Brackett, M. A., Ashton-James, C., Schneiderman, E., &
Salovey, P. (2007). On emotionally intelligent time travel: Individual
differences in affective forecasting ability. Personality and Social Psy-
chology Bulletin, 33, 85–93.
Einstein, D., & Lanning, K. (1998). Shame, guilt, ego development, and the
five-factor model of personality. Journal of Personality, 66, 555–582.
Eisenberg, N. (2000). Emotion, regulation, and moral development. Annual
Review of Psychology, 51, 665–697.
Eisenberg, N., Smith, C. L., Sadovsky, A., & Spinrad, T. L. (2004).
Effortful control: Relations with emotion regulation, adjustment, and
socialization in childhood. In R. F. Baumeister & K. D. Vohs (Eds.),
Handbook of self-regulation: Research, theory, and applications (pp.
259–282). New York, NY: Guilford Press.
Eisenberger, R., Armeli, S., Rexwinkel, B., Lynch, P. D., & Rhoades, L.
(2001). Reciprocation of perceived organizational support. Journal of
Applied Psychology, 86, 42–51.
Ekman, P., & O’Sullivan, M. (1991). Who can catch a liar? American
Psychologist, 48, 384–392.
Elfenbein, H. A. (2008). Emotion in organizations: A review and theoret-
ical integration. In J. P. Walsh & A. P. Brief (Eds.), Academy of
Management annals (Vol. 1, pp. 315–386). New York, NY: Erlbaum.
Emmons, R. A. (1989). The personal strivings approach to personality. In
L. A. Pervin (Ed.), Goal concepts in personality and social psychology
(pp. 87–117). Mahwah, NJ: Erlbaum.
Equal Opportunity Employment Commission, Civil Service Commission,
U.S. Department of Labor, & U.S. Department of Justice. (1978).
Uniform guidelines on employee selection procedures. Federal Register,
43, 38290–38309.
Erez, A., & Isen, A. M. (2002). The influence of positive affect on the
components of expectancy motivation. Journal of Applied Psychology,
87, 1055–1067.
*Extremera, N., & Fernandez-Berrocal, P. (2005). Perceived emotional
intelligence and life satisfaction: Predictive and incremental validity
using the Trait Meta-Mood Scale. Personality and Individual Differ-
ences, 39, 937–948.
Eysenck, H. J. (1970). The structure of human personality. London, En-
gland: University Press.
*Farrelly, D., & Austin, E. J. (2007). Ability EI as an intelligence?
Associations of the MSCEIT with performance on emotion processing
and social tasks and with cognitive ability. Cognition and Emotion, 21,
1043–1063.
*Fillion, F. (2001). The construct validation of two measures of emotional
intelligence. Unpublished master’s thesis, University of Guelph, Guelph,
Ontario, Canada.
Fine, S. E., Izard, C. E., & Trentacosta, C. J. (2006). Emotion situation
knowledge in elementary school: Models of longitudinal growth and
preschool correlates. Social Development, 15, 730–751.
Fineman, S. (2004). Getting the measure of emotion—and the cautionary
tale of emotional intelligence. Human Relations, 57, 719–740.
*Fox, S., & Spector, P. E. (2000). Relations of emotional intelligence,
practical intelligence, general intelligence, and trait affectivity with
interview outcomes: It’s not all just “G.” Journal of Organizational
Behavior, 21, 203–220.
Fredrickson, B. L. (2001). The role of positive emotions in positive
psychology: The broaden-and-build theory of positive emotions. Amer-
ican Psychologist, 56, 218–226.
*Freudenthaler, H. H., & Neubauer, A. C. (2005). Emotional intelligence:
The convergent and discriminant validities of intra- and interpersonal
emotional abilities. Personality and Individual Differences, 39, 569–
579.
Frijda, N. H. (1986). The emotions. Cambridge, England: Cambridge
University Press.
*Fukunishi, I., Wise, T. N., Sheridan, M., Shimai, S., Orake, K., Utsuki, N.,
& Uchiyama, K. (2001). Validity and reliability of the Japanese version
of the Emotional Intelligence Scale among college students and psychi-
atric outpatients. Psychological Reports, 89, 625–632.
*Furnham, A. (2005). Self-estimated intelligence, psychometric intelli-
gence and personality. Psychologia, 48, 182–192.
Gailliot, M. T., Mead, N. L., & Baumeister, R. F. (2008). Self-regulation.
In O. P. John, R. W. Robins, & L. A. Pervin (Eds.), Handbook of
personality (3rd ed., pp. 472–491). New York, NY: Guilford Press.
*Gannon, N., & Ranzijn, R. (2005). Does emotional intelligence predict
unique variance and life satisfaction beyond IQ and personality? Per-
sonality and Individual Differences, 38, 1353–1364.
George, J. M. (1991). State or trait: Effects of positive mood on prosocial
behaviors at work. Journal of Applied Psychology, 76, 299–307.
