European Journal of Personality

Eur. J. Pers. 16: 103–125 (2002)

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/per.434

The Relationship of Emotional Intelligence

with Academic Intelligence and the Big Five

KAREN VAN DER ZEE1*, MELANIE THIJS2 AND LOLLE SCHAKEL2

1
University of Groningen, The Netherlands
2
PiMedia, Maarssen, The Netherlands

Abstract
The present study examines the relationship of self- and other ratings of emotional
intelligence with academic intelligence and personality, as well as the incremental validity
of emotional intelligence beyond academic intelligence and personality in predicting
academic and social success. A sample of 116 students filled in measures for emotional
and academic intelligence, the Big Five, and indicators of social and academic success.
Moreover, other ratings were obtained from four different raters on emotional intelligence
and social success. Factor analysis revealed three emotional intelligence dimensions that
were labelled as ‘Empathy’, ‘Autonomy’, and ‘Emotional Control’. Little evidence was
found for a relationship between emotional and academic intelligence. Academic
intelligence was low and inconsistently related to emotional intelligence, revealing both
negative and positive interrelations. Strong relationships were found of the emotional
intelligence dimensions with the Big Five, particularly with Extraversion and Emotional
Stability. Interestingly, the emotional intelligence dimensions were able to predict both
academic and social success above traditional indicators of academic intelligence and
personality. Copyright # 2002 John Wiley & Sons, Ltd.

INTRODUCTION

For everyday problem solving academic intelligence seems not sufficient. Everyone knows
examples of brilliant persons who outperformed fellow students in college but are not very
successful in a work setting, obviously lacking certain practical and social skills.
Therefore, researchers have introduced constructs that refer to more common sense
approaches to intelligence. In this tradition, measures have been developed for social
intelligence (Archer, 1980; Cantor and Kihlstrom, 1987; Kerr and Speroff, 1954; Stricker
and Rock, 1990), practical intelligence (Sternberg and Wagner, 1986) and emotional
intelligence (see e.g. Davies, Stankov and Roberts, 1998; Salovey and Meyer, 1990;

*Correspondence to: Prof. Dr Karen I. van Oudenhoven-van der Zee, Department of Psychology, University of
Groningen, Grote Kruisstraat 2/1, 9712 TS Groningen, The Netherlands.
E-mail: K.I.van.Oudenhoven-van.der.Zee@ppsw.rug.nl

Received 17 July 2000

Copyright # 2002 John Wiley & Sons, Ltd. Accepted 14 August 2001
104 K. van der Zee et al.

Schutte et al., 1998). These constructs refer to cognitive skills that are needed to solve the
problems that are typically encountered in life, for example in solving conflicts at work,
collaborating with others, or adjustment to new cultural environments and work settings.
Whereas academic problem-solving is characterized by well defined problems for which
one correct answer is usually available, daily problem-solving is characterized by poorly
defined problems, that can be approached from different perspectives and for which no one
best answer is available (Neisser, 1976; Sternberg, 1985, 1997; Wagner and Sternberg,
1986).
The present study focused on the concept of emotional intelligence. Salovey and Mayer
(1990) define emotional intelligence as ‘the ability to monitor one’s own and other’s
emotions, to discriminate among them and to use the information to guide one’s thinking
and actions’ (p. 189). Emotional intelligence is generally regarded as a construct within
the broad framework of human cognitive abilities (Goleman, 1996; Mayer, Caruso and
Salovey, 2000; Mayer and Geher, 1996; Mayer and Salovey, 1993; Salovey and Mayer,
1994). It must be noted, however, that empirical studies (e.g. Davies et al., 1998; Schutte
et al., 1998; Van der Zee, Van Leest and Van Oudenhoven, manuscript submitted for
publication) failed to reveal a clear relationship between emotional intelligence and
academic intelligence. Moreover, studies have revealed a clear relationship between
emotional intelligence and personality. Emotional intelligence seems to be related to
Neuroticism, Extraversion and Agreeableness (Davies et al., 1998) and there is some
evidence that the construct is also related to Openness to Experience (Schutte et al., 1998).
The aim of the present study is threefold. First, an instrument for emotional intelligence
was developed and its internal structure was evaluated. The majority of studies on
emotional intelligence has relied on self-ratings. Self-report assessment of emotional
intelligence has been criticized (see e.g. Mayer and Cobb, 2000; Mayer et al., 2000;
Petrides and Furnham, 2000). Generally, the target person him or herself undoubtfully has
the greatest access to information that is relevant for judging traits and competences.
However, the relevance of such information for judging the self may be better perceived by
others than by the person him or herself (see Hofstee, 1994; Jones and Nisbett, 1972). In
addition, some of the richness of self-experience is lost as a consequence of limitations of
the judge’s working memory upon performing the self-rating. Thus self-views may be less
accurate than outsiders’ views (see e.g. Hofstee, 1994; John and Robins, 1993; Thorne,
1989). For this reason, self-ratings have often been supplemented with ratings by
knowledgeable informants such as peers and spouses (Borkenau and Liebler, 1993;
Cattell, 1946; Funder, 1989; McCrae, 1982). The present study was aimed at the
psychometric qualities of both self- and other ratings of emotional intelligence. Second,
the present study focused on the relationship of emotional intelligence with academic
intelligence and personality. Third, the incremental validity of emotional intelligence
beyond intelligence and personality in predicting both academic and social success was
examined. The central assumption underlying common sense approaches to intelligence is
that these approaches are better able to predict real life criteria than traditional measures of
academic intelligence and personality, but there is little evidence to support this claim.

A conceptual framework of emotional intelligence

It seems that two important distinctions underlie the definition of emotional intelligence by
Salovey and Mayer (1990). First, the authors refer to cognitive processes of monitoring
emotions and discriminating among them in contrast to the behavioural process of actually

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
 Emotional intelligence 105

using the information to guide one’s thinking and actions. Although the revised model of
Mayer and Salovey (1997) gives more emphasis to cognitive processes, the same
distinction can be recognized. This model distinguishes between four dimensions of
emotional intelligence: perception, appraisal, and expression of emotion; emotional
facilitation of thinking; understanding, analysing, and employing emotional knowledge;
and, finally, reflective regulation of emotions so as to promote further emotional and
intellectual growth. Particularly the elements of expression of emotion and employing
emotional knowledge seem to refer to behavioural processes, whereas perception and
appraisal of emotions, emotional facilitation of thinking and understanding, and analysing
emotional knowledge primarily concern cognitive processes.
Second, Salovey and Mayer distinguish between processes regarding one’s own
emotions and the emotions of others. In this regard, Gardner (1983) distinguished between
intrapersonal intelligence referring to the ability to recognize and label one’s own
emotions, and interpersonal intelligence, referring to the ability to understand other
individual’s emotions and intentions. Departing from the definition of Salovey and Mayer
(1990), in the present study emotional intelligence was defined as the ability to perceive
one’s own and other’s emotions, to interpret one’s own emotions and the emotions of
others, and and to cope with emotions of self and others effectively. As a theoretical
framework the dimensions of cognitive–behavioural and self–others were used to classify
subscales that fit into this definition of emotional intelligence.

