ArticlePDF Available

The Science of Emotional Intelligence: Current Consensus and Controversies


Abstract and Figures

Almost from its inception, the emotional intelligence (EI) construct has been an elusive one. After nearly 2 decades of research, there still appears to be little consensus over how EI should be conceptualized or assessed and the efficacy of practical applications in real life settings. This paper aims at providing a snapshot of the state-of-the-art in research involving this newly minted construct. Specifically, in separate sections of this article, we set out to distinguish what is known from what is unknown in relation to three paramount concerns of EI research, i.e., conceptualization, assessment, and applications. In each section, we start by discussing assertions that may be made with some degree of confidence, elucidating what are essentially sources of consensus concerning EI. We move then to discuss sources of controversy; those things for which there is less agreement among EI researchers. We hope that this "straight talk" about the current status of EI research will provide a platform for new research in both basic and applied domains.
Content may be subject to copyright.
M. Zeidner et al.: EI: Co nsensus and ControversiesEuropean Psychologist 2008; Vol. 13(1):64–78© 2008 Hogrefe & Huber Publishers
The Science of Emotional Intelligence
Current Consensus and Controversies
Moshe Zeidner
, Richard D. Roberts
, and Gerald Matthews
University of Haifa, Israel,
Educational Testing Service, Princeton, NJ, USA,
University of Cincinnati, OH, USA
Abstract. Almost from its inception, the emotional intelligence (EI) construct has been an elusive one. After nearly 2 decades of research,
there still appears to be little consensus over how EI should be conceptualized or assessed and the efficacy of practical applications in
real life settings. This paper aims at providing a snapshot of the state-of-the-art in research involving this newly minted construct.
Specifically, in separate sections of this article, we set out to distinguish what is known from what is unknown in relation to three
paramount concerns of EI research, i.e., conceptualization, assessment, and applications. In each section, we start by discussing assertions
that may be made with some degree of confidence, elucidating what are essentially sources of consensus concerning EI. We move then
to discuss sources of controversy; those things for which there is less agreement among EI researchers. We hope that this “straight talk”
about the current status of EI research will provide a platform for new research in both basic and applied domains.
Keywords: emotional intelligence, emotions, intelligence, emotional competencies, consensus and controversy
Over the past 17 years or so, emotional intelligence (EI)
has emerged as one of the most visible and high-profile
constructs in individual differences research (Matthews,
Zeidner, & Roberts, 2002, 2007). Broadly defined, EI rep-
resents a set of competencies for identifying, processing,
and managing emotion. Research into EI has prospered, in
part, because of the increasing importance of intelligence
for people in modern society. It has also been claimed that
EI predicts clinical, educational, and occupational criteria
above and beyond that predicted by general intelligence.
However, despite a high level of public and scientific in-
terest, the science of EI is in its early years, with many
issues still being debated and cardinal questions remaining
As the field matures, it has become increasingly common-
place for academic psychologists to reject early, overstated
claims made in popular works about the importance of emo-
tional intelligence. Goleman (1995), for example, claimed EI
may be as important as IQ in determining life success (see
also e.g., Cooper & Sawaf, 1997). Such claims have mainly
been disconfirmed by the decade of rigorous empirical re-
search that has followed. Decisively, Van Rooy and Viswes-
varan’s (2004) meta-analysisshowedthat,aftercorrectingfor
statistical artifacts, the average correlation between EI and
work performance was modest (i.e., around 0.24).
Notwithstanding, research on EI goes back to entirely
respectable and sober research on social abilities and com-
petencies pursued by a line of distinguished intelligence
researchers (Landy, 2006). Contemporary researchers have
also made a persuasive case that EI measures add some-
thing new to conventional understanding of human individ-
ual differences (e.g., Mayer, Salovey, & Caruso, 2000a,b).
There are now several hundred peer-reviewed journal arti-
cles that deal with key issues, including models of EI, as-
sessment of EI, personality and cognitive ability correlates
of EI, validation of tests of EI against measures of social
functioning and adaptation, and applications in real life set-
tings (especially work, education, and clinical practice).
Although work on EI is open to various criticisms (e.g.,
Matthews et al., 2002), the field may still possess a suffi-
cient groundswell of scientific support to constitute a legit-
imate branch of psychological science.
This paper aims at providing a snapshot of the state-of-
the-art in this newly minted construct. Specifically, in sep-
arate sections of this article, we set out to distinguish what
is known from what is unknown in relation to three para-
mount concerns of EI research, i.e., conceptualization, as-
sessment, and applications. In each section, we start by dis-
cussing assertions that may be made with some degree of
confidence, elucidating what are essentially sources of con-
sensus concerning EI. We move then to discuss sources of
controversy; those things for which there is less agreement
among EI researchers. We hope that this “straight talk”
about the current status of EI research will help the reader
disentangle rhetoric from reality and fiction from fact in
contemporary EI research.
Conceptualization of Emotional
Despite nearly 2 decades of research, there appears little
consensus over how EI should be defined and conceptual-
DOI 10.1027/1016-9040.13.1.64
European Psychologist 2008; Vol. 13(1):64–78 © 2008 Hogrefe & Huber Publishers
ized (Matthews et al., 2002; Matthews, Roberts, & Zeidner,
2004; Roberts, Schulze, Zeidner, & Matthews, 2005).
Thus, it is presently unclear whether EI is cognitive or non-
cognitive; whether it refers to explicit or implicit knowl-
edge of emotion; and whether it refers to a basic aptitude
or to some adaptation to a specific social and cultural mi-
lieu (Zeidner, Matthews, & Roberts, 2001). Arguably, one
of the primary initial tasks in any scientific endeavor is
systematic mapping of the major components and facets in
the universe of discourse under consideration (Kerlinger,
1973). However, it is difficult to obtain a satisfactory jus-
tification or definitional framework for the construct of EI.
In fact, most researchers would likely agree that popular
definitions are too over-inclusive to be useful. Defining EI
as a “laundry list of virtually every positive quality of
character except for cognitive intelligence, also accom-
plishes little (cf. Goleman, 1995).
Although research definitions are not immune to over-
inclusiveness, the majority of attempts to operationalize EI
have tried to start from some conceptual analysis of the
construct. The four-branch model of Mayer et al. (2000a)
is the best-known and most influential conceptualization of
this kind. It has inspired not just the ability tests for which
Mayer et al. (2000b) are renowned, but questionnaire mea-
sures as well (see e.g., Schutte et al., 1998). Mayer et al.
(2000b) divide ability models, which view EI as a standard
intelligence, from mixed models that define EI more broad-
ly as a package of personal qualities including both abilities
and personality traits that facilitate expression of EI. A fur-
ther conceptualization, trait EI (Petrides & Furnham,
2003), embeds EI exclusively within the personality do-
main, and, thus, separates EI rather sharply from abilities
as normally defined. It is certainly legitimate to study
stable perceptions of emotional functioning as an aspect of
personality, but the label “trait EI” may be misleading, giv-
en its connotations of ability (O’Sullivan, 2007). Next, we
discuss some sources of consensus, followed by sources of
controversy, in conceptualizing the EI construct.
Conceptualization: Sources of Consensus
Emotional intelligence is a multifaceted construct that may
best be studied from multiple perspectives. A variety of dif-
ferent conceptions and definitions of EI have been pro-
posed, each leading to different operational measures of the
construct. The various psychometrically adequate scales
for EI appear to be measuring several different constructs.
This assertion is supported most clearly by the weakness
of correlations between objective tests and questionnaires
based on self-report (Brackett, Mayer, & Warner, 2004;
Warwick & Nettelbeck, 2004), but other instances may be
found. For example, Austin (2005) studied chronometric
(i.e., time-based) measures that, by analogy with cognitive
research, should correlate with scores on EI tests. Although
a distinct emotional-information processing factor was
identified, it failed to correlate significantly with overall EI
scores from two leading assessments.
There is probably a consensus that “trait” EI may be
studied separately from ability-based EI (e.g., Petrides &
Furnham, 2003). However, it remains unclear if there is a
single over-arching general “trait” EI dimension, or if the
various questionnaires may be picking up multiple, largely
independent traits. Some questionnaires do exhibit a strong
general factor (e.g., Bar-On, 2004; Petrides & Furnham,
2003), whereas others do not. Tett, Fox, and Wang (2005)
extracted three independent factors from their question-
naire, which they labeled as self-orientation, other-orienta-
Table 1. Multiple conceptualizations of emotional intelligence
Construct Possible current
Equivalent in
IQ research
Key processes Adaptive significance Developmental
Temperament Scales for Big Five
EQ-i (Bar-On, 1997)
None Neural and cognitive
processes controlling
arousal, attention and
reinforcement sensi-
Mixed: most tempera-
mental factors confer
a mixture of costs
and benefits
Genetics and early
Information process-
JACBART, Emotion-
al Stroop, RAFL
Choice RT, inspec-
tion time, working
Specific processing
Uncertain: Is speed of
processing necessari-
ly adaptive?
Genetics and early
Emotional self-regula-
Selected scales from
questionnaires for EI
(e.g., SSRI, TEIQue)
Self-assessed intelli-
Monitoring and regu-
lation of internal
states; self-efficacy in
such processes
Predominantly but
not exclusively posi-
tive: presumed simi-
lar to self-esteem
Learning and social-
ization: e.g., mastery
experiences, model-
ing, direct reinforce-
ment (in emotive con-
Emotional knowledge
and skills
MSCEIT Gc and/or Gk Multiple acquired pro-
cedural and declara-
tive skills
Adaptive within con-
text for learning: may
be irrelevant or count-
er-productive in other
Learning, socializa-
tion and training of
specific skills and
M. Zeidner et al.: EI: Consensus and Controversies 65
© 2008 Hogrefe & Huber Publishers European Psychologist 2008; Vol. 13(1):64–78
tion, and emotional-sharing. As they point out, the first two
factors correspond to Gardners (1983) intra- and inter-per-
sonal intelligences, conceived as separate personal intelli-
gences. By contrast, the Mayer et al. (2000a,b) model does
not sharply separate emotion regulation in self and others.
Thus, although it is agreed that EI may encompass multiple
domains (e.g., Austin & Saklofske, 2005), it remains un-
clear how many separate domains should be discriminated,
and whether general factors may be identified within each
domain. Elsewhere (Matthews, Zeidner, & Roberts, 2006),
we have tentatively distinguished at least four, largely sep-
arate domains, as shown in Table 1.
Emotional intelligence overlaps with other psychological
constructs. A major issue is how EI should be aligned
with other dimensions of ability and personality. The de-
gree of overlap of EI with other constructs appears to be
measure-dependent, with multiple EI constructs appear-
ing to relate differently to other factors. For example, ob-
jective measures of EI such as the Multifactor Emotional
Intelligence Scale (MEIS) and Mayer-Salovey-Caruso
Emotional Intelligence Test (MSCEIT), in keeping with
an “ability model,” correlate between 0.30 and 0.40 with
general intelligence, and rather less with personality
traits (e.g., Roberts, Zeidner, & Matthews, 2001). Per-
haps one might envisage including an EI ability factor
inside the multistratum model for ability advanced by
Carroll (1993).
By contrast, questionnaire measures for EI overlap with
standard personality traits to a degree that is often exces-
sive. The most egregious example is Bar-On’s (2004) Emo-
tional Quotient Inventory (EQ-i), which correlates around
0.80 with low trait anxiety and general psychopathology
(e.g., Bar-On, 1997; Newsome, Day, & Catano, 2000).
Most of its reliable variance may be attributed to the Five-
Factor model (FFM) of personality (Dawda & Hart, 2000;
Matthews et al., 2002). Similarly, Petrides and Furnham
(2003) reported that their trait EI questionnaire correlated
at –0.70 with neuroticism and 0.68 with extraversion. Other
questionnaires appear to possess more unique variance, but
still show substantial intercorrelation with personality (see
Matthews et al., 2002, 2004), raising the thus-far unan-
swered question of how personality and EI traits should be
interrelated within a common structural model.