Gignac, G. E. (2005). Evaluating the MSCEIT V2.0 via CFA: Comment on
Mayer et al. (2003). Emotion, 5, 233–235.
Goldberg, L. S., & Grandey, A. A. (2007). Display rules versus display
autonomy: Emotion regulation, emotional exhaustion, and task perfor-
mance in a call center simulation. Journal of Occupational Health
Psychology, 12, 301–318.
*Goldenberg, I., Matheson, K., & Mantler, J. (2006). The assessment of
emotional intelligence: A comparison of performance-based and self-
report methodologies. Journal of Personality Assessment, 86, 33–45.
*Goldsmith, T. B. (2008). Relationships between emotional intelligence
and individual workplace performance. Unpublished doctoral disserta-
tion, Lynn University.
Goleman, D. (1995). Emotional intelligence: Why it can matter more than
IQ. New York, NY: Bantam Books.
Goleman, D. (1998). Working with emotional intelligence. New York, NY:
Bantam Books.
Grandey, A. A. (2000). Emotion regulation in the workplace: A new way
to conceptualize emotional labor. Journal of Occupational Health Psy-
chology, 5, 95–110.
Grandey, A. (2003). When “the show must go on”: Surface and deep acting
74 JOSEPH AND NEWMAN
as predictors of emotional exhaustion and service delivery. Academy of
Management Journal, 46, 86–96.
Grandey, A., Dickter, D., & Sin, H.-P. (2004). The customer is not always
right: Customer verbal aggression toward service employees. Journal of
Organizational Behavior, 25, 397–418.
Grandey, A. A., Fisk, G. M., & Steiner, D. D. (2005). Must “service with
a smile” be stressful? The moderating role of personal control for
American and French employees. Journal of Applied Psychology, 90,
893–904.
Gross, J. J. (1998a). Antecedent- and response-focused emotion regulation:
Divergent consequences for experience, expression, and physiology.
Journal of Personality and Social Psychology, 74, 224–237.
Gross, J. J. (1998b). The emerging field of emotion regulation: An inte-
grative review. Review of General Psychology, 2, 271–299.
Gross, J. J. (2008). Emotion and emotion regulation: Personality processes
and individual differences. In O. P. John, R. W. Robins, & L. A. Pervin
(Eds.), Handbook of personality (3rd ed., pp. 701–724). New York, NY:
Guilford Press.
Gross, J. J., & John, O. (2003). Individual differences in two emotion
regulation processes: Implications for affect, relationships, and well-
being. Journal of Personality and Social Psychology, 85, 348–362.
Gross, J. J., Sutton, S. K., & Ketelaar, T. (1998). Relations between affect
and personality: Supper for the affect-level and affective-reactivity
views. Personality and Social Psychology Bulletin, 24, 279–288.
Gross, J. J., & Thompson, R. A. (2007). Emotion regulation: Conceptual
foundations. In J. J. Gross (Ed.), Handbook of emotion regulation (pp.
3–24). New York, NY: Guilford Press.
*Grubb, W. L., & McDaniel, M. A. (2007). The fakability of Bar-On’s
Emotional Quotient Inventory Short Form: Catch me if you can. Human
Performance, 20, 43–59.
*Hader, E. M. (2006). Emotional intelligence and its relationship to
cognitive and social task requirements. Unpublished doctoral disserta-
tion, Alliant International University.
Hall, J. A. (1978). Gender effects in decoding nonverbal cues. Psycholog-
ical Bulletin, 85, 845–857.
Hall, J. A. (1984). Nonverbal sex differences: Communication accuracy
and expressive style. Baltimore, MD: Johns Hopkins University Press.
Hirt, E. R., Melton, R. J., McDonald, H. E., & Harackiewicz, J. M. (1996).
Processing goals, task interest, and the mood–performance relationship:
A mediation analysis. Journal of Personality and Social Psychology, 71,
245–261.
Hochschild, A. R. (1983). The managed heart: Commercialization of
human feeling. Berkeley: University of California Press.
Hogan, J. C., & Holland, B. (2003). Using theory to evaluate personality
and job performance relations: A socioanalytic perspective. Journal of
Applied Psychology, 88, 100–112.
Holbrook, W. L. (1997). A study of the relationships between emotional
intelligence and basic writers’ skills. Unpublished doctoral dissertation,
Ball State University.
*Hong, Y. (2005). Motivation to lead: Antecedents resulting in leader
emergence. Unpublished doctoral dissertation, Saint Mary’s University.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance
structure analysis: Conventional criteria versus new alternatives. Struc-
tural Equation Modeling, 6, 1–55.
Humphreys, L. G. (1979). The construct of general intelligence. Intelli-
gence, 3, 105–120.
Hunter, J. E., & Hunter, R. F. (1984). Validity and utility of alternative
predictors of job performance. Psychological Bulletin, 96, 72–98.