Emotional intelligence and academic intelligence

The concept of intelligence can be described as a general ability, that is largely inherited,
to transform information into new concepts and skills. Analytical and emotional
intelligence are assumed to be different kinds of intelligence (Davies et al., 1998, Jones
and Day, 1997; Mayer and Salovey, 1993). What academic and emotional intelligence
have in common is that they both make use of declarative and procedural knowledge and
that they apply this knowledge in a flexible manner (Cantor and Harlow, 1994; Cantor and
Kihlstrom, 1987; Lohman, 1993). Declarative knowledge concerns understanding the
nature of specific situations, whereas procedural knowledge refers to understanding what
to do in specific situations. Flexible application of declarative and procedural knowledge
means understanding what is going on in new and unknown situations and knowing what
to do in such situations. An important difference between emotional intelligence and
analytical intelligence lies in the use and application of knowledge. Analytical intelligence
usually concerns problems for which one best solution is available (Sternberg, Wagner,
Williams and Horvath, 1995). In contrast, emotionally intelligent behaviour concerns
problems that can be interpreted in several different ways, for which no objective solution
is available and whereby different responses may lead to the desired outcome. Moreover,
no absolute standards are available to evaluate emotional intelligence: what constitutes
intelligent behaviour is determined by the environment in which the behaviour is exhibited
(Jones and Day, 1997).
Theoretically, in order to be regarded as a separate construct within the framework of
cognitive ability, emotional intelligence should be moderately, not too highly, associated
with measures of academic intelligence. Only a few studies have examined the
relationship between emotional and academic intelligence and, in general, these studies
failed to establish any support for a place of emotional intelligence within the framework
of cognitive abilities. Schutte et al. (1998) found that a measure of emotional cognition

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
106 K. van der Zee et al.

based on the model of Salovey and Mayer (1990) was unrelated to cognitive ability, as
measured by mathematical and verbal scores on the SAT. Davies et al. (1998) also studied
the relationship between indicators of emotional intelligence and academic intelligence.
Consistent with the findings of Schutte et al. (1998), they found no meaningful
relationship between emotional and crystallized intelligence in one study and only low
loadings of emotional intelligence items on a verbal factor in a second study. Finally, Van
der Zee et al. (manuscript submitted for publication) studied the relationship between
emotional intelligence and indicators of academic intelligence and found no significant
relationships. The majority of studies that focus on related concepts, such as social
intelligence (Wong, Day, Maxwell and Meara, 1995), and practical intelligence or ‘tacit
knowledge’ (Sternberg and Wagner, 1986, 1993; Sternberg et al., 1995) also failed to show
any relationship with academic intelligence. In the present study, it was tried to establish
support for a place for the concept of emotional intelligence in general taxonomies of
intelligence. Therefore, the relationship between emotional and academic intelligence was
examined.

Emotional intelligence and personality

Interestingly, although emotional intelligence is usually regarded as a construct within the
framework of intelligence, the evidence for a relationship between emotional intelligence
and personality variables is much stronger than the evidence for a relationship with
intelligence (see e.g. Boyle, Stankov and Cattell, 1995; Eysenck, 1994a, 1994b; Stankov
and Crawford, 1998). Personality can be described in terms of five basic factors, often
labelled as the ‘Big Five’ (see e.g. De Raad, 1992; Digman, 1990; John, 1990; McCrae,
1992; McCrae and Costa, 1984). The current labels for the factors are (I) Extraversion, (II)
Agreeableness, (III) Consientiousness, (IV) Emotional Stability, and (V) Intellect/
Autonomy, or Openness to Experience. In the present study, a positive relationship
between emotional intelligence and Extraversion and Agreeableness was expected. With
respect to Extraversion, Extraverts are open to others and tend to be unreserved and
informal in their contacts with other people. Probably, these characteristics are related to
what Gardner (1983) referred to as ‘interpersonal intelligence’. Several studies have
revealed a relationship between emotional intelligence and Extraversion (see e.g. Bagby,
Taylor and Parker, 1994; Davies et al., 1998; Roger and Najarian, 1989; Van der Zee et al.,
manuscript submitted for publication).
Individuals high in Agreeableness tend to be friendly and warm, tend to have respect for
others, and tend to be sensitive to other people’s wishes. Again, these characteristics will
probably be related to the cognitive and behavioural processes directed at the emotions of
others, as distinguished in our model. Although the evidence for a relationship between
emotional intelligence and Agreeableness is scarce, there are some studies in support of it
(Davies et al., 1998; Van der Zee et al., manuscript submitted for publication).
There is little reason to expect a relationship between emotional intelligence and
Conscientiousness. The carefulness, reliability, persistence, and goal-directedness of
individuals high in this trait seem to bear no conceptual resemblance to the monitoring and
interpreting of and coping with emotions. Only the study by Van der Zee et al. (2000)
reports a significant relationship between conscientiousness and emotional intelligence. It
must be noted that this study used an operationalization of conscientiousness that included
helpfulness and honesty, and thereby differed from more common approaches to the third
factor.