Individual emotional intelligence constructs relate mean-
ingfully to various external criteria. There is a growing
body of evidence showing that various scales for EI corre-
late robustly with a variety of outcomes that plausibly sig-
nal social-emotional success (Day, 2004). Both MSCEIT
subscales and various questionnaires purportedly assessing
EI correlate with measures of well-being and social en-
gagement (e.g., Brackett, Rivers, Shiffmann, Lerner, & Sa-
lovey, 2006; Lopes, Salovey, & Straus, 2003; Saklofske,
Austin, & Minski, 2003). Validity coefficients tend to be
higher for questionnaires, but, in this case, they are ampli-
fied by confounding with personality traits linked to social
adjustment, such as extraversion. Given the overlap with
personality and ability already noted, a critical issue is
whether tests for EI show incremental predictive validity
with key factors, including general intelligence and the Big
Five personality factors, statistically controlled. Evidence
for incremental validity is, however, clearly mixed. Several
studies fail to show that scales for EI predict important cri-
teria (e.g., academic success) with other traits controlled
(Barchard, 2003), while still other studies demonstrate the
obverse (e.g., Ciarrochi, Chan, & Caputi, 2000; Saklofske
et al., 2003).
Emotional intelligence relates to exceptionality in theoret-
ically meaningful, and practically important, ways. A com-
mon tactic for gathering validity evidence for ability and
personality measures is to show that they discriminate
groups that are exceptional in terms of, for example, their
intellectual capabilities, pathology, or social maladjust-
ment. Support for the validity of EI may be found from
similar sources. Case studies of emotionally gifted individ-
uals provide some informal but persuasive evidence (Oat-
ley, 2004). Quantitative differences in EI scores have been
shown for groups such as therapists, psychiatric patients,
and prisoners (e.g., Bar-On, 2004; Schutte et al., 1988).
These studies have typically, however, neglected the pos-
sible confounding influences of personality and ability. The
association between EI and (lack of) alexithymia has been
important in establishing the clinical relevance of the con-
struct (Parker, Taylor, & Bagby, 2001). However, for each
exceptional group, findings may vary with the measure of
EI employed. For example, Zeidner, Shani-Zinovich, Mat-
thews, and Roberts (2005) showed that intellectually gifted
children obtained elevated scores on the MSCEIT, but were
lower in self-reported EI, demonstrating a failure of con-
Emotional intelligence follows a well-defined develop-
mental trajectory. There is an extensive literature on so-
cial-emotional development in children that predates
work on EI (e.g., Denham, 1998). There appears to be a
fairly well-defined sequence of markers of emotional de-
velopment, beginning with the simple expressive and
regulatory behaviors of the infant, and culminating in ac-
tive, insightful self-regulation sensitive to the social and
cultural environment (Saarni, 2000). We have proposed
that emotional competencies may be attached to the fol-
lowing three aspects of the developmental sequence: (1)
basic temperaments shaped by innate biological attri-
butes; (2) social learning of rule-based adaptive behav-
iors (e.g., emotion display rules); and (3) development of
self-reflective insight (Zeidner, Matthews, Roberts, &
MacCann, 2003).
There are empirical literatures focusing on such key is-
sues as the nature of temperament (Rothbart & Bates,
1998), the shaping of emotional competencies by parent-
child interaction (Eisenberg, Fabes, & Loyosa, 1997), and
the role of language in shaping emotional awareness in old-
66 M. Zeidner et al.: EI: Consensus and Controversies
European Psychologist 2008; Vol. 13(1):64–78 © 2008 Hogrefe & Huber Publishers
er children (Shipman & Zeaman, 1999). The work of Izard
(e.g., 2001) on emotional adaptiveness provides a clear and
convincing link between the developmental literature and
emotional intelligence.
Conceptualization: Sources of Controversy
Are researchers who investigate emotional intelligence on
the right track or following blind alleys? Various commen-
tators have questioned the value of focusing on resolving
the conceptual status of EI. Brody (2004), for example,
challenges the psychometric status and predictive validity
of ability tests of EI. By contrast, Landy (2005) claims that
much of the applied work fails to meet elementary scien-
tific standards, such as availability of data to other re-
searchers. Even if it is accepted that empirical studies are
rigorous, the suspicion may be that the phenomena ob-
served (e.g., rating one’s response to vignettes of doubtful
personal relevance) lack ecological validity and are of little
relevance to real-life emotional functioning.
Is there evidence for multiple emotional “intelligences”
and do they share any common element(s)? We have al-
ready indicated that there seems to be little overlap between
ability-based performance measures of EI and self-report
measures. However, even if we transfer questionnaire mea-
sures to the personality domain, as trait EI, significant dif-
ficulties remain. As we have pointed out in previous cri-
tiques (e.g., Matthews et al., 2002, 2004; Roberts et al.,
2005), there are a variety of distinct ways of conceptualiz-
ing EI as a construct open to objective measurement. These
include as explicit emotional knowledge (as featured in the
MSCEIT), as implicit emotional knowledge and procedural
skills, as an “ecological” construct related to person-envi-
ronment fit in specific social contexts, as components of
information-processing measured chronometrically (Aus-
tin, 2005), and other perspectives (see Table 1). It is un-
known whether constructs measured within each domain
correlate positively with one another; certainly a future re-
search topic that might profitably be explored. A possible
basis for EI in implicit knowledge or processing has been
especially neglected, despite recent interest in implicit fea-
tures of personality and unconscious priming processes.
There are additional reasons from emotion theory to
question whether an overarching EI construct is expected
to generalize across different domains and measurement
methods. One feature of many emotions theories (e.g., Oat-
ley & Johnson-Laird, 1995) is that there are multiple basic
emotions supported by distinct neurological and cognitive
systems. For example, anxiety and anger, respectively, are
supported by different fear and rage neurobiological sub-
systems (Panksepp, 1998). Thus, it is not self-evident that
individual differences in the functioning of, say, the fear
system, which is localized in structures including the
amygdala, will relate to individual differences in other sys-
tems, such as anger, or disgust. It could reasonably be ar-
gued that EI relates not to individual emotions but to a su-
per-ordinate emotion-regulation system located anatomi-
cally in the frontal lobes (cf., Rolls, 1999). However, in our
view, EI researchers have done too little to separate emo-
tion from meta-emotional regulation, which may be sup-
ported by different brain systems. Also surprising is the
paucity of research evaluating whether individual differ-
ences in regulation generalize across the basic emotions.
Clearly, these are important research questions that would
advance the field if addressed with rigorously designed
What are the most appropriate criteria for evaluating the
importance of emotional intelligence? Another fundamen-
tal difficulty confronting researchers is deciding what in-
dependent criteria for social-emotional competence should
be predicted by EI measures (Matthews et al., 2002). Ex-
acerbating this problem, most research uses self-report
scales to assess both the predictors and criteria, with crite-
rion contamination, thus, coming into play. For example,
questionnaires for EI include items that refer to positive
mood, optimism, and confidence; the very same criteria
that many researchers wish to predict. Indeed, too little re-
search has used objective behavioral criteria, though in the
few isolated instances where this information has been col-
lected, evidence tends to be mixed. For example, the
MSCEIT fails to predict attentional and working memory
performance under stress (Matthews, Emo, Funke et al.,
2006). EI questionnaires also fail to predict learning to use
emotional information in a multi-cue discrimination task
(Fellner, Perez, Emo, & Matthews, 2006). By contrast,
studies have linked questionnaire measures to cortisol se-
cretion (Salovey, Stroud, Woolery, & Epel, 2002) and
speed of facial processing (Petrides & Furnham, 2003).
What should the status of ethical and moral behavior be in
emotional intelligence models? Another issue is the extent
to which ethical and moral behavior should be part and
parcel of the EI construct. Equally important is whether it
should be differentiated from character and ethical behav-
ior, and from conformity to social norms. The selfish, Ma-
chiavellian individual may possess EI in the sense of per-
ceiving others emotional weaknesses and then proceeding
to manipulate them. At the extreme, individuals with anti-
social personality disorder may possess social skills that
allow them to exploit others (Harpur, Hart, & Hare, 2002),
despite deficiencies in other areas of emotional function-
ing. As recently noted by Waterhouse (2006, p. 253):
“nothing in any EI construct precludes someone with high
EI from being an immoral person.”
There is also a tension between EI in the sense of fitting
in with the social expectations of others and consensus
norms and EI in the sense of making insightful autonomous
decisions about the value of social norms. Some Germans
opposed the Nazis and were executed as a consequence.
Certainly, these people were ethical and even heroic; but
were they emotionally intelligent?
M. Zeidner et al.: EI: Consensus and Controversies 67
© 2008 Hogrefe & Huber Publishers European Psychologist 2008; Vol. 13(1):64–78
What are the dynamic processes underlying emotional in-
telligence? It is perhaps ironic that much EI research re-
visits one of the less attractive aspects of intelligence re-
search. This a tendency to present ability as a static set
of constructs, while ignoring the processes that support
intelligent interaction with the external environment (cf.
Corno et al., 2002). Describing EI as a list of desirable
qualities, unrelated to any independent psychological
theory, is the most obvious example of this problem.
More subtly, research on EI frequently ignores the per-
son-situation interaction that has become fundamental to
personality research (Magnusson, 1976). The basic point
is that the expression of EI may vary – perhaps radically
– depending on the surrounding environment. Individual
differences in EI may be more apparent in some contexts
than in others, and EI may be adaptive in some settings
but harmful in others (Ciarrochi, Dean, & Anderson,
2002; Petrides & Furnham, 2003).
Borrowing from Caspi and Bem’s (1990) account of
person-situation interaction, we can identify at least three
forms of interaction, related to the above, that remain al-
most entirely unexplored. How do high EI persons filter
and interpret the social world around them (reactive in-
teraction)? What kinds of behaviors does the high EI per-
son provoke in others (evocative interaction)? Perhaps
(consciously or unconsciously), the high EI person elicits
more cooperation and support from other people, where-
as the low EI person rubs them the wrong way. How does
EI relate to choosing and shaping social environments
(proactive interaction)? For example, low EI may be as-
sociated not just with a liking for harmful drugs (Trini-
dad, Unger, Chou, Azen, & Johnson, 2004), but also for
picking friends that are a bad influence.
Is emotional intelligence related to adaptation? Calling
EI an ability signals that it refers to individual differences
in adaptation. Within this perspective, the high EI person
is, in some sense, better adapted to social-emotional func-
tioning than the person of lower EI. This assumption
needs empirical support, especially as self-reports of
competence are sometimes viewed as circumspect (see
Dunning, Heath, & Suls, 2004). The vision of the pio-
neers of tests for EI appears to have been that high EI is
unequivocally adaptive (e.g., Mayer, Caruso, & Salovey,
1999). To enjoy high EI is to enjoy a variety of benefits,
including life satisfaction, emotional connections with
others, and occupational success. As we have previously
shown, researchers differ sharply in their assessments of
whether these predictions have been confirmed (e.g.,
Brody, 2004; Day, 2004; Mayer, Salovey, & Caruso,
2004). Furthermore, there appears to be a “dark side” of
EI (or some of its specific operationalizations), including
over-confidence, narcissism, Machiavellian social ma-
nipulation, and inflated self-esteem (Zeidner, Roberts, &
Matthews, in press).
Assessment of Emotional Intelligence
The case that reliable and valid assessment of EI is central
to building a science of EI is straightforward to make (Mat-
thews et al., 2002), and is generally accepted by researchers
(cf. however, Oatley, 2004). As noted above, approaches
to measurement of the EI construct have generally been
understood within Mayer et al.’s (2000b) distinction be-
tween ability and mixed models of the construct. Indeed,
reviews of the various measures of EI (e.g., McCann, Mat-
thews, Zeidner, & Roberts, 2003) have generally been
structured around this distinction. The ability model sug-
gests that EI may be assessed by objective tests, whereas
mixed models have inspired self-report approaches. Mayer
et al. (2000b) introduce the possibility of measuring abili-
ties by self-report, but we are not aware of any question-
naires for EI that demonstrably measures abilities rather
than personality-like traits.
Several developments suggest that it may be time to
move on from the dichotomy between mixed and ability
models of EI (Mayer et al., 2000b) at least in some respects.