Hunter, J. E., & Schmidt, F. L. (2004). Methods of meta-analysis. Thou-
sand Oaks, CA: Sage.
Hunter, J. E., Schmidt, F. L., & Le, H. (2006). Implications of direct and
indirect range restriction for meta-analysis methods and findings. Jour-
nal of Applied Psychology, 91, 594–612.
Hurtz, G. M., & Donovan, J. J. (2000). Personality and job performance:
The Big Five revisited. Journal of Applied Psychology, 85, 869–879.
*Ivcevic, Z., Brackett, M. A., & Mayer, J. D. (2007). Emotional intelli-
gence and emotional creativity. Journal of Personality, 75, 199–236.
Izard, C. E. (1991). The psychology of human emotions. New York, NY:
Plenum Press.
James, L. R., Mulaik, S. A., & Brett, J. M. (2006). A tale of two methods.
Organizational Research Methods, 9, 233–244.
*Janovics, J., & Christiansen, N. D. (2002, January). Emotional intelli-
gence in the workplace. Paper presented at the annual conference of the
Society of Industrial and Organizational Psychology, San Diego, CA.
Jensen, A. R. (1994). Humphreys’ “behavioral repertoire”: An epiphenom-
enon of g. Psychological Inquiry, 5, 208–210.
Jensen-Campbell, L. A., & Malcolm, K. T. (2007). The importance of
conscientiousness in adolescent interpersonal relationships. Personality
and Social Psychology Bulletin, 33, 368–383.
*Johnson, H. M., & Spector, P. E. (2007). Service with a smile: Do
emotional intelligence, gender, and autonomy moderate the emotional
labor process? Journal of Occupational Health Psychology, 12, 319–
333.
*Jordan, P. J., Ashkanasy, N. M., Hartel, C. E. J., & Hooper, G. S. (2002).
Workgroup emotional intelligence: Scale development and relationship
to team process effectiveness and goal focus. Human Resource Man-
agement Review, 12, 195–214.
Joseph, D. L., & Newman, D. A. (in press). Discriminant validity of
self-reported emotional intelligence: A multitrait–multisource study. Ed-
ucational and Psychological Measurement.
Judge, T. A., Jackson, C. L., Shaw, J. C., Scott, B. A., & Rich, B. L. (2007).
Self-efficacy and work-related performance: The integral role of indi-
vidual differences. Journal of Applied Psychology, 92, 107–127.
Judge, T. A., Thoresen, C. J., Bono, J. E., & Patton, G. K. (2001). The job
satisfaction–job performance relationship: A qualitative and quantitative
review. Psychological Bulletin, 127, 376–407.
Kahneman, D. (1973). Attention and effort. Englewood Cliffs, NJ:
Prentice-Hall.
Kanfer, R., & Ackerman, P. L. (1989). Motivation and cognitive abilities:
An integrative/aptitude–treatment approach to skill acquisition [Mono-
graph]. Journal of Applied Psychology, 74, 657–690.
Kanfer, R., & Ackerman, P. L. (1990). Ability and metacognitive determi-
nants of skill acquisition and transfer [Final report]. Washington, DC:
Air Force Office of Scientific Research.
Kanfer, R., & Ackerman, P. L. (1996). A self-regulatory skills perspective
to reducing cognitive interference. In I. G. Sarason, B. R. Sarason, &
G. R. Pierce (Eds.), Cognitive interference: Theories, methods, and
findings (pp. 153–171). Mahwah, NJ: Erlbaum.
Kanfer, R., Ackerman, P. L., & Heggestad, E. D. (1996). Motivational
skills and self-regulation for learning: A trait perspective. Learning and
Individual Differences, 8, 185–209.
*King, D. H. (1999). Measurement of differences in emotional intelligence
of preservice educational leadership students and practicing adminis-
trators as measured by the Multifactor Emotional Intelligence Scale.
Unpublished doctoral dissertation, East Carolina University.
Kintsch, W. (2000). The control of knowledge activation in discourse
comprehension. In W. J. Perrig & A. Grob (Eds.), Control of human
behavior, mental processes, and consciousness (pp. 137–146). Mahwah,
NJ: Erlbaum.
*Kluemper, D. H. (2006). An examination of ability-based emotional
intelligence in the structured employment interview. Unpublished doc-
toral dissertation, Oklahoma State University.
*Kohan, A. (2002). Emotional intelligence: An investigation of discrimi-
nant and concurrent validity. Unpublished doctoral dissertation, Lake-
head University, Thunder Bay, Ontario, Canada.
*Kostman, J. T. (2004). Multi-dimensional performance requires multi-
dimensional predictors: Predicting complex job performance using cog-
75
EMOTIONAL INTELLIGENCE: AN INTEGRATIVE REVIEW
nitive ability, personality, and emotional intelligence assessment instru-
ments as combinatorial predictors. Unpublished doctoral dissertation,
City University of New York.