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
 Emotional intelligence 107

Emotionally unstable individuals are worried, easily provoked, depressive, and

vulnerable. The ability to cope with emotions seems to lie at the heart of the Emotional
Stability construct. We therefore expected a positive relationship between emotional
intelligence and Emotional Stability. Several studies have indeed supported the
relationship between emotional intelligence and Emotional Stability (e.g. Davies et al.,
1998; Roger and Najarian, 1989). Bagby et al. (1994) found a negative relationship
between scales of the Toronto Alexithymia Scale and Emotional Stability. This
alexithymia scale measures a pathological lack of emotional skills.
Finally, a positive relationship between emotional intelligence and Factor V can be
expected. The fifth factor of the Big Five is sometimes referred to as Intellect (Goldberg,
1983; Peabody, 1984). This quasi-ability factor of personality makes it somewhat similar
to intelligence. Clearly, the Intellect interpretation of Factor V strongly resembles
intelligence, or at least the words people use to describe intelligent people (e.g. smart,
bright, quick, etc) (see e.g. Ashton, Lee and Vernon, 1999). One could argue that Factor V
trait terms refer to regularities that people find to be indicative of intelligence in the
everyday life of other individuals. The evidence for a relationship between emotional
intelligence and Intellect/Autonomy is scarce (see e.g. Schutte et al., 1998; Shafer, 1999).
Shafer (1999), for example, studied the relationship between Social Intelligence and found
that the fifth factor was a strong and the most potent Big Five predictor of Practical
Cognition and that factor V was also more predictive of social intelligence than general
intelligence. To summarize, in the present study we were interested into what extent and in
what way emotional intelligence is related to the Big Five. Relationships between
emotional intelligence and four of the Big Five factors were expected: Extraversion,
Agreeableness, Emotional Stability, and Intellect/Autonomy. In order to be considered as a
separate construct, these correlations between emotional intelligence and the Big Five
should not be too high.
Emotional intelligence, academic and social success
The central assumption underlying common sense approaches to intelligence is that these
approaches are better able to predict real life criteria than traditional measures of academic
intelligence and personality. It has been suggested and found that non-academic abilities,
although they differ from the abilities that are needed to solve academic problems, are
good predictors of educational outcome measures (Feshbach and Feshbach, 1987; Ford,
1986; Safer, 1986; Schutte et al., 1998). For example, Schutte et al. (1998) showed that
scores on indices of emotional intelligence were predictive of end-of-year grade point
averages (GPA). Unfortunately, these authors make no reference of the predictive value of
scores for emotional intelligence after partialling out the effect of intelligence and
personality. There is clear empirical evidence for a relationship between intelligence and
GPA (Goff and Ackerman, 1992; Martell, Mickelvie and Standing, 1987; Rothstein,
Paunonen, Rush and King, 1994). It must be noted that most of these studies were focused
on crystallized rather than on fluid intelligence. Moreover, there are also some studies
pointing at a weak though significant relationship between personality variables such as
extraversion and conscientiousness and academic results (Goff and Ackerman, 1992). The
present study examined the incremental value of emotional intelligence above academic
intelligence and personality in predicting academic success. It was expected that, although
academic intelligence is a stronger predictor of academic success than emotional
intelligence, emotional intelligence would nevertheless be able to explain additional
variance in this success criterion.

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
108 K. van der Zee et al.

Earlier we argued that the concept of emotional intelligence particularly became

popular because it was felt that cognitive skills were required in dealing with problems that
are typically encountered in life, such as solving conflicts, collaborating with others, or
adjustment to new environments, that were not captured by traditional approaches to
intelligence. Many of these problems are social in nature and it is generally assumed that
emotional intelligence is predictive of social success and, even more so, more strongly
than academic intelligence. It is surprising how little effort has been made to prove that this
assumption is correct. McCown, Johnson and Austin (1986) present evidence that young
adults with social and criminal problems lack skills in monitoring and interpreting
emotions in others. They are not capable of interpreting facial expressions. McCown et al.
(1986) attributed this inability to a lack of emotional intelligence. Derksen (1998) revealed
a positive relationship between scores for emotional intelligence and social success. His
questionnaire contained items referring to success at work, in the family, in recreation, and
in friendships. Ford and Tisak (1983) compared the predictive value of academic
intelligence and ratings of social competence of social behaviours. They concluded that
peer-ratings of social competence obtained from friends, parents, and teachers are better
predictors of social success than academic intelligence. On the basis of these findings, it
was expected that emotional intelligence would be able to predict variance in social
success, above and beyond personality.

METHOD

Participants
The present findings were based on a sample of 116 students, who were approached for the
study at a number of faculties. The majority of the group were students from the Faculty of
Arts (41%), 28% studied Psychology, 28% studied Business Administration and
Economics; 22% were involved in other studies. The age of the sample varied between
18 and 32 years (M ¼ 22.8; SD ¼ 2.7). The majority of the respondents were female (72%).
All students who participated in the study received feedback regarding their personal test
results.

Instruments
First, as indicators of Academic Intelligence the General Aptitude Test-Battery (GATB;
US Department of Labor, 1970) was used. Six subscales from the Dutch version of the
General Aptitude Test-Battery were included. Participants’ verbal abilities were measured
by scales for Name Comparisons (M ¼ 5.94, SD ¼ 2.40) and Vocabulary (M ¼ 3.82,
SD ¼ 2.26), numerical abilities by scales for Elementary Arithmetic (M ¼ 3.00, SD ¼ 2.06)
and Verbal Arithmetic (M ¼ 3.02, SD ¼ 1.93) and, finally, figural/spatial abilities were
measured by scales for Three-Dimensional Space (M ¼ 3.85, SD ¼ 2.03) and Figure
Detection (M ¼ 4.05, SD ¼ 2.00). Departing from the theory of Horn and Catell (1966),
empirical data suggest that the scales for vocabulary, elementary arithmetic, and verbal
arithmetic can be considered as crystallized intelligences, whereas the scales for figure
detection and three-dimensional space belong to the fluid intelligences. Finally, the name
comparison test seems to call on both intelligences (Van der Flier and Boomsma-Suerink,
1994). The GATB is a speeded test (45 minutes are allowed for the six subscales). For all
scales, decile (norm) scores were computed.

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
 Emotional intelligence 109