First, some questionnaire researchers (e.g., Petrides &
Furnham, 2003) claim that self-report inventories belong
to the domain of personality (trait EI) and do not measure
abilities, even of the “mixed” kind. Second, questions re-
main about whether Mayer et al.’s (e.g., 2000a,b) tests mea-
sure abilities akin to those assessed by conventional intel-
ligence tests (Brody, 2004; Matthews et al., 2002). Some
fresh thinking on how to measure social-emotional abilities
or competencies may be required. Third, conceptual anal-
yses of the kind discussed in the previous section imply that
the current range of tests for EI may not adequately sample
the full range of constructs that may be labeled as “EI”;
again, new assessment methods may be needed. Thus, a
review of assessment methods needs not only to evaluate
extant tests against standard psychometric criteria, but to
examine the fundamental principles being used to sample
the domain of emotional competence as a basis for test de-
velopment. Next, we take up these issues in more depth by
reviewing sources of consensus and controversy in the as-
sessment of emotional intelligence.
Assessment: Sources of Consensus
The discipline of psychometrics provides relatively uncon-
troversial principles for determining what constitutes
sound assessment practices. Indeed, almost all major orga-
nizations concerned with educational and psychological
testing have endorsed a set of standards for determining the
efficacy of a given measure (AERA/APA/NCME Test
Standards, 1999). These standards lay out a framework for
interpreting reliability and validity; in essence, how the
corpus of research should confirm the status of the instru-
ment. Issues of consequential validity (i.e., demonstrating
that the construct assessed by the test has meaningful soci-
68 M. Zeidner et al.: EI: Consensus and Controversies
European Psychologist 2008; Vol. 13(1):64–78 © 2008 Hogrefe & Huber Publishers
etal consequences), fairness (i.e., showing that items are
not biased against a particular subpopulation for inappro-
priate reasons), and how to appropriately document test de-
velopment are also critical components of developing an
EI measure (Matthews, Emo, Roberts, & Zeidner, 2006).
All of these various processes are ongoing and should feed
back to guide theoretical refinements, test development,
and future cycles of research. It is equally important to es-
tablish each piece of evidence.
Emotional intelligence assessments need to demonstrate
adequate reliability. The vast majority of research on self-
reports has concentrated in particular on establishing inter-
nal consistency reliability (e.g., Bar-On, 1995; Schutte et
al., 1998; Tett et al., 2005). Those subscribing to perfor-
mance-based measures most generally calculate split-half
reliability (e.g., Palmer, Gignac, Manocha, & Stough,
2005). Mayer et al. (2004) provide a rationale for the latter
approach when using performance-based assessments ad-
ministered in the fashion that they describe. Of note, we are
unaware of studies that report the standard error of the mea-
surement of any EI test, although this could have been
readily calculated from the summary statistics and test re-
liability coefficient.
It is not incidental that many of the subscales have mar-
ginal (i.e., less than 0.60) reliabilities both for self-report
and performance-based measures, though superordinate
constructs such as Experiential EI, general EI, and the like
have high reliability coefficients (i.e., in excess of 0.90).
Strategies for improving the reliability of subscales, such
as increasing the number of test items, have curiously been
neglected. By way of illustration, the MSCEIT actually has
fewer items than its predecessor the MEIS for several com-
mon subscales; and their reliabilities were marginal. There
are also relatively few studies of the test-retest reliability
of any measure. Moreover, although those that have been
conducted are suggestive, they generally come from stud-
ies with fairly small sample sizes (e.g., Tett et al., 2005).
The jury is still out on whether many existing tests have
high enough reliability coefficients for use in applied set-
Emotional intelligence assessments require various forms
of validity evidence. The AERA/APA/NCME Test Stan-
dards (1999) acknowledge construct validity as an all-en-
compassing, unifying concept overarching all types of va-
lidity evidence. Regrettably, not all forms of validity evi-
dence have been the subject of extant empirical research
on EI. In the quest for construct validity evidence, research
has tended, instead, to focus on factorial validity, conver-
gent and discriminant validity evidence, and test-criterion
relationships (Matthews, Zeidner, & Roberts, 2007). Next,
we briefly discuss some of the validity evidence for EI mea-
Factorial validity. Overall, the data attesting to the factorial
validity of virtually every single extant measure of EI are
equivocal. For example, Schutte et al. (1998) postulated a
single, general factor for the self-report inventory that they
developed. However, Petrides and Furnham (2000b) pro-
vide evidence instead for a four-factor solution, while still
other commentators (e.g., Saklofske et al., 2003) have not
been entirely successful in replicating this, or other, factor
solutions for this measure (see e.g., Gignac, Palmer, Ma-
nocha, & Stough, 2005; who lamentably consider their
study the first to address this issue). Similarly, Matthews et
al. (2002) reanalyzed data from Bar-On’s (1995) EQ-i tech-
nical manual to reveal a number of inconsistencies in the
hypothetical structure purportedly underlying this instru-
ment (see also Livingstone & Day, 2005; Palmer, Manocha,
Gignac, & Stough, 2003). Given the high degree of overlap
between questionnaire assessments of EI and standard per-
sonality traits, the onus is on researchers working with a
mixed-model or trait EI to develop factor models that spec-
ify the relationship between EI and standard personality
traits as latent constructs. However, little progress has been
made in this direction. Petrides and Furnham (2001) claim
that trait EI is a lower-order construct within the FFM, but
their factor analytic data actually show that different facets
of trait EI load on different factors. Some facets attach to
the standard Big Five, whereas others define a separate fac-
tor. Researchers have yet to explore the multistratum factor
models that have helped to clarify psychometric confusions
in the standard intelligence field (Carroll, 1993).
Performance-based measures fare no better when it
comes to showing robust, theoretically-defensible factor
structure. For example, although there are several studies
that allege the MSCEIT has four recoverable branches
(Mayer, Salovey, Caruso, & Sitarenios, 2001), virtually no
published study has been able to find evidence for an inde-
pendent Emotional Facilitation (i.e., using emotions to fa-
cilitate thought) construct (see e.g., Palmer et al., 2005;
Roberts et al., 2006). Problems in factor structure also hold
true for the MSCEIT’s predecessor, the MEIS (Roberts et
al., 2001; Zeidner et al., 2001). Furthermore, research is
sorely needed to test for factorial invariance of current
measures of EI across different sociocultural groups.
Convergent and discriminant validity evidence. Perfor-
mance-based tests show convergent validity against ability
measures, correlating positively with verbal, knowledge-
based tests (i.e., crystallized intelligence), particularly for
Understanding Emotion (Branch 3), while at the same time
being relatively weakly related to tests of reasoning ability
(fluid intelligence; see Ciarrochi et al., 2000; Mayer et al.,
2000b). Generally, the evidence suggests that ability-based
EI-measures index emotional knowledge, which is related
to crystallized intelligence. By contrast, self-report mea-
sures of EI have shown poor convergent validity. For ex-
ample, thus far, they show low (near zero) correlations with
traditional forms of intelligence (see e.g., Davies, Stankov,
& Roberts, 1998; Derksen, Kramer, & Katzko, 2002; Sala,
2002; Zeidner et al., 2005).
There is a growing body of evidence that self-report as-
sessments of EI assess dispositional traits rather than a form
M. Zeidner et al.: EI: Consensus and Controversies 69
© 2008 Hogrefe & Huber Publishers European Psychologist 2008; Vol. 13(1):64–78
of intelligence. Unfortunately, the magnitude of correlation
between the vast majority of self-report assessments of EI
and (a lack of) Neuroticism (particularly, anxiety) is strong,
with moderate to high correlations also evident between the
self-report assessments and Agreeableness, Conscientious-
ness, and Extraversion, for a good deal of the available
scales (see MacCann et al., 2003; Matthews et al., 2002).
Given correlations with several personality variables it
seems possible that once the variance associated with per-
sonality is partialled out, EI-related variance would be min-
imal. Collectively, these findings suggest that self-report
measures of EI, whether based on mixed or trait models,
have questionable discriminant validity.
A final piece of convergent validity evidence is the re-
lations between different EI assessments. EI measures do
not fare well in this respect. The correlation between per-
formance-based and self-report assessments is surprisingly
low, ranging somewhere between 0.20 and 0.30 across
many studies. More problematic, we have shown that op-
posite conclusions can be reached on the basis of these two
different assessment approaches when one also looks at ex-
ternal variables like intelligence or at group differences
(Zeidner et al., 2005).
Test-criterion relations. The test-criterion relation refers to
what was formally known as predictive validity. For both
self-report and ability scales, this is often mixed. For ex-
ample, Brackett and Mayer (2003) tested both kinds of
measure as predictors of various criteria for emotional
competence, controlling for the FFM and cognitive ability.
In each case, EI predicted only one out of six criteria, with
small effect sizes. By contrast, other studies have demon-
strated modest test-criterion relations. In the ability field,
several studies (see Lopes et al., 2004; Rivers, Brackett,
Salovey, & Mayer, 2007) have shown that the Managing
Emotions branch of the MSCEIT predicts social function-
ing with personality and ability controlled. Likewise, stud-
ies have shown that self-report scales for EI can add mod-
estly to prediction of well-being criteria, with the Big Five
controlled (e.g., Saklofske et al., 2003). Although uncor-
rected validity coefficients for questionnaire assessments
of EI tend to be higher than for the MSCEIT, most of their
predictive validity derives from overlap with personality.
(Some self-report scales, as we have previously argued, are
also vulnerable to criterion contamination, in that they in-
clude items referring to well-being or social success). It is
striking that, with personality and ability controlled, both
types of measure provide only modest incremental validity
at best, typically adding 5% or so to the variance ex-
Assessments of emotional intelligence need to be premised
on justifiable scoring rubrics. There is little doubt that sim-
ply by virtue of the methods, and the widespread use of this
technique across psychology, investigating self-reported EI
can be justified as a research tactic. However, by contrast,
compared to more traditional intelligence tests (e.g., vocab-
ulary, matrices), the scoring of performance-based EI tests
is difficult, as there is no algorithm for determining the
correct answer (MacCann, Roberts, Matthews, & Zeidner,
2004; Roberts et al., 2001; Zeidner et al., 2001). Objectiv-
ity, or “the problem of the correct answer” (Mayer et al.,
2000b), has proven hugely problematic in investigations of
social intelligence, and has, thus far, proven an equally dif-
ficult issue to traverse in emotional intelligence research.
The MSCEIT (and the earlier MEIS) deal with this problem
in one of three ways: (1) assuming emotion experts know
the answer (expert scoring), (2) assuming that the stimulus
creators know the answer (target scoring), or (3) assuming
that the correct answer is what people generally agree is
correct (consensus scoring). Each of these scoring tech-
niques has problems (see Matthews et al., 2007).
Assessment: Sources of Controversy
Is the proposed dichotomy between ability and mixed mod-
els useful? The current trend toward relating self-reports of
EI to the personality domain (Petrides & Furnham, 2003)
leaves the earlier mixed models of EI (e.g., Bar-On, 2000)
in limbo. Should we abandon Bar-Ons idea that question-
naires may be used to assess abilities, or does the notion
still have credence? Again, it may be useful to explore dif-
ferentiation of constructs within the self-report domain.
Thus, although the split between performance and ques-
tionnaire tests looms large in the present context, other un-
certainties over the optimal choice of assessment methods
are appearing as the field develops.
For example, assessing one’s geographic knowledge by
asking a series of question along the lines of “what is the
capital city of Greenland” seems more cost-effective and
valid than asking a person to “rate how good you are in
geography on a seven-point scale from awful to brilliant”
(the answer to which might suddenly change if you were
offered a decent sum of money). Given the choice to assess
intelligence with a question that is factually verifiable or a
subjective rating, even the staunchest advocate of the later
approach is forced to concede this is a no-brainer. Besides
having a good deal more face and ecological validity, ve-
ridical items are less impervious to faking, coaching, or
self-deception biases. Notwithstanding, self-estimates of
intelligence (or related constructs, stressing in particular
cognitive engagement) have been used in research settings
to generate a variety of theoretically meaningful findings
(e.g., Ackerman & Goff, 1994; Furnham & Rawles, 1999;
Rammstedt & Rammsayer, 2000). Thus, it is important not
to dismiss self-report approaches to EI out of hand.