*Lam, L. T. (1998). Emotional intelligence: Implications for individual
performance. Unpublished doctoral dissertation, Texas Tech University.
*Lam, L. T., & Kirby, S. L. (2002). Is emotional intelligence an advantage?
An exploration of the impact of emotional and general intelligence on
individual performance. Journal of Social Psychology, 142, 133–143.
Landy, F. J. (1986). Stamp collecting versus science: Validation as hy-
pothesis testing. American Psychologist, 41, 1183–1192.
Landy, F. J. (2005). Some historical and scientific issues related to research
on emotional intelligence. Journal of Organizational Behavior, 26,
411–424.
Landy, F. J. (2006). The long, frustrating, and fruitless search for social
intelligence: A cautionary tale. In K. R. Murphy (Ed.), A critique of
emotional intelligence (pp. 81–123). Mahwah, NJ: Erlbaum.
*Law, K. S., Wong, C., Huang, G. H., & Li, X. (2008). The effects of
emotional intelligence on job performance and life satisfaction for the
research and development scientists in China. Asia Pacific Journal of
Management, 25, 51–69.
*Law, K. S., Wong, C. S., & Song, L. J. (2004). The construct and criterion
validity of emotional intelligence and its potential utility for manage-
ment studies. Journal of Applied Psychology, 89, 483–496.
Lazarus, R. S. (1968). Emotions and adaptation. In W. J. Arnold (Ed.),
Nebraska Symposium on Motivation (Vol. 16, pp. 175–265). Lincoln:
University of Nebraska Press.
Lazarus, R. S. (1991). Emotion and adaptation. New York, NY: Oxford
University Press.
Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New
York, NY: Springer-Verlag.
*Lindley, L. D. (2001). Personality, other dispositional variables, and
human adaptability. Unpublished doctoral dissertation, Iowa State Uni-
versity.
*Livingstone, H. A., & Day, A. L. (2005). Comparing the construct- and
criterion-related validity of ability-based and mixed-model measures of
emotional intelligence. Educational and Psychological Measurement,
65, 757–779.
Locke, E. A. (2005). Why emotional intelligence is an invalid concept.
Journal of Organizational Behavior, 26, 425–431.
*Lopes, P. N., Brackett, M. A., Nezlek, J., Schutz, A., Sellin, I., & Salovey,
P. (2004). Emotional intelligence and social interaction. Personality and
Social Psychology Bulletin, 30, 1018–1034.
*Lopes, P. N., Grewal, D., Kadis, J., Gall, M., & Salovey, P. (2006).
Evidence that emotional intelligence is related to job performance and
affect and attitudes at work. Psicothema, 18, 132–138.
*Lopes, P. N., Salovey, P., & Straus, R. (2003). Emotional intelligence,
personality and the perceived quality of social relationships. Personality
and Individual Differences, 35, 641–658.
*Lumley, M. A., Gustavson, B. J., Partridge, R. T., & Labouvie-Vief, G.
(2005). Assessing alexithymia and related emotional ability constructs
using multiple methods: Interrelationships among measures. Emotion, 5,
329–342.
*MacCann, C., Roberts, R. D., Matthews, G., & Zeidner, M. (2004).
Consensus scoring and empirical option weighting of performance-
based emotional intelligence (EI) tests. Personality and Individual Dif-
ferences, 36, 645–662.
*Mack-Allen, J. (2001). Relationship between emotional intelligence and
personality factors in a community-based sample. Unpublished doctoral
dissertation, Widener University.
Marco, C. A., & Suls, J. (1993). Daily stress and the trajectory of mood:
Spillover, response assimilation, contrast, and chronic negative affectiv-
ity. Journal of Personality and Social Psychology, 64, 1053–1063.
*Martinez-Pons, M. (1997). The relation of emotional intelligence with
selected areas of personal functioning. Imagination, Cognition, and
Personality, 17, 3–13.
Matsumoto, D., LeRoux, J., Wilson-Cohn, W., Raroque, J., Kooken, K.,
Ekman, P., . . . & Goh, A. (2000). A new test to measure emotion
recognition ability: Matsumoto and Ekman’s Japanese and Caucasian
Brief Affect Recognition Test (JACBART). Journal of Nonverbal Be-
havior, 24, 179–209.
*Matthews, G., Emo, A. K., Funke, G., Zeidner, M., Roberts, R. D., Costa,
P. T., & Schulze, R. (2006). Emotional intelligence, personality, and
task-induced stress. Journal of Experimental Psychology: Applied, 12,
96–107.
Matthews, G., Emo, A. K., Roberts, R. D., & Zeidner, M. (2006). What is
this thing called emotional intelligence? In K. R. Murphy (Ed.), A
critique of emotional intelligence (pp. 3–36). Mahwah, NJ: Erlbaum.
Matthews, G., Zeidner, M., & Roberts, R. D. (2002). Emotional intelli-
gence: Science and myth. Cambridge, MA: MIT Press.