Personality
Personality was measured by the Connector-P3 (141 items), developed by PiMedia
(t Mannetje and Schattenberg, 1996). The questionnaire has scales for the Big Five factors
Extraversion, Agreeableness, Conscientiousness, Emotional Stability, and Autonomy/
Intellect (see e.g. De Raad, 1992; Digman, 1990; Goldberg, 1990; John, 1990). The
corresponding Connector-P scales have items that are adjusted to the work context and are
referred to as Extraversion, Respect, Work attitude, Stability, and Openness. For each
personality dimension, items referred to concrete behaviours or tendencies that are
relevant to work settings. Participants could give their answers on a five point scale ranging
from not applicable to me (1) to entirely applicable to me (5). The number of items per
scale varied between 18 and 32. It took approximately 30 minutes to complete the
questionnaire. For each personality factor t-scores were computed. Cronbach’s alpha’s for
the different scales varied between 0.77 and 0.84.
Emotional Intelligence
The instrument for Emotional Intelligence contained 17 scales each consisting of five
items. Subdimensions were derived from the literature and were classified according to the
cognitive–behavioural and self–other distinction. Each scale item consisted of two
opposite behavioural descriptions referring to behaviours that were assumed to be
indicative of the specific dimension. Participants could indicate on a five-point scale which
position between the two behavioural poles best described the focal person’s behaviour.
Three scales were included for monitoring and interpreting one’s own emotions. First, the
scale for Emotional Self-Consciousness referred to being open to one’s emotions,
understanding one’s emotions, and recognizing the effects of emotions on one’s
performance. A sample item of this scale is ‘Takes emotions into consideration in
decision making’ ( þ ) versus ‘Is guided by rational arguments in making decisions’ (
).
The scale for Self-Reliance indicated trust in one’s own appearance and confidence in
one’s capabilities, for example ‘Emphasizes own strengths’ ( þ ) versus ‘Talks mostly
about own weaknesses’ (
). Positive Thinking referred to a tendency to focus on positive
aspects of life, for example ‘Speaks with enthusiasm about his/her work’ ( þ ) versus
‘Speaks primarily about the negative aspects of his/her work’ (
).
Dealing with one’s own emotions was measured with scales for Stress Management,
Autonomy, Adaptation, Self-Motivation, Self-Development and Self-Control. Stress
Management was defined as regulating one’s emotions in order to prevent them from
interfering with performance and effective performance in stressful circumstances, and
was measured by items such as ‘Easily recovers from setbacks’ ( þ ) versus ‘Has
difficulties recovering from setbacks’ (
). Autonomy indicated a tendency not to be
emotionally influenced by the environment and to undertake actions that are based on
one’s own beliefs (e.g. ‘Works independently and autonomously’ (
) versus ‘Seeks
confirmation of others’ (
)). Adaptation refers to reconciling one’s emotions with
changing situations and conditions, for example ‘Easily adjusts to new work situations’
( þ ) versus ‘Has difficulties in adjusting to new work situations’ (
). Self-Motivation was
defined as regulating one’s own emotions in order to reach goals. A sample item is ‘Works
because he or she enjoys it’ ( þ ) versus ‘Is motivated for work because of its financial
rewards’ (
). Self-Development referred to insight into one’s own emotional strengths

3
For information about the instrument for Emotional Intelligence that was developed in this study and about the
Connector-P, please contact PiMedia, PO Box 1087, 3600 BB Maarssen, The Netherlands. Tel: þ 31 346 559010.
Fax: þ 31 346 559015. E-mail: info@pimedia.nl, internet: www.pimedia.nl

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
110 K. van der Zee et al.

and weaknesses and a tendency to undertake actions to improve one’s performance

(e.g. ‘Frequently asks for feedback from coworkers’ ( þ ) versus ‘Shows no interest in how
his/her behaviour is observed by others’ (
)). Finally, the scale for Self-Control was
intended to measure the ability to control one’s emotions in order to retain a maximum
level of performance. A sample item of this scale is ‘Is able to concentrate on work without
being distracted by his or her emotions’ ( þ ) versus ‘Is easily distracted’ (
).
Perceiving and interpreting the emotions of others was measured by scales for Affective
Listening, Empathy, and Understanding Human Character. Affective Listening means
showing verbally and nonverbally that one has noticed important emotional information in
the messages other people communicate and being able to react effectively to this
information, for example ‘Pays attention to mood and relations during conversations’ ( þ )
versus ‘Pays attention to content during conversations, is business-wise’ (
). The scale
for Empathy referred to being conscious of the emotions of others and showing
appreciation of those emotions (e.g. ‘Senses if matters are delicate’ ( þ ) versus ‘Does not
sense if matters are delicate’ (
)). Understanding Human Character indicated the ability
to sense the nature and motives of other. A sample item is ‘Is able to assess if the intentions
of others are good ( þ )’ versus ‘Has difficulties assessing other people’s motives’ (
).
Finally, the ability to deal with the emotions of others was measured by the scales for
Commitment, Team Skills, Influencing Others, Conflict Management, and Development
of Others. Commitment was defined as expressing concern for the benefit of others and
actively helping others, for example ‘Inquires after the progression of projects of
colleagues and helps if necessary’ ( þ ) versus ‘Shows no interest in projects of colleagues’
(
). Team Skills referred to contributing actively to team goals, even when these goals do
not coincide with personal goals (e.g. ‘Enhances team spirit by paying attention to mutual
relationships’ ( þ ) versus ‘Mainly concentrates on task-related outcomes of the team’
(
)). The scale for Influencing Others indicated the ability to influence opinions of others
by using the right tactics and arguments. A sample item is ‘Presents arguments that others
are susceptible to’ ( þ ) versus ‘Presents arguments that are based on his/her own frame of
reference’ (
). Conflict Management refers to an ability to prevent and to solve conflicts
and is measured by items such as ‘Takes the time to listen to the arguments of the
counterparty and involves these arguments in his/her proposals’ ( þ ) versus ‘Makes
proposals to the counterparty that are based on his/her own arguments’ (
). Development
of Others indicates motivating others to enhance their performance (e.g. ‘Gives positive
feedback on high performance of colleagues’ ( þ ) versus ‘Is only responsive to low
performance of colleagues’ (
)). It took participants approximately 30 minutes to
complete the questionnaire.

Academic and social success

Academic Success was indicated by students’ mean grade and the number of grade points
in the present year and in their first year. As an indication of study pace, students were
asked when they had started their studies, whether they were on schedule, and when they
expected to graduate. Mean grade, grade points, and study pace were combined into one
success indicator by computing decile scores on the three measures (departing from the
score distribution in the present sample) and multiplying them. The decile scores for time
were recoded, in order that a high score indicated high success (fast progression).
Social success was measured by a questionnaire for social activities and a scale for
social competence. Five items were included for social activities. These items for example
referred to being a member of student organizations, committees, or being engaged in