Are the stakes at which assessment is targeted defensible?
Basic research aside, psychological testing is generally
conducted for some practical purpose, with varying impli-
cations. In general, practitioners and policy makers talk of
the tests falling into one of three categories, corresponding
to the “fidelity” of the instrument in question for decision-
70 M. Zeidner et al.: EI: Consensus and Controversies
European Psychologist 2008; Vol. 13(1):64–78 © 2008 Hogrefe & Huber Publishers
making purposes. These are (1) high-stakes (e.g., determin-
ing whether an individual can enter the college of their
choice after taking an assessment); (2) medium-stakes
(e.g., ascertaining whether an individual may move into a
higher position on the corporate ladder on the basis of a test
score); or (3) low-stakes (e.g., a test designed to give the
individual a certain level of self-insight). Currently, there
is considerable push to bring EI measures into the high
stakes testing arena. Clearly, however, in the absence of
more carefully documented validity evidence, such usage
is questionable.
Is it not time for more advanced psychometric analyses?
As alluded to at several points, thus far a good deal of EI
research has been conducted without particularly advanced
psychometrics. We are aware, for example, of no published
study using item response theory (Embertson & Reeise,
2000; van der Linden, 1996), differential item functioning
(Holland & Wainer, 1993), or equating (Kolen & Brennan,
2004), to name but a few of the statistical procedures com-
monly employed in cognitive assessment today. These pro-
cedures are especially important for high-stakes assess-
ment, and feed in to a previous concern that we had that
contemporary EI assessments should be used at a justifiable
Applications of Emotional Intelligence
As alluded to in the introduction, a major driving force fu-
eling the growing public and scientific interest in EI, and
cognate emotional and social competencies, is their poten-
tial for improving personal and societal well-being. Thus,
EI has been claimed to play a pivotal role in such diverse
domains as job performance, interpersonal relationships,
educational achievements, and clinical disorders. This final
section will address applications of EI to these various do-
mains of human endeavor.
As already demonstrated, empirical research has not al-
ways supported many of the validity claims surrounding
this concept. Nevertheless, there is sufficient evidence for
incremental validity of the better EI measures to suggest
that a focus on EI may be relevant to enhancing personal,
social, and organizational functioning and adaptation
(Mayer et al., 2000a,b). There are also well-validated inter-
vention programs that are designed to improve emotional
functioning, especially in education (Zins, Weissberg,
Wang, & Walberg, 2004). Such programs lend weight to
the idea that elevating EI may be a valuable practical strat-
egy in many real-life settings.
EI has been claimed to be predictive of individual task
performance at the workplace, especially in settings requir-
ing leadership, teamwork, or effective communication, as
well as contextual or tacit performance (e.g., Abraham,
2005; Daus & Ashkanasy, 2005). EI may also relate to cit-
izenship behavior, integrity, and effective personal rela-
tionships in organizational settings.
In education, EI skills and competencies that are culti-
vated and trained in social and emotional learning pro-
grams are commonly believed to be able to help motivate
students to reach higher levels of achievement, become
more socially and emotionally competent, and to become
more responsible and productive members of society. It is
thought that elevating EI will impact both overt academic
goals, such as better grades, and the student’s broader per-
sonal development (Zins et al., 2004). Reviews of the evi-
dence on programs for social-emotional learning, including
those using meta-analysis to demonstrate change in out-
come criteria, support their efficacy in improving mental
health, academic performance, and remediation of various
behavior problems (Greenberg, Weissberg, O’Brien, &
Zins, 2003).
EI may also have considerable potential for clinical ap-
plications. Assuming EI is related to disordered affect and
dysfunctional affect regulation (which, in turn, is related to
psychopathology) it might play an important role in clinical
diagnosis and treatment (Parker, 2005). Research on alex-
ithymia highlights how difficulties in understanding and
communicating emotion may be important in affective dis-
orders (Taylor & Bagby, 2004). However, while many
mental disorders are related to emotional dysfunctions and
expressions of negative affect, the diversity of these disor-
ders may mitigate against an unambiguous relationship
with low EI (Matthews et al., 2002).
We now point to a number of areas in which there ap-
pears to be a general consensus about applications, before
again touching on what appear as major sources of contro-
Applications: Sources of Consensus
There has been irrational enthusiasm surrounding the
practical utility of emotional intelligence. Many scholars
now working in the area agree that there has been an initial,
irrational exuberance regarding the practical value of EI in
applied settings (see e.g., Landy, 2006). Barrett, Gross,
Christensen, and Benvenuto’s (2001) review suggests that
much of the existing evidence bearing on the role of EI in
occupational success is anecdotal, impressionistic, or col-
lected by consulting companies and not published in the
peer-reviewed literature. Further, Zeidner, Roberts, and
Matthews (2002) have pointed out that despite popular
claims, most of the programs touted as effective EI pro-
grams lack clear conceptual frameworks, implementation
analyses and checks, and sound evaluation designs. At pre-
sent, there are few EI training programs that have been sys-
tematically constructed, implemented, and assessed. For
example, some (though certainly not all) EI programs are
being implemented in school settings without sufficient
theoretical grounding, intervention hypotheses, or rigorous
evaluation studies. Nevertheless, the failings of current ap-
M. Zeidner et al.: EI: Consensus and Controversies 71
© 2008 Hogrefe & Huber Publishers European Psychologist 2008; Vol. 13(1):64–78
plied work and training programs do not negate the possi-
bility that more modest practical gains may be attainable.
EI appears related (albeit weakly) to performance outcomes
in a variety of applied settings.Overall, research suggeststhat
EI modestly predicts outcomes in a variety of real-life set-
tings, with evidence available mainly for occupational (Daus
& Ashkanasy, 2005) and educational contexts (Zins et al.,
2004). In occupational settings, the additional contribution of
EI to prediction over and above personality and ability is
typically limited (Day, 2004). The majority of studies have
used criteria that are based on self-report; objective behavior-
al criteria are regrettably neglected. A concern with the occu-
pational studies reviewed by Van Rooy and Viswesvaran
(2004) is that many use supervisor ratings that may be influ-
enced by the likability of the employee, rather than their job
competence. At the same time, there seems to be growing
confidence among organizational psychologists that tests for
EI predict job performance to an extent that is practically
useful (Daus & Ashkanasy, 2005). Our view is that the scope
and importance of the validity coefficients for EI remains
open for debate, but the proponents of EI have made progress
in demonstrating that the scales have sufficient criterion va-
lidity to be taken seriously.
EI has been claimed to be directly predictive of work
performance and job satisfaction, organizational citizen-
ship, truancy at work, and prosocial behavior. A review by
Daus and Ashkanasy (2005) suggests that for jobs that
would appear logically to require a high level of EI (e.g.,
police officers) relationships between EI and job perfor-
mance and satisfaction may be higher than those where
emotional demands are less obvious. There are, however,
specific problems for the use of EI in occupational settings
(Zeidner, Matthews, & Roberts, 2004). These include: (1)
failure to provide an adequate theoretical rationale for their
use in a particular occupational setting, (2) lack of occupa-
tional-specificity, (3) absence of normative data for differ-
ent occupational groups, and (4) failure to provide evidence
for predictive and discriminate validity (both within and
among occupational clusters). Thus, the applied psycholo-
gist must fall back on clinical or professional judgment to
gauge the role of EI in many contexts, lacking a proper
science-based analysis.
In education, EI appears a rather weak predictor of aca-
demic success per se. Thus, based on a limited number of
studies, the best estimate of the true validity coefficient for
the relationship between EI and academic success (i.e.,
grades) is small (around 0.10). In fact, of all the perfor-
mance domains, academic success may be the one where
EI has the least potential. Furthermore, the EI-performance
relationship appears to be both measure dependent (higher
for self-report than ability measures), as well as criteria de-
pendent (Van Rooy, Dilchert, Viswesvaran, & Ones, 2006).
A more compelling argument, though, may be that EI in-
directly mediates success by protecting students from bar-
riers to learning such as mental distress, substance abuse,
delinquency, teen pregnancy, and violence (Hawkins,
Smith, & Catalano, 2004). Equally, the criterion space for
studying academic success has so far been narrow; there is
more to academic success than grades. Thus, retention, cit-
izenship, and psychological well-being all appear impor-
tant outcome variables to consider in the educational sec-
tor, each of which have so far received short shrift (Roberts
et al., 2005).
It has also been claimed that EI has merit and practical
utility in predicting a broad set of outcomes in the social
domain, such as quality of social relationships, marital suc-
cess, prosocial behaviors, and delinquent behaviors. Com-
pared to academic performance and occupational criteria,
these outcomes have often been difficult to measure and
operationalize (Van Rooy et al., 2006). Clearly, as we sug-
gested earlier, the theoretical rationale for why EI should
be predictive of criteria in the social realm needs to be more
fully delineated. Thus, a broadside approach should not be
adopted whenever EI is used to try to predict any possible
outcome without specifying how and why these outcomes
are important.
Systematic approaches are required to match emotional in-
telligence (sub)constructs to applications. In the special
contexts of organization and industry, there is currently no
empirically validated taxonomy of job types corresponding
to separate components of EI. Thus, we cannot ascertain
what facets of EI are requisite for any given job cluster.
Generally, the level of EI apparently critical for occupa-
tional success should be a function of how central EI may
be to work activities. Thus, EI may be more important in
service organizations than others. Furthermore, EI may be
more important for those occupying lower positions in the
hierarchy relative to those much higher up in the organiza-
tional chain (Daus, 2006). Thus, a more systematic ap-
proach to matching emotional competencies to career com-
ponents is needed. For example, a fine-grained analysis of
the emotional demands imposed on police officers might
support development of a measure of emotional regulation
that could be used to assess and select police officers. This
measure, in turn, could be validated against job-specific
behavioral criteria, such as frequency of angry verbal be-
haviors during encounters with the public. Just as tradition-
al job analysis is increasingly being supplemented by cog-
nitive task analysis, so too there appears a need for “emo-
tional task analysis” to ascertain the affective requirements
of different occupations. At present, practitioners may need
to rely on a relatively superficial dissection of emotional
requirements. However, as theories of emotional compe-
tence become more fully articulated, more theory-driven
analysis of emotional tasks at work may become possible.
Applications: Sources of Controversy
Are current research designs adequate? Studies assessing
the predictive validity of EI have failed to employ measures
of EI that predict career success or other important educa-
72 M. Zeidner et al.: EI: Consensus and Controversies
European Psychologist 2008; Vol. 13(1):64–78 © 2008 Hogrefe & Huber Publishers
tional and social criteria. The most basic task for validation
research is to show that EI measures reliably differentiate
between low- and high-performing groups on particular
criteria. In the occupational domain, such studies should
focus on predicting success both across jobs and within
jobs, identifying the occupations for which EI is more and
less important (e.g., social workers vs. financial analysts).
The use of EI component sub-tests also needs to be val-
idated using large-scale, trait-performance validation de-
signs. It is highly plausible that effective performance in
different occupations involves different patterns of emo-
tional (or social) characteristics. The criteria against which
EI predictors in occupational selection and placement are
validated should likewise be valid, reliable, and uncontam-
inated. The same holds for educational, clinical, and social
outcomes (Zeidner, Matthews, & Roberts, 2004).
The impetus of proponents of EI in applied settings should
be on testing the validity of EI in predicting a wide array of
meaningful criteria. As a first step, it would seem important
to look for the variance explained by EI with regard to con-
ventional criteria (e.g., in the workplace: supervisor’s ratings
of performance, objective criteria such as absenteeism) and
whether EI remains predictive with general intelligence and
personality factors statistically controlled.
To what extent is the predictive validity of emotional intel-
ligence moderated by the nature of the applied setting? Be-
fore employing EI measures in a particular setting, it is
essential to precisely identify the specific contexts, needs,
and purposes for which the EI test is being used. Without
sounding trite, different jobs call for varying levels of social
and emotional involvement and activity. Disparate occupa-
tions also require different types of interpersonal interac-
tion. In some jobs (e.g., social work), one interacts emo-
tionally with others during most of the time on the job. In
such professions, there is a real need to have frequent in-
terchanges with clients at an emotional level. Incumbents
in some jobs not only need to talk with others face-to-face
and exhibit positive, prosocial behavior (e.g., receptionist),
but also assess the reactions of others, and attempt to influ-
ence others’ emotions and motives (e.g., insurance agent).