*Mayer, J. D., Caruso, D. R., & Salovey, P. (1999). Emotional intelligence
meets traditional standards for emotional intelligence. Intelligence, 27,
267–298.
Mayer, J. D., & Geher, G. (1996). Emotional intelligence and the identi-
fication of emotion. Intelligence, 22, 89–113.
Mayer, J. D., Roberts, R. D., & Barsade, S. G. (2008). Human abilities:
Emotional intelligence. Annual Review of Psychology, 59, 507–536.
Mayer, J. D., & Salovey, P. (1993). The intelligence of emotional intelli-
gence. Intelligence, 17, 433–442.
Mayer, J. D., & Salovey, P. (1997). What is emotional intelligence? In P.
Salovey & D. Sluyter (Eds.), Emotional development and emotional
intelligence: Implications for educators (pp. 3–34). New York, NY:
Basic Books.
Mayer, J. D., Salovey, P., & Caruso, D. R. (2000). Models of emotional
intelligence. In R. J. Sternberg (Ed.), Handbook of intelligence (pp.
392–420). Cambridge, England: Cambridge University Press.
Mayer, J. D., Salovey, P., Caruso, D. R., & Sitarenios, G. (2003). Modeling
and measuring emotional intelligence with the MSCEIT V2.0. Emotion,
3, 97–105.
McClure, E. B. (2000). A meta-analytic review of sex differences in facial
expression processing and their development in infants, children, and
adolescents. Psychological Bulletin, 126, 424–453.
McCrae, R. R. (2000). Emotional intelligence from the perspective of the
five-factor model of personality. In R. Bar-On & J. D. A. Parker (Eds.),
The handbook of emotional intelligence: Theory, development, assess-
ment, and application at home, school, and in the workplace (pp.
263–276). San Francisco, CA: Jossey-Bass.
McCrae, R. R., & Costa, P. T., Jr. (1992). Discriminant validity of NOE-
PIR facet scales. Educational and Psychological Measurement, 52,
229–237.
*Mestre, J. M., Guil, R., Lopes, P. N., Salovey, P., & Gil-Olarte, P. (2006).
Psicothema, 16, 112–117.
*Moss, M., Ritossa, D., & Ngu, D. (2006). The effect of follower regula-
tory focus and extraversion on leadership behavior: The role of emo-
tional intelligence. Journal of Individual Differences, 27, 93–107.
*Muniz, M., & Primi, R. (2007). Inteligeˆncia emocional e desempenho
em policiais militares: Validade de criterio do MSCEIT. Aletheia, 25,
66–81.
Muraven, M., & Baumeister, R. F. (2000). Self-regulation and depletion of
limited resources: Does self-control resemble a muscle? Psychological
Bulletin, 126, 247–259.
Muraven, M., Tice, D. M., & Baumeister, R. F. (1998). Self-control as
limited resource: Regulatory depletion patterns. Journal of Personality
and Social Psychology, 74, 774–789.
*Murensky, C. L. (2000). The relationship between emotional intelligence,
personality, critical thinking ability, and organizational leadership per-
formance at upper levels of management. Unpublished doctoral disser-
tation, George Mason University.
76 JOSEPH AND NEWMAN
Murphy, K. R. (2006). Four conclusions about emotional intelligence. In
K. R. Murphy (Ed.), A critique of emotional intelligence (pp. 345–354).
Mahwah, NJ: Erlbaum.
Murphy, K. R., & Sideman, L. (2006). The two EIs. In K. R. Murphy (Ed.),
A critique of emotional intelligence (pp. 37–58). Mahwah, NJ: Erlbaum.
Newman, D. A., Jacobs, R. R., & Bartram, D. (2007). Choosing the best
method for local validity estimation: Relative accuracy of meta-analysis
versus a local study versus Bayes analysis. Journal of Applied Psychol-
ogy, 92, 1394–1413.
*Newsome, S., Day, A. L., & Catano, V. M. (2000). Assessing the
predictive validity of emotional intelligence. Personality and Individual
Differences, 29, 1005–1016.
Nigg, J. T., John, O. P., Blaskey, L. G., Huang-Pollock, C. L., Willcutt,
E. G., Hinshaw, S. P., & Pennington, B. (2002). Big Five dimensions
and ADHD symptoms: Links between personality traits and clinical
symptoms. Journal of Personality and Social Psychology, 83, 451–469.
Norman, D. A., & Bobrow, D. B. (1975). On data-limited and resource-
limited processes. Cognitive Psychology, 7, 44–64.
Norman, W. T. (1963). Toward an adequate taxonomy of personality
attributes: Replicated factor structure in peer nomination personality
ratings. Journal of Abnormal Social Psychology, 66, 574–583.
Nunnally, J. C. (1978). Psychometric theory. New York, NY: McGraw-
Hill.
*O’Connor, R. M., Jr., & Little, I. S. (2003). Revisiting the predictive
validity of emotional intelligence: Self-report versus ability-based mea-
sures. Personality and Individual Differences, 35, 1893–1902.