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
 Emotional intelligence 111

voluntary work, and participants were asked to answer with yes or no to the question if
they were involved in such social activities or not. Participants received one point if they
had been involved in a particular social activity; they received two points if they were for
example members of two or more student organizations. A total score was obtained by
summing up across activities, resulting in a scale of modest reliability (
¼ 0.60). Second,
a five-item social competence scale was included, defined as the ability to establish and
maintain contacts with other people. Each scale item consisted of two opposite
behavioural descriptions referring to behaviours that were assumed to be indicative of
the specific scale. Participants could indicate on a five-point scale which position between
the two behavioural poles best described the focal person’s own behaviour, for example
‘Starts a conversation with another person, even if he/she has never met this person before’
( þ ) versus ‘Waits until others approach him/her to start a conversation’. Both self- and
other ratings on this scale were collected. Cronbach’s alpha of this scale was 0.83 for self-
ratings and 0.87 for other ratings. Self- and other ratings were highly interrelated
(r ¼ 0.64). Self- and other ratings of social competence were moderately related to the
social activities scale, r ¼ 0.21, p < 0.05 and r ¼ 0.24, p < 0.05, respectively for self- and
other ratings, and unrelated to the measure of academic success.
Procedure
Upon arrival in the laboratory, students received a short introduction on the study
purposes. Next, they received a short questionnaire that contained indicators of academic
and social success. This questionnaire was followed by computerized intelligence tests and
a personality questionnaire. In order to control for fatigue symptoms, the order of the
intelligence tests was varied. Next, students filled out the personality questionnaire. At the
end of the session, students received five questionnaires that included scales for emotional
intelligence together with an envelope. The students were instructed to return their
completed questionnaire together with preferably four additional questionnaires that were
filled out by persons close to them. The students were free in choosing the persons that
provided the other ratings, but received the instruction to select persons from different
social networks, such as room mates, sport mates, family members, and close friends. In
the instruction to the participating others, it was stressed that their evaluations had no
consequences for the assessed person and that the ratings would be treated confidentially
and would not be revealed to the person they had assessed.

RESULTS

Internal structure of the emotional intelligence scales

Table 1 reveals the internal consistencies of self- and other ratings on the 17 scales. Other
ratings were obtained by computing mean item scores across the four raters from whom
other ratings were obtained. Not surprisingly, on the whole these other ratings were more
reliable than self-ratings. Self-ratings were insufficiently reliable: only the scale for
positive thinking had an alpha value exceeding 0.70. The other ratings were also of
moderate reliability, with the exception of the scales for positive thinking, autonomy, stress
management, empathy and involvement. In the majority, both for self- and for other
ratings, scale means were clearly above the midpoint of the scale, pointing at possible
social desirability bias. A MANOVA revealed no significant differences between scale
scores based on self- versus other ratings, except for the scale for conflict management.

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
112 K. van der Zee et al.

Table 1. Means, standard deviations and scale reliabilities for self- and other ratings on the
emotional intelligence scales
Self-ratings Other ratings

M SD
M SD

1. Emotional Self-Consciousness 3.50 0.60 0.53 3.44 0.95 0.65

2. Self-Reliance 3.01 0.74 0.57 3.01 0.46 0.52
3. Positive Thinking 3.59 0.47 0.74 3.53 0.33 0.82
4. Stress Management 3.30 0.80 0.64 3.35 0.51 0.80
5. Autonomy 3.64 0.68 0.65 3.58 0.46 0.78
6. Adaptation 3.62 0.63 0.65 3.53 0.39 0.72
7. Self-Motivation 3.74 0.64 0.54 3.68 0.48 0.67
8. Self-Development 3.67 0.56 0.37 3.69 0.34 0.48
9. Self-Control 3.50 0.72 0.60 3.48 0.47 0.59
10. Affective Listening 3.76 0.41 0.56 3.74 0.61 0.77
11. Empathy 4.04 0.64 0.59 4.02 0.36 0.80
12. Understanding Human Character 3.61 0.62 0.59 3.60 0.36 0.66
13. Commitment 3.80 0.62 0.65 3.78 0.41 0.81
14. Team Skills 3.45 0.65 0.31 3.43 0.47 0.70
15. Influencing Others 3.23 0.47 0.17 3.26 0.29
0.07
16. Conflict Management 3.60 0.40 0.25 3.67 0.33 0.64
17. Development of Others 3.62 0.50 0.36 3.59 0.35 0.67

In order to examine the internal structure of the measure scale intercorrelations were
computed on the level of the individual ratings. Table 2 reveals the resulting correlations.
As can be derived from this table, the pattern of intercorrelations did not support our a
priori classification of the scales into cognitive versus behavioural and self versus others.
Intercorrelations between scales with a similar position on both dimensions did not exceed
correlations between scales from different quadrants. In addition, an exploratory factor
analysis was performed, that also failed to support the hypothesized two-dimensional
classification. The scree plot pointed at a three factor structure. Together, these three
factors explained 28.03% of variance (eigenvalue > 4)4. After oblique rotation, this
solution appeared to be readily interpretable. Table 3 shows sample items with the highest
loadings on each factor. On the first factor, high loadings were found for items that referred
to being sensitive to the emotions of others and expressing concern towards others.
Particularly items from the scales for affective listening, empathy, and commitment
showed high loadings on these factors. This factor was labelled Empathy. High loadings on
the second factor were found for items referring to autonomous and self-motivational
behaviours and cognitions, and this factor will be further referred to as Autonomy.
The third factor could be interpreted as lack of Emotional Control. Both items referring to
keeping calm and retaining a high level of performance under stress and items
referring to dealing with stress or conflict in the social environment are included in this
factor.
Scales for each factor were formed by selecting items with the highest loading on the
specific factor, provided that this loading exceeded 0.35. Factor analyses on the selected

4
Coefficient alpha is a simple function of its eigenvalue (Kaiser and Caffrey, 1965; see also Hofstee, Ten Berge
and Hendriks, 1998):
¼ [n/(n
1)][1
1/Ej], whereby n represents the number of components and Ej the jth
eigenvalue ( j ¼ 1, . . . , m). An eigenvalue of 4 corresponds to a reasonable alpha of 0.76.

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
 Table 2. Intercorrelations between the emotional intelligence scales
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

1. Emotional Self-
0.17**
0.10*
0.16** 0.07
0.16** 0.18**
0.25** 0.27** 0.50** 0.54** 0.30** 0.44** 0.34** 0.12** 0.38** 0.38**
Consciousnessa
2. Self-Reliance 0.33** 0.39** 0.20** 0.49** 0.19** 0.18**
0.03
0.20**
0.14** 0.11
0.12*
0.12* 0.10*
0.07
0.13**
3. Positive Thinking 0.45*** 0.39** 0.29*** 0.41*** 0.20** 0.34** 0.26** 0.26** 0.37** 0.31** 0.27** 0.20** 0.29** 0.39**
4. Stress Management 0.32** 0.34** 0.40** 0.42** 0.22** 0.12* 0.07 0.43** 0.10* 0.22** 0.08 0.27 0.12
5. Autonomy 0.35** 0.31** 0.43** 0.32** 0.21** 0.20** 0.30** 0.26** 0.26** 0.03 0.26** 0.26**

Copyright # 2002 John Wiley & Sons, Ltd.