Some jobs require matching one’s own behavior to the
needs of others (e.g., psychotherapist), creatively influenc-
ing others by engaging their emotions, and transforming
one’s own emotions and also those of others. In other jobs
(e.g., statistician), one interacts with people a smaller per-
centage of time, such that the need to be able to recognize
and manipulate others’ feelings is relatively unimportant,
but one may need to manage personal frustrations. By the
same token, EI may be effective in predicting educational
outcomes in some domains (e.g., social sciences) yet not
be predictive of outcomes in others (e.g., math and physical
How are emotional intelligence measures best used in
practical situations? Should EI measures be used together
with other variables in the predictor stock in a multiple re-
gression prediction equation of relevant job or educational
performance criteria, or used in a noncompensatory “mul-
tiple-hurdle” framework? In this case, a sequential model
is adopted for integration of multiple measures used in any
test battery that assesses, in turn, job or educational-rele-
vant abilities, performance-relevant issues, and appropriate
measures of EI. While time consuming, this process will
most likely result in more accurate assessment.
To what degree can we develop and train emotional intel-
ligence? Programs for helping managers and would-be
leaders, as well as students, to become more emotionally
intelligent have mushroomed in recent years. Although
many of these programs are promising, few have been
modeled upon EI theory or designed in a way that is likely
to lead to long-term change. Furthermore, intervention pro-
grams that seek to raise EI sometimes lack a clear theoret-
ical and methodological rationale, and employ a miscellany
of techniques, whose psychological bases are not always
clear (Zeidner et al., 2002). EI, and the competencies linked
to it, are based on temperament, learning experiences, and
reflective goal-oriented experiences. One-day seminars or
workshops can be valuable in educating people and raising
awareness, but they may not by themselves lead to the kind
of reprogramming that is required for significant improve-
ment (Cherniss & Adler, 2000).
Which components of emotional intelligence are the most
malleable or susceptible to intervention(s)? At present, it
remains uncertain which of the components of EI are most
malleable and responsive to training; what the threshold
level of EI is for training; or what age level EI components
are most responsive to instruction and training. Equally,
little is known of the following key facets of training: spe-
cific goals, specific EI components most responsive to
training, most effective interventions to use for low vs. av-
erage EI clients, and the minimal level of EI that a client
needs to benefit from therapy. There is also a need for de-
veloping standards for program implementation as well as
employing cost-benefit analysis for assessing the return for
costs associated with delivering EI programs.
Our review suggests that there are several difficulties in
trying to evaluate the overall scientific contribution of EI
research (Matthews et al., 2002). These are summarized,
using the three key domains covered in the main body of
this review concepts, assessments, and applications in
the passages that follow. Because we believe some of these
represent tractable research issues, these passages also con-
tain suggestions for future research directions.
M. Zeidner et al.: EI: Consensus and Controversies 73
© 2008 Hogrefe & Huber Publishers European Psychologist 2008; Vol. 13(1):64–78
Currently, there is no agreed upon definition of the EI con-
struct. Thus, it is presently unclear whether EI is cognitive
or noncognitive; whether it refers to explicit or implicit
knowledge of emotion; and whether it refers to a basic ap-
titude or to some adaptation to a specific social and cultural
milieu (Zeidner et al., 2001). Arguably, one of the primary
initial tasks in any scientific endeavor is the systematic
mapping out of the major components and facets in the
universe of discourse under consideration (Kerlinger,
1973). However, it is difficult to obtain a satisfactory jus-
tification or definitional framework for the construct of EI
(Zeidner et al., 2001).
A precondition for any large-scale applications of EI in
applied settings is the development of well-defined theo-
retical frameworks and assessment tools. We believe that
resolving uncertainties over the conceptualization and the-
ory of EI requires better structural models than are present-
ly available or being employed in EI research. The analysis
represented in Table 1 suggests that there may be quite dif-
ferent “EI” constructs to be found in different research do-
mains. Various accounts of the key components of EI exist,
but these models are in need of systematic comparison and
integration. We also see a need for weeding out those con-
structs that are not well-supported by research or lack a
sound theoretical base. Building better conceptual models
of EI also requires better psychometric models that specify
both the overlaps and the uniqueness of EI constructs in
relation to personality and ability.
Presently, it is uncertain how EI may best be measured or
assessed. Various tests of the construct fail to converge and
these diverse measures may, in fact, be assessing different
constructs. There is less consensus regarding appropriate
vehicles and methods for assessing EI than there appears
to be for even the rather under-developed conceptual mod-
els that saturate the field. Arguably, this is an unsatisfactory
state of affairs, one that may be conceived as inhibitory to
the development of a systematic framework for conducting
EI research. Certain points of consensus are virtually dic-
tated by the scientific community, most especially with re-
spect to standards set for psychological and educational
The two different approaches taken by researchers to the
assessment of EI has led to two separate scientific litera-
tures emerging on the topic. Often the findings coming out
of these two emerging research traditions do not converge.
Thus, it is perhaps an important undertaking of future re-
search to provide a synthesis of these approaches, should
this be possible. Conversely synthesis may not really be
possible; perhaps it is best to adopt one approach over an-
other according to a cogent set of arguments. Oatley (2004)
has suggested over-emphasis on measurement issues has
led to neglect of the psychology of EI. The initial question,
of course, calls for a Kuhnian answer, namely, only time
will tell whether the evidence is sufficient to persuade the
research community that a paradigm-shift is required.
Regardless, the sheer number of self-report measures so
far developed in this field, without attendant concerns for
the procurement of compelling validity evidence, suggests
that it is time to call for a moratorium on the development
of still further instruments of this type. By contrast, the
number of performance measures is surprisingly small
(though clearly some of those used by emotions researchers
could be reconceptualized as measures of specific EI con-
structs). Developing further objective measures of EI
would appear to be an important future research endeavor.
Clearly, the practical utility of tests of EI is limited by the
conceptual and psychometric deficiencies described above.
Thus, available tests generally have questionable validity
evidence for the tests to be used with confidence in making
real-world decisions (Zeidner, Matthews, & Roberts,
2004). Moreover, intervention programs targeted to en-
hance EI in educational (Zeidner et al., 2002) and occupa-
tional settings (Zeidner, Matthews & Roberts, 2004) appear
sometimes to lack clearly articulated theoretical and meth-
odological rationales.
Arguably, in order for EI to be useful in applied settings,
the theoretical foundations of EI need to be secured first;
otherwise it will remain a fuzzy and slippery construct,
with little practical value. It is critical that definition issues
are settled before proceeding to operationalize and apply
the construct in real-life settings. Of note, research has yet
to demonstrate the added value of using EI measures in
applied settings relative to narrower measures of individual
There is a danger that EI may be no more than a fad of the
type common in business and education (Murphy & Side-
man, 2006). The explosive growth of EI matches the three
defining characteristics of such fads identified by these au-
thors: (1) a fast growth trajectory, (2) promise of a great deal
more than can be delivered, and (3) provocation of intense
reactions, both positive and negative. Such fads, Murphy and
Sideman (2006) go on to argue, generally follow a natural
life-cycle. Interest tends to plateau, followed by a precipitous
decline, although the fad may re-emerge years or decades
later. While the optimist may believe EI is here to stay, the
pessimist may believe that EI will burn out before too long.
Concluding Comments
Finally, research surrounding EI is still new, and the optimist
may take the diversity of concepts and measures as a sign of
the vitality of the field. As research addresses the controver-
74 M. Zeidner et al.: EI: Consensus and Controversies
European Psychologist 2008; Vol. 13(1):64–78 © 2008 Hogrefe & Huber Publishers
sies we have highlighted here, we can expect to see a shake-
out of misconceptions, the maturing of the science,and great-
er consensus between advocates for EI and skeptics. Progress
depends on greater rigor in conceptualization and measure-
ment, and in validation of scales against objective criteria for
social-emotional functioning. In applied fields, the focus
should be on evidence-based studies that show that interven-
tions specifically directed toward EI improve over other,
well-attested techniques for improving social functioning.
The spirit of “letting a hundred flowers bloom should not be
used as an excuse for poor science or for practical interven-
tions that promise more than they deliver. Will EI stay or will
it go? Only time will tell.
Authors’ Note
This paper is an expanded version of a presentation given
by the author at the 26th International Congress of Applied
Psychology, Athens, 2006. The views presented are those
of the authors and do not represent the views of their insti-
tutional affiliations.
Abraham, R. (2005). Emotional intelligence in the workplace: A
review and synthesis. In R. Schulze & R.D. Roberts (Eds.), In-
ternational handbook of emotional intelligence (pp. 255–270).
Cambridge, MA: Hogrefe & Huber.
Ackerman, P.L., & Goff, M. (1994). Typical intellectual engage-
ment and personality: Reply to Rocklin. Journal of Education-
al Psychology, 86, 150–153.
American Educational Research Association, American Psycho-
logical Association, National Council on Measurement in Ed-
ucation. (1999). Standards for educational and psychological
testing. Washington, DC: American Educational Research As-
Austin, E.J. (2005). Emotional intelligence and emotional infor-
mation processing. Personality and Individual Differences, 39,
Austin, E.J., & Saklofske, D.H. (2005). Far too many intelligenc-
es? On the communalities and differences between social,
practical, and emotional intelligences. In R. Schulze & R.D.
Roberts (Eds.), International handbook of emotional intelli-
gence (pp. 107–128). Cambridge, MA: Hogrefe & Huber.
Austin, E.J., Saklofske, D.H., & Egan, V. (2005). Personality,
well-being, and health correlates of trait emotional intelli-
gence. Personality and Individual Differences, 38, 547–558.
Barchard, K. (2003). Does emotional intelligence assist in the pre-
diction of academic success? Educational and Psychological
Measurement, 63, 840–858.
Bar-On, R. (1995). EQ-i: The emotional quotient inventory man-
ual. A test of emotional intelligence. New York: Multi-Health
Systems, Inc.
Bar-On, R. (1997). The Emotional Intelligence Inventory (EQ-i):
Technical manual. Toronto, Canada: Multi-Health Systems.
Bar-On, R. (2000). Emotional and social intelligence: Insights
from the Emotional Quotient Inventory. In R. Bar-On & J.D.A.
Parker (Eds.), The handbook of emotional intelligence
(pp. 363–388). San Francisco: Jossey-Bass.
Bar-On, R. (2004). The Bar-On Emotional Quotient Inventory (EQ-
i): Rationale, description, and summary of psychometric proper-
ties.InG.Geher(Ed.),Measuring emotional intelligence: Com-
mon ground and controversy (pp. 111–142). Hauppauge, NY:
Nova Science Publishers.
Barrett, L.F., Gross, J., Christensen, T.C., & Benvenuto, M.
(2001). Knowing what you’re feeling and knowing what to do
about it: Mapping the relation between emotion differentiation
and emotion regulation. Cognition & Emotion, 15, 713–724.
Brackett, M.A., & Mayer, J.D. (2003). Convergent, discriminate,
and incremental validity of competing measures of emotional
intelligence. Personality and Social Psychology Bulletin, 29,
Brackett, M.A., Mayer, J.D., & Warner, R.M. (2004). Emotional
intelligence and its relation to everyday behaviour. Personality
and Individual 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 emo-
tional intelligence. Journal of Personality and Social Psychol-
ogy, 91,780–95.
Brody, N. (2004). What cognitive intelligence is and what emo-
tional intelligence is not. Psychological Inquiry, 15, 234–238.
Carroll, J.B. (1993). Human cognitive abilities: A survey of fac-
tor-analytic studies. New York: Cambridge University Press.
Caspi, A., & Bem, D. (1990). Personality continuity and change
acrossthelifecourse.InL.A.Pervin(Ed.),Handbook of person-
ality theory and research (pp. 549–575). New York: Guilford.