Olson-Buchanan, J. B., Drasgow, F., Moberg, P. J., Mead, A. D., Keenan,
P. A., & Donovan, M. A. (1998). Interactive video assessment of conflict
resolution skills. Personnel Psychology, 51, 1–24.
Ones, D. S. (1993). The construct validity of integrity tests. Unpublished
doctoral dissertation, University of Iowa.
Palmer, B. R., Gignac, G., Manocha, R., & Stough, C. (2005). A psycho-
metric evaluation of the Mayer–Salovey–Caruso Emotional Intelligence
Test Version 2.0. Intelligence, 33, 285–305.
*Pasanen, S. E. (2000). Emotional intelligence, conscientiousness and
integrity as predictors of organizational citizenship behavior. Unpub-
lished master’s thesis, University of California, Los Angeles.
Paulhus, D. L. (1984). Two-component models of socially desirable re-
sponding. Journal of Personality and Social Psychology, 46, 598–609.
*Pellitteri, J. (2002). The relationship between emotional intelligence and
ego defense mechanisms. Journal of Psychology, 136, 182–194.
*Petrides, K. V., & Furnham, A. (2001). Trait emotional intelligence:
Psychometric investigation with reference to established trait taxono-
mies. European Journal of Personality, 15, 425–448.
*Petrides, K. V., & Furnham, A. (2006). The role of trait emotional
intelligence in a gender-specific model of organizational variables. Jour-
nal of Applied Social Psychology, 36, 552–569.
*Petrides, K. V., Niven, L., & Mouskounti, T. (2006). The trait emotional
intelligence of ballet dancers and musicians. Psicothema, 18, 101–107.
*Petrides, K. V., Pita, R., & Kokkinaki, F. (2007). The location of trait
emotional intelligence in personality factor space. British Journal of
Psychology, 98, 273–289.
*Rego, A., Sousa, G., Cunha, M. P., Correia, A., & Sair-Amaral, I. (2007).
Leader self-reported emotional intelligence and perceived employee
creativity: An exploratory study. Creativity and Innovation Manage-
ment, 16, 250–264.
Roberts, B. W., Jackson, J. J., Fayard, J. V., Edmonds, G., & Meints, J. (in
press). Conscientiousness. In M. Leary & R. Hoyle (Eds.), Handbook of
individual differences in social behavior. New York, NY: Guilford
Press.
*Roberts, R. D., Zeidner, M., & Matthews, G. (2001). Does emotional
intelligence meet traditional standards for an intelligence? Some new
data and conclusions. Emotion, 1, 196–231.
*Rode, J. C., Mooney, C. H., Arthaud-Day, M. L., Near, J. P., Baldwin,
T. T., Rubin, R. S., & Bommer, W. H. (2007). Emotional intelligence
and individual performance: Evidence of direct and moderated effects.
Journal of Organizational Behavior, 28, 399–421.
Roseman, I. J. (1979, September). Cognitive aspects of emotion and
emotional behavior. Paper presented at the meeting of the American
Psychological Association, New York.
Roseman, I. J. (1984). Cognitive determinants of emotions: A structural
theory. In P. Shaver (Ed.), Review of personality and social psychology:
Vol. 5. Emotions, relationships, and health (pp. 11–36). Beverley Hills,
CA: Sage.
*Rosete, D., & Ciarrochi, J. (2005). Emotional intelligence and its rela-
tionship to workplace performance outcomes of leadership effectiveness.
Leadership and Organization Development Journal, 26, 388–399.
Rossen, E., Kranzler, J. H., & Algina, J. (2008). Confirmatory factor
analysis of the Mayer–Salovey–Caruso Emotional Intelligence Test V
2.0 (MSCEIT). Personality and Individual Differences, 44, 1258–1269.
Roth, P. L., BeVier, C., Bobko, P., Switzer, F., & Tyler, P. (2001). Ethnic
group differences in cognitive ability in employment and educational
settings: A meta-analysis. Personnel Psychology, 54, 297–330.
*Rovnak, A. (2007). A psychometric investigation of the Emotional Quo-
tient Inventory in adolescents: A construct validation and estimate of
stability. Unpublished doctoral dissertation, University of Akron.
Sackett, P. R., & Yang, H. (2000). Correction for range restriction: An
expanded typology. Journal of Applied Psychology, 85, 112–118.
*Saklofske, D. H., Austin, E. J., Galloway, J., & Davidson, K. (2007).
Individual difference correlates of health-related behaviors: Preliminary
evidence for links between emotional intelligence and coping. Person-
ality and Individual Differences, 42, 491–502.
*Saklofske, D. H., Austin, E. J., & Minski, P. S. (2003). Factor structure
and validity of trait emotional intelligence measure. Personality and
Individual Differences, 34, 707–721.
Salgado, J. F. (1997). The five-factor model of personality and job perfor-
mance in the European community. Journal of Applied Psychology, 82,
30–43.