6. Adaptation 0.20** 0.28** 0.15**
0.13**
0.13** 0.14**
0.09
0.14** 0.09
0.04
0.12*
7. Self-Motivation 0.06 0.32** 0.34** 0.31** 0.43** 0.36** 0.39** 0.28** 0.39** 0.25**
8. Self-Development 0.18** 0.06 0.00 0.24** 0.02 0.09
0.06 0.13** 0.01
9. Self-Control 0.42** 0.32** 0.34** 0.38** 0.46** 0.16** 0.48** 0.38**
10. Affective Listening 0.67** 0.56** 0.70** 0.64** 0.15** 0.67** 0.50**
11. Empathy 0.51** 0.59** 0.51** 0.22** 0.55** 0.51**
12. Understanding 0.46*** 0.49** 0.20** 0.61** 0.39**
Human Character
13. Commitment 0.65** 0.19** 0.56** 0.52**
14. Team Skills 0.16** 0.63** 0.55**
15. Influencing Others 0.20** 0.21**
16. Conflict Management 0.46**
17. Development of Others
a
Scales 1–4 refer to the monitoring and interpretation of one’s own emotions, scales 5–9 to the ability to deal with one’s own emotions. Scales 10–13 concern the ability to accurately
perceive and interpret the emotions of others, scales 14–17 to the ability to deal with the emotions of others.
Significance level *p < 0.05, **p < 0.01.
Emotional intelligence
113

Eur. J. Pers. 16: 103–125 (2002)

114 K. van der Zee et al.

Table 3. Sample items with the highest loadings on the emotional intelligence factors

I. Empathy (31 items)

Shows understanding in response to emotional reactions 0.66
Keeps asking questions when others are reluctant to talk about their feelings 0.63
Takes his/her time for a colleague who expresses a need to talk 0.61
II. Autonomy (19 items)
Expects to be considered for an aspired project 0.63
Works independently and autonomously 0.59
Relies on his/her own opinion in making a decision 0.59
III. (Lack of ) Emotional Control (13 items)
Inhibits emotional outbursts such as anger or tears
0.60
Reacts calmly to criticism
0.50
Is able to keep exciting news to him/herself
0.47

Table 4. Means, standard deviations, and scale reliabilities for self- and other ratings on the
emotional intelligence dimensions
M SD
2 3

Self-ratings
1. Empathy 3.91 0.39 0.83 0.11 0.04
2. Autonomy 3.36 0.51 0.84 0.46**
3. Emotional control 3.52 0.47 0.69
Other ratings
1. Empathy 3.70 0.39 0.92
0.10 0.14
2. Autonomy 3.46 0.40 0.90 0.38**
3. Emotional control 3.50 0.42 0.81
Significance levels *p < 0.05, **p < 0.01.

item set supported the three factor solution (eigenvalue > 3, 32.29% explained variance).
The pattern matrix revealed a factor solution that perfectly fitted the a priori allocation of
the items to the three factors. Table 4 shows reliabilities and scale intercorrelations of the
scales that were formed by adding up the selected items, both for self- and for other
ratings. The scale score corresponding to the (lack of ) emotional control factor was coded
in the positive direction. High scores corresponded to high emotional control. For the self-
ratings, the internal consistencies of the scales were moderate to high, with a modest alpha
for emotional control. Reliabilities for the other ratings were high: all alpha’s exceeded
0.80. The emotional intelligence dimensions were moderately interrelated.
Next, we were interested in the correspondence between self- and other ratings. Table 4
shows means for self- and other ratings. MANOVA with rater as the within-subject
independent variable, and the emotional intelligence dimensions as the dependent
variables, revealed significantly higher scores for self-ratings as compared with other
ratings for empathy, F(1, 85) ¼ 32.52, p < 0.001, and significantly lower self-ratings for
autonomy, F(1, 85) ¼ 4.13, p < 0.05. No significant difference was found between self-
and other ratings of emotional control. Both for self- and for other ratings, the present data
revealed elevated scale means. Possibly, both self- and other ratings suffered from social
desirability bias. Table 5 shows the correlations between self- and other ratings. All the
values on the principal diagonal exceed the so-called 0.3 barrier (e.g. McCrae and Costa,
1989), pointing at the validity of self- versus other ratings.

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
 Emotional intelligence 115

Table 5. Correlations between self- and other ratings on the emotional intelligence dimensions
Other ratings

1 2 3

Self-ratings
1. Empathy 0.65**
0.06 0.00
2. Autonomy 0.01 0.59** 0.30**
3. Emotional Control
0.12 0.22* 0.56**
Significance levels *p < 0.05, **p < 0.01.

Table 6. Results of stepwise regression of the combined ratings of emotional intelligence on

academic intelligence
R R2  step 1 r

Self-ratings
Empathy
Name Comparisons 0.30 0.09
0.30**
0.30**
Autonomya
Emotional Control
Vocabulary 0.24 0.06
0.24**
0.24**
Other ratings
Empathy
Vocabulary 0.24 0.06
0.24**
0.24**
Autonomy
Figure Detection 0.22 0.05 0.22* 0.22*
Emotional Controla
a
None of the academic intelligence dimensions revealed significant beta-weights.
Significance levels *p < 0.05, **p < 0.01.

Emotional intelligence and academic intelligence

In order to examine the relationship of the emotional intelligence dimensions with
academic intelligence, stepwise regressions were performed with the different test scores
for academic intelligence as independent variables and the emotional intelligence
dimensions as dependent variables. Table 6 shows the results of these regression analyses.
As expected, academic intelligence was only modestly related to emotional intelligence,
with explained variances of 9% or less. Unexpectedly, in most cases negative correlations
between indicators of academic and emotional intelligence were found. Only for the scale
for figure detection that did show a positive relation with autonomy, findings were in the
expected direction.

Emotional intelligence and personality

Next, we were interested in the relationship between emotional intelligence and the Big
Five personality factors. Stepwise regressions were performed with the Big Five as
independent variables and the emotional intelligence dimensions as dependent variables.
Table 7 shows the result of these regression analyses. In line with earlier findings, as
compared with academic intelligence, the Big Five were far more predictive of emotional
intelligence. Empathy seems to be primarily a combination of extraversion and respect.
Emotional control seems to be a mixture of the Big Five, with stability as its best predictor.