Cherniss, C., & Adler, M. (2000). Promoting emotional intelli-
gence in organizations. Alexandria, VA: American Society for
Training and Development.
Ciarrochi, J., Chan, A., & Caputi, P. (2000). A critical evaluation
of the emotional intelligence construct. Personality and Indi-
vidual Differences, 28, 539–561.
Ciarrochi, J., Deane, F.P., & Anderson, S. (2002). Emotional intelli-
gence moderates the relationship between stress and mental
health. Personality and Individual Differences, 32, 197–209.
Cooper, R.K., & Sawaf, A. (1997). Executive EQ: Emotional intel-
ligence in leaders and organizations. New York: Grosset/Put-
Corno, L., Cronbach, L.J., Kupermintz, H., Lohman, D., Mandi-
nach, E.B., Porteus, A.W., et al. (2002). Remaking the concept
of aptitude: Extending the legacy of R.E. Snow. Mahwah, NJ:
Daus, C.S. (2006). The case for an ability-based model of emo-
tional intelligence. In K.R. Murphy (Ed.), A critique of emo-
tional intelligence: What are the problems and how can they
be fixed? (pp. 301–324). Mahwah, NJ: Erlbaum.
Daus, C.S., & Ashkanasy, N.M. (2005). The case for the ability-
based model of emotional intelligence in organizational behav-
ior. Journal of Organizational Behavior, 26, 453–466.
Davies, M., Stankov, L., & Roberts, R.D. (1998). Emotional in-
telligence: In search of an elusive construct. Journal of Per-
sonality and Social Psychology, 75, 989–1015.
Dawda, D., & Hart, S.D. (2000). Assessing emotional intelli-
gence: Reliability and validity of the Bar-On Emotional Quo-
tient Inventory (EQ-i) in university students. Personality &
Individual Differences, 28, 797–812.
M. Zeidner et al.: EI: Consensus and Controversies 75
© 2008 Hogrefe & Huber Publishers European Psychologist 2008; Vol. 13(1):64–78
Day, A. (2004). The measurement of emotional intelligence: The
good, the bad and the ugly. In G. Geher (Ed.), Measuring
emotional intelligence: Common ground and controversy
(pp. 245–270). New York: Nova Science Publishers.
Denham, S.A. (1998). Emotional development in young children.
New York: Guilford
Derksen, J., Kramer, I., & Katzko, M. (2002). Does a self-report
measure for emotional intelligence assess something different
than general intelligence? Personality and Individual Differ-
ences, 32, 37–48.
Dunning, D., Heath, C., & Suls, J. (2004). Flawed self-assess-
ment: Implications for health, education, and the workplace.
Psychological Science in the Public Interest, 69, 106.
Eisenberg, N., Fabes, R.A., & Losoya, S. (1997). Emotional re-
sponding: Regulation, social correlates, and socialization. In P.
Salovey & D.J. Sluyter (Eds.), Emotional development and emo-
tional intelligence: Educational implications (pp. 129–167).
New York: Basic Books.
Embretson, S., & Reise, S.P. (2000). Item response theory for psy-
chologists. Mahwah, NJ: Erlbaum.
Fellner, A.N., Perez, J.C., Emo, A., & Matthews, G. (2006, April).
How is trait emotional intelligence related to personality and to
mood states? Paper presented at the 97th Annual Meeting of the
Southern Society for Philosophy and Psychology, Durham, NC.
Furnham, A., & Rawles, R. (1999). Correlations between self-es-
timated and psychometrically measured IQ. Journal of Social
Psychology, 139, 405–410.
Gardner, H. (1983). Frames of mind: The theory of multiple in-
telligences. New York: Basic Books.
Gignac, G.E., Palmer, B.R., Manocha, R., & Stough, C. (2005). An
examination ofthe factor structureof the Schutte self-report emo-
tional intelligence (SSREI) scale via confirmatory factor analy-
sis. Personality and Individual Differences, 39, 1029–1042.
Goleman, D. (1995). Emotional intelligence. New York: Bantam.
Greenberg, M.T., Weissberg, R.P., O’Brien, M.U., & Zins, J.E.
(2003). Enhancing school based prevention and youth devel-
opment through coordinated social, emotional, and academic
learning. American Psychologist, 58, 466–474.
Harpur, T.J., Hart, S.D., & Hare, R.D. (2002). Personality of the
psychopath. In P.T. Costa, Jr. & T.A. Widiger (Eds.), Person-
ality disorders and the five-factor model of personality (2nd
ed., pp. 299–324). Washington, DC: American Psychological
Hawkins, J.D., Smith, B.H., & Catalano, R.F. (2004). Social de-
velopment and social and emotional learning. In J.E. Zins,
M.R. Bloodworth, R.P. Weissberg, & H.J. Walberg (Eds.),
Building academic success on social and emotional learning:
What does the research say? (pp. 135–150). New York: Teach-
ers College Press.
Holland, P.W., & Wainer, H. (Eds.). (1993). Differential item func-
tioning. Hillsdale, NJ: Erlbaum.
Izard, C.E. (2001). Emotional intelligence or adaptive emotions?
Emotion, 1, 249–257.
Kerlinger, F.N. (1973). Foundations of behavioral research (2nd
ed.). New York: Holt, Rinehart, and Winston, Inc.
Kolen, M.J., & Brennan, R.L. (2004). Test equating: Methods and
practices (2nd ed.). New York: Springer-Verlag.
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.),
The emotional intelligence bandwagon: The struggle between
science and marketing for the soul of EI (pp. 81–123). Mah-
wah, NJ: Erlbaum.
Livingstone, H.A., & Day, A.L. (2005). Comparing the construct
and criterion-related validity of ability-based and mixed-mod-
el measures of emotional intelligence. Educational and Psy-
chological Measurement, 65, 757–779.
Lopes, P.N., Brackett, M.A., Nezlek, J., Schutz, A., Sellin, I., &
Salovey, P. (2004). Emotional intelligence and social interac-
tion. Personality and Social Psychology Bulletin, 1018–1034.
Lopes, P.N., Salovey, P., & Straus, R. (2003). Emotional intelli-
gence, personality, and the perceived quality of social relation-
ships. Personality and Individual Differences, 35, 641–658.
Lowman, R.L. (1991). The clinical practice of career assessment.
Washington, DC: American Psychological Association.
MacCann, C., Matthews, G., Zeidner, M., & Roberts, R.D.
(2003). Psychological assessment of emotional intelligence: A
review of self-report and performance-based testing. The In-
ternational Journal of Organizational Analysis, 11, 247–274.
MacCann, C., Roberts, R.D., Matthews, G., & Zeidner, M.
(2004). Consensus scoring and empirical option weighting of
performance-based emotional intelligence (EI) tests. Person-
ality and Individual Differences, 36, 645–662.
Magnusson, D. (1976). The person and the situation in an inter-
actional model of behavior. Scandinavian Journal of Psychol-
ogy, 17, 253–271.
Matthews, G., Emo, A.K., Funke, G., Zeidner, M., Roberts, R.D.,
Costa, P.T. Jr., et al. (2006). Emotional intelligence, personal-
ity, and task-induced stress. Journal of Experimental Psychol-
ogy Applied, 12, 96–107.
Matthews, G., Emo, A., Roberts, R.D., & Zeidner, M. (2006). What
is this thing called emotional intelligence? In K.R. Murphy (Ed.),
A critique of emotional intelligence: What are the problems and
how can they be fixed? (pp. 3–6). Mahwah, NJ: Erlbaum.
Matthews G., Roberts, R.D., & Zeidner, M. (2004). Seven myths
about emotional intelligence. Psychological Inquiry, 15,
Matthews, G., Zeidner, M., & Roberts, R.D. (2002). Emotional
intelligence: Science and myth. Cambridge: MIT Press.
Matthews, G., Zeidner, M., & Roberts, R.D. (2006). Measuring
emotional intelligence: Promises, pitfalls, solutions? In A.D.
Ong & M. Van Dulmen (Eds.), Handbook of methods in pos-
itive psychology (pp. 189–204). Oxford, UK: Oxford Univer-
sity Press.
Matthews, G., Zeidner, M., & Roberts, R.D. (Eds.). (2007). The
science of emotional intelligence: Knowns and unknowns. Se-
ries in affective science. New York: Oxford University Press.
Mayer, J.D., Caruso, D.R., & Salovey, P. (1999). Emotional in-
telligence meets traditional standards for an intelligence. Intel-
ligence, 27, 267–298.
Mayer, J.D., Caruso, D., & Salovey, P. (2000). Selecting a mea-
sure of emotional intelligence: The case for ability scales. In
R. Bar-On & J.D.A. Parker (Eds.), The handbook of emotional
intelligence (pp. 320–342). New York: Jossey-Bass.
Mayer, J.D., Salovey, P., & Caruso, D.R. (2000a). Competing
models of emotional intelligence. In R.J. Sternberg (Ed.),
Handbook of human intelligence (2nd ed., pp. 396–422). New
York: Cambridge University Press.
Mayer, J.D., Salovey, P., & Caruso, D.R. (2000b). Emotional in-
telligence as Zeitgeist, as personality, and as a mental ability.
76 M. Zeidner et al.: EI: Consensus and Controversies
European Psychologist 2008; Vol. 13(1):64–78 © 2008 Hogrefe & Huber Publishers
In R. Bar-On & J.D.A. Parker (Eds.), The handbook of emo-
tional intelligence (pp. 92–117). San Francisco: Jossey-Bass.
Mayer, J.D., Salovey, P., & Caruso, D.R. (2004). Emotional intel-
ligence: Theory, findings, and implications. Psychological In-
quiry, 15, 197–215.
Mayer, J.D., Salovey, P., Caruso, D.R., & Sitarenios, G. (2001).
Emotional intelligence as a standard intelligence. Emotion, 1,
Murphy, K.R., & Sideman, L. (2006). The fadification of emo-
tional intelligence. In K.R. Murphy (Ed.), A critique of emo-
tional intelligence: What are the problems and how can they
be fixed? (p. 283–299). Mahwah, NJ: Erlbaum Publishers.
Newsome, S., Day, A.L., & Catano, V.M. (2000). Assessing the
predictive validity of emotional intelligence. Personality and
Individual Differences, 29, 1005–1016.
Oatley, K. (2004). Emotional intelligence and the intelligence of
emotions. Psychological Inquiry, 15, 216–238.
Oatley, K., & Johnson-Laird, P.N. (1995). The communicative
theory of emotion: Empirical tests, mental models, and impli-
cations for social interaction. In L.L. Martin & A. Tesser
(Eds.), Striving and feeling: Interactions among goals, affect,
and self-regulation (pp. 363–393). Mahwah, NJ: Erlbaum.
O’Sullivan, M. (2007). Trolling for trout, trawling for tuna: The
methodological morass in measuring emotional intelligence.
In G. Matthews, M. Zeidner, & R.D. Roberts.(Eds), Emotional
intelligence: Knowns and unknowns (pp. 258–287). Cam-
bridge, MA: Oxford University Press.
Palmer, B.R., Gignac, G., Manocha, R., & Stough, C. (2005). A
psychometric evaluation of the Mayer-Salovey-Caruso Emo-
tional Intelligence Test Version 2.0. Intelligence, 33, 285–305.
Palmer, B.R., Manocha, R., Gignac, G., & Stough, C. (2003). Ex-
amining the factor structure of the Bar-On Emotional Quotient
Inventory with an Australian general population sample. Per-
sonality and Individual Differences, 35, 1191–1210.
Panksepp, J. (1998). Affective neuroscience: The foundations of
human and animal emotions. New York: Oxford University
Parker, J.D.A. (2005). The relevance of emotional intelligence for
clinical psychology. In R. Schulze & R.D. Roberts (Eds.), In-
ternational handbook of emotional intelligence (pp. 271–288).
Cambridge, MA: Hogrefe & Huber.
Parker, J.D.A., Taylor, R., & Bagby, M. (2001). The relationship
between emotional intelligence and alexithymia. Personality
and Individual Differences, 30, 107–115.
Petrides, K.V., & Furnham, A. (2000a). Gender differences in
measured and self-estimated trait emotional intelligence. Sex
Roles, 42, 449–461.