Salovey, P., & Mayer, J. D. (1990). Emotional intelligence. Imagination,
Cognition, and Personality, 9, 185–211.
Schmidt, F. L., & Hunter, J. (2004). General mental ability in the world of
work: Occupational attainment and job performance. Journal of Person-
ality and Social Psychology, 86, 162–173.
Schmidt, F. L., Hunter, J. E., & Outerbridge, A. N. (1986). Impact of job
experience and ability on job knowledge, work sample performance, and
supervisory ratings of job performance. Journal of Applied Psychology,
71, 432–439.
*Schmidt, L. L. (2006). Self-reported emotional intelligence as an indica-
tor of social exchange quality at work. Unpublished doctoral disserta-
tion, University of Houston.
*Schulte, M. J., Ree, M. J., & Carretta, T. R. (2004). Emotional intelli-
gence: Not much more than gand personality. Personality and Individ-
ual Differences, 37, 1059–1068.
*Schumacher, L. A. (2005). The relationship between supply managers’
emotional intelligence and their performance. Unpublished doctoral
dissertation, Bowling Green State University.
*Schutte, N. S., Malouff, J. M., Hall, L. E., Haggerty, D. J., Cooper, J. T.,
Golden, C. J., & Dornheim, L. (1998). Development and validation of a
measure of emotional intelligence. Personality and Individual Differ-
ences, 25, 167–177.
*Semadar, A., Robins, G., & Ferris, G. R. (2006). Comparing the validity
of multiple social effectiveness constructs in the prediction of manage-
rial job performance. Journal of Organizational Behavior, 27, 443–461.
Shadish, W. R. (1996). Meta-analysis and the exploration of causal medi-
ating processes: A primer of examples, methods, and issues. Psycholog-
ical Methods, 1, 154–169.
*Shaikh, A. (2004). Emotional intelligence: Is it intelligence or a person-
77
EMOTIONAL INTELLIGENCE: AN INTEGRATIVE REVIEW
ality trait? Unpublished doctoral dissertation, Laurentian University,
Sudbury, Ontario, Canada.
Shewchuk, R. M., Elliott, T. R., MacNair-Semands, R. R., & Harkins, S.
(1999). Trait influences on stress appraisal and coping: An evaluation of
alternative frameworks. Journal of Applied Social Psychology, 29, 685–
704.
*Shi, J., & Wang, L. (2007). Validation of emotional intelligence scale in
Chinese university students. Personality and Individual Differences, 43,
377–387.
*Sjoberg, L. (2001a). Emotional intelligence: A psychometric analysis.
European Psychologist, 6, 79–95.
*Sjoberg, L. (2001b). Emotional intelligence measured in a highly com-
petitive testing situation. Unpublished manuscript, Center for Economic
Psychology, Stockholm School of Economics, Stockholm, Sweden.
Smith, C. A., & Ellsworth, P. C. (1985). Patterns of cognitive appraisal in
emotion. Journal of Personality and Social Psychology, 48, 5812–5838.
*Smith, J. E. (2000). Emotional intelligence and behavior: An exploratory
study of people on parole. Unpublished doctoral dissertation, Kansas
State University.
*Stewart, J. H. (1997). Practical intelligence: Assessing its convergent and
discriminant validity with social, emotional, and academic intelligence.
Unpublished doctoral dissertation, University of Southern Mississippi.
*Sue-Chan, C., & Latham, G. P. (2004). The situational interview as a
predictor of academic and team performance: A study of the mediating
effects of cognitive ability and emotional intelligence. International
Journal of Selection and Assessment, 12, 312–320.
Suls, J., Green, P. J., & Hillis, S. (1998). Emotional reactivity to everyday
problems, affective inertia, and neuroticism. Personality and Social
Psychology Bulletin, 24, 127–136.
*Sy, T., Tram, S., & O’Hara, L. A. (2006). Relation of employee and
manager emotional intelligence to job satisfaction and performance.
Journal of Vocational Behavior, 68, 461–473.
*Tett, R. P., Fox, K. E., & Wang, A. (2005). Development and validation
of a self-report measure of emotional intelligence as a multidimensional
trait domain. Personality and Social Psychology Bulletin, 31, 859–888.
*Thomas, M. D. (2005). Response biases and the five-factor personality
construct. Unpublished master’s thesis, Mississippi State University.
*Thompson, B. L., Waltz, J., Croyle, K., & Pepper, A. C. (2007). Trait
meta-mood and affect as predictors of somatic symptoms and life
satisfaction. Personality and Individual Differences, 43, 1786–1795.
Thorndike, R. L. (1920). Intelligence and its use. Harper’s Magazine, 140,
227–235.
Thorndike, R. L. (1949). Personnel selection: Test and measurement
techniques. New York, NY: Wiley.