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
116 K. van der Zee et al.

Table 7. Results of stepwise regression of self- and other ratings of emotional intelligence on the
Big Five
R R2 step 1 step 2 step 3 step 4 step 5 r

Self-ratings
Empathy
step 1 Extraversion 0.39 0.15 0.39** 0.42** 0.39**
step 2 Respect 0.53 0.28 0.35** 0.32**
Autonomy
step 1 Stability 0.69 0.48 0.69** 0.47** 0.70**
step 2 Openness 0.77 0.59 0.40** 0.67**
Emotional Control
step 1 Stability 0.53 0.28 0.53** 0.65** 0.65** 0.72** 0.62** 0.53**
step 2 Extraversion 0.60 0.37
0.32**
0.31**
0.35**
0.43**
0.08
step 3 Respect 0.65 0.42 0.23* 0.23* 0.26** 0.22*
step 4 Work Attitude 0.68 0.47
0.23*
0.19*
0.09
step 5 Openness 0.70 0.49 0.23* 0.31**
Other ratings
Empathy
step 1 Extraversion 0.32 0.10 0.32** 0.24* 0.39** 0.32**
step 2 Respect 0.37 0.14 0.21* 0.27** 0.30**
step 3 Stability 0.43 0.19
0.29** 0.06
Autonomy
step 1 Openness 0.44 0.20 0.69** 0.47** 0.44**
step 2 Respect 0.53 0.28
0.31** 0.03
Emotional Control
step 1 Stability 0.29 0.09 0.29** 0.53** 0.64** 0.29**
step 2 Extraversion 0.42 0.18
0.38**
0.40**
0.06
step 3 Work Attitude 0.46 0.22
0.22* 0.00
Significance levels *p < 0.05, **p < 0.01.

It must be noted that for extraversion and work attitude, scales for which negative loadings
were found, the zero-order correlations with empathy were not significant. Therefore,
these beta-weights should be noted with caution. Not surprisingly, for other ratings it was
found that autonomy was best predicted by the fifth factor of the Big Five, that is often
referred to as Autonomy/Intellect (De Raad, Hendriks and Hofstee, 1992; Goldberg, 1990).
For self-ratings, however, emotional stability appeared equally predictive (and even
slightly more predictive) of autonomy. Although the Big Five explained a significant
amount of variance in emotional intelligence, it seems not warranted to infer from the
present data that emotional intelligence is nothing more than a bunch of socially desirable
traits. The Big Five explained most variance in autonomy (59% and 28% for self- and
other ratings, respectively).

Emotional intelligence and academic and social success

To examine the concurrent validity of the emotional intelligence scales, stepwise
regressions were performed of four success criteria on the emotional intelligence
dimensions. The combined measure of mean grade, number of grade points, and study
time was used as an indicator of academic success; the social activities scale as well as
self- and other ratings of social competence were used as indicators of social success. With
respect to academic success, self-rated emotional intelligence appeared to be unrelated to
academic performance. By contrast, autonomy as rated by others did appear to be a

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
 Emotional intelligence 117

Table 8. Results of stepwise regression of academic success on self- and other ratings of emotional
intelligence
R R2  step 1  step 2  step 3 r

Self-ratings
Social Activities
step 1 Empathy 0.29 0.08 0.29* 0.29*
Social Competence (self-ratings)
step 1 Empathy 0.58 0.32 0.58** 0.53** 0.54** 0.58**
step 2 Autonomy 0.67 0.44 0.35** 0.44** 0.41**
step 3 Emotional Control 0.69 0.48
0.20* 0.02
Social Competence (other ratings)
step 1 Empathy 0.46 0.21 0.46** 0.46**
Other ratings
Social Activities
step 1 Empathy 0.29 0.03 0.29*
Social Competence (self-ratings)
step 1 Empathy 0.39 0.16 0.39** 0.42** 0.49** 0.39**
step 2 Autonomy 0.54 0.29 0.36** 0.53** 0.33**
step 3 Emotional Control 0.64 0.41
0.39*
0.14
Social Competence (other ratings)
step 1 Empathy 0.54 0.29 0.54** 0.59** 0.65** 0.54**
step 2 Autonomy 0.69 0.47 0.43** 0.55** 0.36**
step 3 Emotional Control 0.74 0.56
0.20*
0.06
Significance levels *p < 0.05, **p < 0.01.

significant predictor of academic success ( ¼ 0.26, p < 0.05). As predicted, emotional

intelligence was more strongly associated with social success than with academic success,
with empathy as the best predictor of both social competence and engagement in social
activities. Table 8 reveals the results of the regression analyses of social competence on the
three scales for emotional intelligence. Not surprisingly, emotional intelligence was less
predictive of social competence if the results were obtained from different raters, but in all
cases particularly empathy and autonomy appeared to be rather strong predictors of social
competence.
Finally, because the interest of the present study was in the incremental validity of
emotional intelligence in predicting success in academic and social life above traditional
measures of academic intelligence and personality, stepwise regressions were performed
of the indicators of academic and social success on the emotional intelligence dimensions,
after controlling for the Big Five dimensions and academic intelligence (Table 9). The Big
Five dimensions were entered in the regression in the first step, the six subscales of the
GATB in the second step. In the third step, the emotional intelligence dimensions were
entered in a stepwise manner. In line with our predictions, emotional intelligence
explained a significant amount of variance above intelligence and personality in both
social competence and academic success. Personality appeared to be a stronger predictor
of academic success than academic intelligence. The higher students’ work motivation, the
more successful they were at university. With respect to social activities it was found that
personality was about equally predictive of this criterion as emotional intelligence
(Tables 8 and 9). The autonomy dimension was nevertheless able to explain independent
variance in social activities. Extraversion was very strongly related to social competence:
this trait explained respectively 48% and 32% of variance in self- and other rated social
competence. Again, emotional intelligence was nevertheless able to explain additional

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
118 K. van der Zee et al.