Petrides, K.V., & Furnham, A. (2000b). On the dimensional struc-
ture of emotional intelligence. Personality and Individual Dif-
ferences, 29, 313–320.
Petrides, K.V., & Furnham, A. (2003). Trait emotional intelli-
gence: Behavioural validation in two studies of emotion rec-
ognition and reactivity to mood induction. European Journal
of Personality, 17, 39–57.
Rammstedt, B., & Rammsayer, E. (2000). Sex differences in self-
estimates of different aspects of intelligence. Personality and
Individual Difference, 20, 869–880.
Rivers, S., Brackett, M.A., Salovey, P., & Mayer, J.D. (2007).
Measuring emotional intelligence as a set of mental abilities.
In G. Matthews, M. Zeidner, & R.D. Roberts (Eds.), The sci-
ence of emotional intelligence (pp. 230–257). New York: Ox-
ford University Press.
Roberts, R.D., Schulze, R., Zeidner, M., & Matthews, G. (2005).
Understanding, measuring, and applying emotional intelli-
gence: What have we learned? What have we missed? In R.
Schulze & R.D. Roberts (Eds.), International handbook of
emotional intelligence (pp. 311–341). Cambridge, MA: Ho-
grefe & Huber.
Roberts, R.D., Schulze, R., O’Brien, K., Reid, J., MacCann, C.,
& Maul, A. (2006). Exploring the validity of the Mayer-Salo-
vey-Caruso Emotional Intelligence Test (MSCEIT) with estab-
lished emotions measures. Emotion, 6, 663–669.
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.
Rolls, E.T. (1999). The brain and emotion. New York: Oxford
University Press.
Rothbart, M., & Bates, J. (1998). Temperament. In W. Damon
(Ed.), Handbook of child psychology, Vol. 3: Social, emotional,
and personality development (5th ed., pp. 105–176). New
York: Wiley.
Saarni, C. (2000). The social context of emotional development.
In M. Lewis & J. Haviland (Eds.), The handbook of emotion
(2nd ed., pp. 306–322). New York: Guilford.
Saklofske, D.H., Austin, E.J., & Minski, P.S. (2003). Factor struc-
ture and validity of a trait emotional intelligence measure. Per-
sonality and Individual Differences, 34, 707–721.
Sala, F. (2002). Emotional Competence Inventory (ECI): Techni-
cal manual. Boston: Hay/Mcber Group.
Salovey, P., Stroud, L.R., Woolery, A., & Epel, E.S. (2002). Per-
ceived emotional intelligence, stress reactivity, and symptom
reports: Further explorations using the trait meta-mood scale.
Psychology and Health, 17, 611–627.
Schutte, N.S., Malouff, J.M., Hall, L.E., Haggerty, D.J., Cooper,
J.T., Golden, C.J., et al. (1998). Development and validation
of a measure of emotional intelligence. Personality and Indi-
vidual Differences, 25, 167–177.
Shipman, K., & Zeman, K. (1999). Emotion understanding: A
comparison of physically maltreating and nonmaltreating
mother-child dyads. Journal of Clinical Child Psychology, 28,
Taylor, G.J., & Bagby, R.M. (2004). New trends in alexithymia
research. Psychotherapy and Psychosomatics, 63, 68–77.
Tett, R.P., Fox, K.E., & Wang, A. (2005). Development and val-
idation of a self-report measure of emotional intelligence as a
multidimensional trait domain. Personality and Social Psy-
chology Bulletin, 31, 859–888.
Trinidad, D.R., Unger, J.B., Chou, C., Azen, S.P., & Johnson,
C.A. (2004). Emotional intelligence and smoking risk factors
in adolescents: Interactions on smoking intentions. Journal of
Adolescent Health, 34, 46–55.
Van der Linden, W.J. (1996). Assembling tests for the measure-
ment of multiple traits. Applied Psychological Measurement,
20, 373–388.
Van Rooy, D.L., Dilchert, S., Viswesvaran, C., & Ones, D.S.
(2006). Multiplying intelligences: Are general, emotional, and
practical intelligence equal? In K.R. Murphy (Ed.), The case
against emotional intelligence: What are the problems and
how can they be fixed? (pp. 235–262). Mahwah, NJ: Erlbaum.
Van Rooy, D.L., & Viswesvaran, C. (2004). Emotional intelli-
M. Zeidner et al.: EI: Consensus and Controversies 77
© 2008 Hogrefe & Huber Publishers European Psychologist 2008; Vol. 13(1):64–78
gence: A meta-analytic investigation of predictive validity and
nomological net. Journal of Vocational Behavior, 65, 71–95.
Warwick, J., & Nettelbeck, T. (2004). Emotional intelligence is
...?Personality and Individual Differences, 37, 1091–1100.
Waterhouse, L. (2006). Inadequate evidence for multiple intelli-
gences, Mozart effect, and emotional intelligence theories. Ed-
ucational Psychologist, 41, 247–255.
Zeidner, M., Matthews, G., & Roberts, R.D. (2001). Slow down
you move too fast: Emotional intelligence remains an “elu-
sive” intelligence. Emotions, 1, 265–275.
Zeidner, M., Matthews G., & Roberts, R.D. (2004). Emotional
intelligence in the workplace: A critical review. Applied Psy-
chology: An International Review, 53, 371–399.
Zeidner, M., Matthews, G., Roberts, R.D., & MacCann, C. (2003).
Development of emotional intelligence: Toward a multilevel
investment model. Human Development, 46, 69–96.
Zeidner, M., Roberts, R.D., & Matthews, G. (2002). Can emo-
tional intelligence be schooled? A critical review. Educational
Psychologist, 37, 215–231.
Zeidner, M., Roberts, R.D., & Matthews, G. (2004). The emotion-
al intelligence bandwagon: Too fast to live, too young to die?
Psychological Inquiry, 15, 239–248.
Zeidner, M., Roberts, R.D., & Matthews, G. (in press). The emo-
tional intelligence primer. Cambridge, MA: MIT Press.
Zeidner, M., Shani-Zinovich, I., Matthews, G., & Roberts, R.D.
(2005). Assessing emotional intelligence in gifted and nongift-
ed high school students: Outcomes depend on the measure.
Intelligence, 33, 369–391.
Zins, J.E., Weissberg, R.P., Wang, M.C., & Walberg, H.J. (2004).
Building academic success on social and emotional learning:
What does the research say? New York: Teachers College Press.
About the authors
Moshe Zeidner, Ph.D., is Professor of Educational and Personal-
ity Psychology and Director of the Center for the Interdisciplinary
Research on Emotions at the University of Haifa, Israel. His main
research areas are personality and individual differences.
Richard D. Roberts, Ph.D, is a Principal Research Scientist in the
Center for New Constructs at the Educational Testing Service in
Princeton, NJ. His area of specialization is applied psychology,
with a special emphasis on psychological and educational mea-
Gerald Matthews, Ph.D., is Professor of Psychology at the Uni-
versity of Cincinnati. His research focuses on the use of cognitive
science models to understand the interplay between personality,
emotion, and information processing.
Professor Moshe Zeidner
Center for Interdisciplinary Research on Emotions
University of Haifa
Mt. Carmel, 31905
Tel. +972 4 824-0897
Fax +972 4 824-0911
78 M. Zeidner et al.: EI: Consensus and Controversies
European Psychologist 2008; Vol. 13(1):64–78 © 2008 Hogrefe & Huber Publishers
... Nevertheless, individuals may be affiliated with multiple heterogeneous profiles of leadership or followership above and beyond the quantitative difference among them in the emotional context. Potential individual differences (person-centered qualitative differences among the subgroups of employees) concerning perceived LEM and followership include, for instance, emotional information processing (Alam & Singh, 2021;Zeidner et al., 2008) and attitude as well as action to follow (e.g., Ehrnrooth et al., 2020;Epitropaki et al., 2013;Ramaprasad et al., 2021). These individual differences are of great interest in research on emotion and behavior at work yet are still to be demonstrated. ...
... The research concerning perceived LEM is informed by a large body of work on Emotional Intelligence (EI) and its desirable effect on organizational outcomes (Choudhary et al., 2017;Goleman & Boyatzis, 2008;Little et al., 2016;O'Boyle et al., 2011). Given the various understandings of EI (Dimitrijević et al., 2019;Mayer et al., 2008;Zeidner et al., 2008), it is rational to consider perceived LEM as consisting of varying components and, hence, not simply as a homogeneous continuum among individuals but as a set of differing emotional processing states perceived by the employees in the leadership context. The heterogeneity of perceived LEM indicates that employees may be Existing studies typically suggest the multidimensional structure of perceived LEM. ...
Full-text available
Leader emotion management (LEM) and followership are complementary processes. Existing studies have principally deemed LEM and followership as homogeneous conceptualizations across employee populations, lacking the attention on the heterogeneity of these phenomena. This study aims to fill the gap and examine the heterogeneous associative patterns among the profiles of employee-perceived LEM and followership. The sample includes 659 employees in South China. Results suggest that high LEM perceivers were more likely to be exemplary followers, and low LEM perceivers were more likely to be passive followers. Both low LEM and high LEM perceivers were more likely to be exemplary followers among males than females, and complicated LEM perceivers were more likely to be exemplary followers among females than males. Complicated LEM perceivers and high LEM perceivers were more likely to be complicated followers among seasoned staff than newcomers. This study enriches theoretical understandings of leader-member relationships by introducing the employee social constructionism process, which sheds light on the diversification of leading and following during the organizational process.
... V empirických studiích je obecně uznáváno, že testová úzkost se skládá ze dvou složek − obavy (odvrácení pozornosti od testu samotného, vtíravé myšlenky) a emocionality (vegetativní příznaky úzkosti, bušení srdce, nauzea) v testové situaci − a bývá spjata např. s kognitivní interferencí, projevuje se špatnými studijními dovednostmi a tendencí k vyhýbání se výkonové situaci (Zeidner et al., 2008). Testovou úzkost lze obecně popsat jako všudypřítomnou formu akademické úzkosti, která má obecně negativní dopad na vzorce víry v sebe sama a chování ve zkouškových situacích (Cassady, 2010). ...
... Patří k ní také intrapersonální proměnné, kupř. motivace a seberegulace (Dull et al., 2015;Khalaila, 2015;Schnell et al., 2015), sociální vlivy (očekávání a hodnoty výkonu, standardy úspěchu a sociální podpora; von der Embse et al., 2018) nebo demografické proměnné (úroveň vzdělání, ekonomický status a kulturní pozadí; Hembree, 1988;Putwain, 2008;Zeidner et al., 2008). ...
Objectives. We investigated the biological response of organism to stress in real situations at school in adolescents. We were interested in how stress is manifested in adolescents without anxiety and in adolescents with a tendency to anxiety experiencing. The aim of the study is to analyse changes in salivary cortisol levels during stressful situation at school. Hypotheses. The analyses verified how cortisol levels change in situations associated with the threat of social assessment in the school environment in individuals without anxiety and anxious; H1a–c: Anxiety individuals will have higher cortisol levels in all monitored situations compared to adolescentswithout anxiety. We also focused on the possible links between an individual’s tendency to experience anxiety and the type of stress response; H2: Anxiety individuals will show a different development of cortisol levels between situations A, B and C compared to adolescents without anxiety. We also examined the relationships between cortisol levels and cognitive and emotional fear. H3: Anxiety individuals will show significant links between cortisol levels and cognitive and emotional fear in the achievement situation. Methods. The research group was comprised of 238 adolescents in the first phase of the study. The final sample for salivary cortisol analysis was comprised of 38 participants aged 12−14 years. The questionnaires battery contains Piers-Harris II, B-JEPI and TAI. Results. The results show that in both groups of adolescents without anxiety and anxious adolescents, cortisol has a demonstrably different development trend during the day. Furthermore, anxious adolescents showed demonstrably lower cortisol levels at school compared to adolescents without anxiety, in adolescents without anxiety, we observed higher cortisol levels and fall of cortisol level during the day. Our analyses showed that stronger relationships between cortisol and test anxiety are shown especially by anxiety adolescents compared to adolescents without anxiety. Limitations. The limitation is a smaller research sample and the selection of situations in which the level of salivary cortisol was evaluated. The important limitation is the fact that the first sampling was taken at different times after awakening, because the cortisol levels change more dynamically in this part of the day.