*Tombs, S. (2004). Challenging the bell curve: An assessment of the role
of emotional intelligence in career placement and performance. Unpub-
lished doctoral dissertation, York University, Toronto, Ontario, Canada.
Totterdell, P. (1999). Mood score: Mood and performance in professional
cricketers. British Journal of Psychology, 90, 317–332.
Totterdell, P. (2000). Catching moods and hitting runs: Mood linkage and
subjective performance in professional sport teams. Journal of Applied
Psychology, 85, 848–859.
Totterdell, P., & Holmann, D. (2003). Emotion regulation in customer
service roles: Testing a model of emotional labor. Journal of Occupa-
tional Health Psychology, 8, 55–73.
Tracy, J. L., & Robins, R. W. (2004). Putting the self in the self-conscious
emotions: A theoretical model. Psychological Inquiry, 15, 103–125.
Tsai, W. C., Chen, C. C., & Liu, H. L. (2007). Test of a model linking
employee positive moods and task performance. Journal of Applied
Psychology, 92, 1570–1583.
*Van Rooy, D. L., Alonso, A., & Viswesvaran, C. (2005). Group differ-
ences in emotional intelligence scores: Theoretical and practical impli-
cations. Personality and Individual Differences, 38, 689–700.
Van Rooy, D. L., Dilchert, S., Viswesvaran, C., & Ones, D. S. (2006).
Multiplying intelligences: Are general, emotional, and practical intelli-
gences equal? In K. R. Murphy (Ed.), A critique of emotional intelli-
gence (pp. 235–262). Mahwah, NJ: Erlbaum.
Van Rooy, D. L., & Viswesvaran, C. (2004). Emotional intelligence: A
meta-analytic investigation of predictive validity and nomological net.
Journal of Vocational Behavior, 65, 71–95.
Van Rooy, D. L., Viswesvaran, C., & Pluta, P. (2005). An evaluation of
construct validity: What is this thing called Emotional Intelligence?
Human Performance, 18, 445–462.
*Vela, R. H., Jr. (2003). The role of emotional intelligence in the academic
achievement of first-year college students. Unpublished doctoral disser-
tation, Texas A&M University.
Viswesvaran, C., & Ones, D. S. (1995). Theory testing: Combining psy-
chometric meta-analysis with structural equation modeling. Personnel
Psychology, 48, 865–885.
Viswesvaran, C., & Ones, D. S. (2000). Measurement error in “big five
factors” personality assessment: Reliability generalization across studies
and measures. Educational and Psychological Measurement, 60, 224–
235.
Vohs, K. D., Baumeister, R. F., Schmeichel, B. J., Twenge, J. M., Nelson,
N., M., & Tice, D. M. (2008). Making choices impairs subsequent
self-control: A limited-resource account of decision making, self-
regulation, and active initiative. Journal of Personality and Social Psy-
chology, 94, 883–898.
Vohs, K. D., & Heatherton, T. F. (2000). Self-regulatory failure: A
resource-depletion approach. Psychological Science, 11, 249–254.
*Walker, M. E. B. (2006). Emotional intelligence and academic success in
college. Unpublished doctoral dissertation, University of Southern Mis-
sissippi.
*Warwick, J., & Nettelbeck, T. (2004). Emotional intelligence is . . .?
Personality and Individual Differences, 37, 1091–1100.
*Wolfradt, U., Felfe, J., & Koster, T. (2002). Self-perceived emotional
intelligence and creative personality. Imagination, Cognition and Per-
sonality, 21, 293–309.
*Wong, C., Foo, M., Wang, C., & Wong, P. (2007). The feasibility of
training and development of EI: An exploratory study in Singapore,
Hong Kong and Taiwan. Intelligence, 35, 141–150.
*Wong, C., & Law, D. S. (2002). The effects of leader and follower
emotional intelligence on performance and attitude: An exploratory
study. Leadership Quarterly, 13, 243–274.
*Wong, C. S., Wong, P. M., & Law, K. S. (2007). Evidence of the practical
utility of Wong’s emotional intelligence scale in Hong Kong and main-
land China. Asia Pacific Journal of Management, 24, 43–60.
*Wraight, M. E. (2006). The impact of trait emotional intelligence and
cognitive style on the academic achievement and life satisfaction of
college students. Unpublished doctoral dissertation, State University of
New York at Buffalo.
*Zeidner, M., & Kaluda, I. (2008). Romantic love: What’s emotional
intelligence (EI) got to do with it? Personality and Individual Differ-
ences, 44, 1684–1695.
Zeidner, M., Matthews, G., & Roberts, R. D. (2004). Emotional intelli-
gence in the workplace: A critical review. Applied Psychology: An
International Review, 53, 371–399.
Received December 2, 2008
Revision received June 1, 2009
Accepted July 29, 2009 䡲
78 JOSEPH AND NEWMAN
A preview of this full-text is provided by American Psychological Association.
Content available from Journal of Applied Psychology
This content is subject to copyright. Terms and conditions apply.