... In the scientific literature EI was usually defined in two different ways, i.e., ability model and trait model (Zeidner et al., 2008). Both models have been used in many domains, such as nursing, teaching, physical activity, and teleworking (Law et al., 2004;Smitha et al., 2009;Cebrian et al., 2020). ...
Full-text available
Introduction Basic psychological needs satisfaction (BPNS) and Emotional intelligence (EI) have been underscored as helpful psychological constructs in explaining academic engagement. However, the joint interaction of BPNS with EI abilities to explain academic engagement has not been tested. Therefore, the present study aimed to investigate the interactive role of BPNS with EI abilities in the prediction of academic engagement in a sample of Chinese university students. Methods A questionnaire survey was administered to a sample of 466 university students. The data were analyzed using the SPSS (version 21.0) software. The first analysis consisted of descriptive statistics (including mean and standard deviation) and Pearson’s correlations among BPNS, EI, and academic engagement. Through structural equation modeling (SEM), direct and indirect effects were calculated. Results The results showed that BPNS was positively associated with academic engagement and that only the Use of emotion dimension of EI mediated these associations. Discussion These results suggest that important interventions incorporated with BPNS and EI abilities, especially the use of emotion ability, may be performed to promote university students’ academic engagement.
... As the field advances, researchers are increasingly interested in the processes that underlie the positive effects of EI (Lievens & Chan, 2017). Therefore, an important question is whether dealing with one's own emotions or those of other individuals is equally important for the prediction of criteria (Brasseur et al., 2013;Zeidner et al., 2008). Pekaar et al (2018) propose that both EI dimensions (i.e., dealing with one's own emotions and dealing with others' emotions) may have a positive effect; however, this effect may occur in different life domains. ...
Full-text available
The Rotterdam Emotional Intelligence Scale (REIS) has a good balance between the emotions of self and others, which solves the problem of previous scales being unable to clearly distinguish the emotional intelligence of self from that of others. In the current study, a short version of the REIS scale was developed by the item response theory (IRT) method according to item fit, item or test information function (IIF) and differential item function (DIF) indicators. The results show that the short scale with 14 items had acceptable structural validity (RMSEA = 0.05, CFI = 0.96, TLI = 0.94, SRMR = 0.04) and high reliability (ω\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\omega$$\end{document} = 0.90). The high correlations (more than 0.7) between the scores of the short scale and EIS and WLEIS also verified the concurrent validity of the short scale. In conclusion, compared with the original version of REIS, the developed short version saves 50% of the items and has competitive validity and reliability. The short form will be useful for research and applied contexts where an efficient, concise version is needed.
... In the last three decades, there have been debates about the definition and measurement of emotional intelligence (Zeidner, Roberts, & Matthews, 2008), but three dominant theoretical models remained: emotional coefficient theory developed by Bar-On and Parker (2012), Mayer and Salovey approach with the main focus on emotional skills (1990), and the mixed model of emotional skills developed by Goleman (1995). ...
Objectives. The presence of the so-called Dark Triad (Narcissism, Machiavellianism, and Psychopathy) can be toxic to those around, can lead to counterproductive behaviors and dysfunctional behaviors. On the other hand, people who are successful in personal and professional life are considered to have a high level of emotional intelligence, numerous studies showing that this is a predictor of good functioning and personal, social and organizational adaptation. This study investigates the relationship between the “dark side” of personality and emotional intelligence, and whether or to what extent Narcissism, Machiavellianism, and Psychopathy are negative predictors of emotional intelligence. Material and methods. One hundred and thirty-four individuals aged 21 to 71 years (M = 42.11, SD = 11.68; Females = 71.1%) took part in this study. The following tools were used: Narcissistic Personality Inventory, The Short Dark Triad Scale, Levenson Self-Report Psychopathy Scale and Battery of Emotional Intelligence profile. Results. The regression coefficients were calculated, controlling the role of the other predictors (age, gender, marital status). Narcissism was not a predictor of emotional intelligence (r = .097, p> 0.01). Age and psychopathy were significantly negatively related with the criterion; the gender did not correlate and was removed from the regression model. Emotional intelligence and Machiavellianism did not correlate either positively or negatively, so this dimension cannot be predictor. Conclusions. As predictors of emotional intelligence, the three dimensions, although in many studies have been negatively correlated, in this study this is partially confirmed. Only psychopathy has a negative relationship, it predicts but to a small extent emotional intelligence.
... Further Emotional intelligence has gained good recognition among individuals, but the awareness level among all the employees should be increased. Therefore, the organization must facilitate workshops and soft skill improvement mechanisms to make the employees aware of it and Training on emotional intelligence can be conducted to increase the emotional stability of the employees and this will increase their performance and also their commitment to the work (Zeidner, Roberts & Matthews, 2008). ...
Full-text available
The present study aims to investigate the impact of emotional intelligence on employee engagement among healthcare workers. Employee engagement is critical for completing and executing activities in any form of organization. It is critical to provide organizations with the opportunity to be in a sustainable state for long-term sustainability and gain a competitive advantage over competitors. Employees must recognize, use, and control their own feelings in productive ways to alleviate tension and resolve obstacles in the workplace. Employees' levels of commitment can be affected by emotional symptoms. They can lead to higher or lower morale, which can positively or negatively impact employee engagement. Though previous studies paid attention to engagement level, little attention was paid to the emotional intelligence of healthcare workers. In addition, this study sheds new light on the emotional intelligence level, engagement level, and effect of emotional intelligence on employee engagement among Sri Lankan healthcare employees. A survey method was used and data were collected from conveniently chosen 212 healthcare employees working in public and private hospitals. The results disclose that individuals with high emotional intelligence are more likely to have high employee engagement. The theoretical and practical implications are discussed at the end of the paper. Keywords: emotional intelligence, employee engagement, healthcare workers
... In the scientific literature, EI has been mainly operationalised through the trait-based and specificability models (Mayer et al., 2008;Zeidner et al., 2008). Correlations between trait EI measures and ability EI measures tend to be low, indicating them to be distinct types of EI (Brannick et al., 2009). ...
Full-text available
Emotional intelligence (EI) is defined by the ability to perceive, manage, and reason about emotions in oneself and others. Studies have reported deficits in EI abilities among certain antisocial populations such as individuals with psychopathy, and enhanced performance among sexual offenders. Despite EI's relevance to offending behaviour, the association between EI and paraphilic offending has been under-studied. We examined the association between EI, sexual offending, and sexual sadism in 80 incarcerated men with sexual offences and 207 incarcerated men with non-sexual offences. EI was assessed using the Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT). Sadism was measured using the Severe Sexual Sadism Scale (SeSaS). Results showed that SeSaS scores were positively associated with Strategic EI (the ability to understand and manage emotions), but were not significantly related to Experiential EI. This may reflect core characteristics of sexual sadism including domination and manipulation, challenging the prevalent notion that higher EI is invariably positive. Practice impact statement The study findings suggest that high EI may not always promote prosocial behaviour, which has significant implications for clinical practice. For example, treatment programmes aimed at generally improving EI-related abilities (e.g. emotion regulation) could be refined towards more specific or individualised strategies (e.g. how to effectively and prosocially use EI skills).
... perceiving, expressing), but goes beyond these in terms of empathy and regulation [12]. For several years, there is a controversial debate concerning the EC construct, as to whether it should be presented solely in terms of ability, or whether it should account for both ability and personality characteristics [13,14]. In the meantime, the term "emotional competence" is used as a strongly related term to emotional intelligence (EI) [15,16]. ...
Full-text available
Background Healthcare organisations, such as hospitals, are largely seen as task-oriented, width different people expected to work in interdependent teams. The objective of this study was to investigate the relevance of individual factors (job satisfaction) and individual competences (emotional competence) for organisational commitment in a sample of healthcare professionals. Methods A cross-sectional survey was conducted among 96 healthcare professionals from March to June 2018 in the catchment area of five clinics in Bavaria, Germany. The present research examined the moderating role of emotional competence on the relationship between job satisfaction and organisational commitment using moderated regression analysis and simple slope analysis. Results Multiple regression analysis indicated that emotional competence moderated the relationship between satisfaction with the job and commitment to the job. The results showed that healthcare professionals with high emotional competence are able to deal more effectively with dissatisfaction in the workplace so that organisational commitment remains unaffected. Conclusions Based on the findings of this study emotional competence of healthcare professionals is important for increasing job satisfaction and commitment to the job. Especially for healthcare professionals whose job satisfaction is low, a high level of emotional competence enables them to maintain a high level of organisational commitment. The findings of the study are discussed at the theoretical level for researchers and practical level for hospital managers interested in fostering emotional competence and improving healthcare professionals’ job satisfaction and their organisational commitment, which ultimately may lead to effective performance.
Full-text available
La inteligencia emocional (IE) ha sido definida como la capacidad para identificar y valorar las emociones de sí mismo, y reconocer las de los otros. Así como el conjunto de habilidades para comprender y regular las emociones, facilitando el pensamiento, la atención, la automotivación y la relación con los demás. De tal forma que la IE ha mostrado relaciones significativas con una mejor salud mental y como mediadora del estrés (Goleman, 2021; Salovey y Mayer, 1997). En el contexto del deporte, se ha explorado la relación entre la IE y la formación deportiva, el desarrollo integral y el rendimiento. El objetivo de esta investigación fue describir las investigaciones actuales sobre la IE y el rendimiento deportivo, identificar los métodos de evaluación para la IE y comparar los hallazgos de las investigaciones. Se realizó la búsqueda en las bases de datos Dialnet, Google Scholar, PubMed y Redalyc utilizando las palabras “inteligencia emocional”, “rendimiento deportivo” y “deporte”. Los criterios de inclusión de los artículos fueron: año de publicación 2016- 2021, tipo de investigación, y población conformada por deportistas universitarios o deportistas de rendimiento. Se obtuvieron 20 artículos que cumplieron con los criterios de inclusión, de los cuales 13 fueron en español y siete en inglés; identificando que los instrumentos más utilizados para medir la IE fueron el TMMS-24 (35%) y el SSRI (35%). Los resultados de los distintos artículos relacionan la IE con variables como motivación, autoestima, ansiedad, género, edad y experiencia deportiva; y de manera menos concluyente, con el rendimiento deportivo.
Several lines of research have examined whether people with depressive symptoms have deficits in social-cognitive abilities, such as emotional reasoning skills. While many depressed people report having such deficits, it is less clear whether depressive symptoms are related to actual objective performance deficits. We examined the relationship between emotional reasoning skills (as assessed with the TEMINT) and depressive symptoms (i.e., BDI-II) in a mini meta-analysis of 11 studies with data from 1503 participants with varying levels of depression, from healthy people to clinical samples with severely depressed people. Using a random effects approach, we found a small but significant correlation between depressive symptoms and TEMINT performance (mean rz = 0.065), indicating that depressive symptoms were associated with higher emotional reasoning skills. These findings suggest that depression is unrelated to deficits in emotional reasoning, though assessed with only one test. If anything, depressive symptoms are associated with improved performance in the TEMINT. The current results point to a discrepancy between depressed people's self-evaluation of their abilities (as shown in previous research) and their actual performance. Our findings also have practical implications as they suggest that clinicians may focus on modifying depressed people's negative views of themselves, rather than on improving their skills.
Emotional intelligence (EI) refers to an individual's capacity to process emotional information in order to enhance cognitive activities and facilitate social functioning. It is defined as the perception, use, understanding, and management of one's own and others' emotional states to solve problems and regulate behaviour. This chapter argues that EI is best described as a set of abilities and therefore best measured by abilitybased assessments. First, it presents an overview of EI theory. Second, the chapter describes the Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT), a theoretically derived and empirically validated assessment tool. Third, it reviews additional ability-based tools that assess a subset of the skills measured by the MSCEIT. Fourth, the chapter describes two alternative measurement approaches to EI assessment: self-judgement and information-processing tasks.