Table 9. Results of hierarchical regression of social success on emotional intelligence, controlled

for the Big Five (step 1) and academic intelligence (step 2)
R R2 step 1 step 2 step 3 step 4 r

Self-ratings
I. Academic Success
step 1 Work Attitude 0.41 0.17 0.41** 0.38** 0.41**
step 2a Empathy 0.46 0.21 0.22* 0.27**
II. Social Activities
step 1 Work Attitude 0.30 0.09
0.30**
0.31**
0.30**
step 2a Autonomy 0.45 0.20 0.33** 0.29**
II. Social Competence
(self-ratings)
step 1 Extraversion 0.66 0.43 0.66** 0.62** 0.52** 0.43** 0.69**
step 2 Name 0.71 0.49
0.26**
0.18*
0.17*
0.35**
Comparisons
step 3 Empathy 0.77 0.59 0.34** 0.38** 0.58**
step 4 Autonomy 0.79 0.63 0.20* 0.41**
III. Social Competence
(other ratings)
step 1 Extraversion 0.56 0.32 0.56** 0.45** 0.56**
step 2a Empathy 0.65 0.42 0.34** 0.46**
Other ratings
I. Academic Success
step 1 Work Attitude 0.31 0.10 0.31** 0.30** 0.31**
step 2a Autonomy 0.41 0.17 0.27** 0.26**
II. Social Activities
step 1 Work Attitude 0.28 0.08
0.28**
0.28**
0.34**
0.37**
0.28**
step 2 Vocabulary 0.43 0.18
0.33**
0.48**
0.43**
0.33**
step 3 Name 0.54 0.29 0.37** 0.39** 0.10
Comparisons
step 4 Autonomy 0.45 0.20 0.33** 0.26** 0.26**
II. Social Competence
(self-ratings)
step 1 Extraversion 0.69 0.48 0.69** 0.76** 0.72** 0.68** 0.69**
Respect 0.71 0.50
0.17*
0.18*
0.21* 0.14
step 2 Name 0.71 0.56
0.24**
0.21*
0.36**
Comparisons
step 3 Empathy 0.77 0.58 0.18* 0.39**
III. Social Competence
(other ratings)
step 1 Extraversion 0.56 0.32 0.56** 0.44** 0.30** 0.23* 0.56**
step 2 Empathya 0.68 0.46 0.41** 0.49** 0.56** 0.54**
step 3 Autonomy 0.74 0.55 0.33** 0.45** 0.36**
step 4 Emotional Control 0.77 0.59
0.23*
0.06
a
None of the academic intelligence dimensions revealed significant beta-weights.
Significance levels *p < 0.05, **p < 0.01.

variance in social success, with empathy and, in the case of other rated social competence,
also autonomy as significant predictors.
To conclude, the expectation that emotional intelligence will predict variance in
academic and social success, above personality and intelligence, was supported by the
data. It must be noted that the incremental validity of emotional intelligence was not very
high. Personality appeared to be a rather strong predictor of both social and academic
success. Extraversion was strongly related to social competence, work attitude to

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
 Emotional intelligence 119

academic success, and (negatively) to social activities. Academic intelligence was not a
very strong predictor of success.

DISCUSSION

The concept of emotional intelligence has evoked considerable debate, for example
leading Davies et al. in their 1998 paper in the Journal of Personality and Social
Psychology to present their work under the title ‘Emotional intelligence in search of an
elusive construct.’ The present study examined the predictive and construct validity of this
elusive but also appealing construct. Departing from the definition of Salovey and Mayer
(1990), in the present study emotional intelligence was defined as the ability to perceive
one’s own and other’s emotions, to interpret the own emotions and the emotions of others,
and to cope with the emotions of self and others effectively. This definition reflects on the
one hand the distinction between the cognitive process of monitoring and interpreting
emotions and the behavioural process of reacting effectively in situations that are
emotionally challenging, and, on the other hand, the distinction between emotions of self
and others. The data provided little support for the a priori distinction between cognitive
and behavioural aspects of intelligence and the classification of aspects of intelligence into
aspects that are self-directed and aspects that are directed at others.
Factor analysis on the items that were developed departing from this classification
resulted in three factors, which were labelled as empathy, autonomy, and emotional
control. Empathy refers to the ability to recognize emotional states of others and to show
sensitivity to these emotions. Related conceptualizations of this emotional intelligence
dimension concern for example ‘interpersonal intelligence’ (Gardner, 1983) or
‘interpersonal skills’ (Bar-On, Brown, Kirkcaldy and Thomé, 2000). Autonomy was
defined as self-motivational and autonomous behaviours and cognitions (cf. ‘motivational
factors’, Bar-On et al., 2000). Emotional control refers to keeping calm and retaining a
high level of performance under stress and to dealing with stress or conflict in the social
environment (cf. ‘stress management’, Bar-On et al., 2000). The dimensions included both
behaviours and the preceeding cognitions, and although empathy particularly seems to
refer to the emotions of others, and autonomy to one’s own emotions, the dimension of
emotional control includes both regulating one’s own emotions and regulating the
emotions of others.
The present data included both self- and other ratings of emotional intelligence. In the
present study, the reliability of the aggregated other ratings of emotional intelligence
dimensions based on judgments of four different raters appeared to be higher than self-
ratings of the same dimensions. This is in line with the Spearman–Brown formula (Brown,
1910; Spearman, 1910), that predicts that the reliability of a scale is a function of the
number of raters. The intercorrelations between self- and other ratings were reasonably
high, with all the values on the principal diagonal exceeding the so-called 0.3 barrier
(see e.g. McCrae and Costa, 1989). Self-ratings on empathy were higher than other ratings.
This finding is consistent with findings from earlier studies that have compared self- and
other ratings, and seems to be due to the fact that individuals tend to present a view of
themselves that is positively biased. However, the opposite findings were obtained for
autonomy, with other ratings exceeding self-ratings. Both self- and other ratings of
emotional intelligence were clearly above the scale mean, suggesting a possible influence
of social desirability bias on both types of rating. It has indeed been suggested in the

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
120 K. van der Zee et al.

literature that self- and other ratings may suffer from the same systematic biases
(see Cronbach, 1955; Funder, 1999). Leniency bias in external judgments may either
spring from favourable attitudes towards other individuals or from social desirability
(see Nunnally, 1978). However, the fact that factor analysis revealed three factors rather
than one general factor suggests that the present findings cannot be fully explained away
by social desirability bias.
One of the basic questions of the present study concerned the relationship between
emotional intelligence on the one hand and academic intelligence and personality on
the other hand. Theoretically, in order to be regarded as a separate construct within the
framework of cognitive ability, emotional intelligence should be moderately, not too
highly, associated with measures of academic intelligence and should not coincide with
indicators of personality. Academic intelligence was weakly and inconsistently related to
emotional intelligence. This pattern of inconsistent findings revealing insignificant or even
negative correlations between academic intelligence and emotional intelligence is in line
with earlier studies that focused both on crystallized and fluid intelligence (e.g. Davies
et al., 1998). Interestingly, in the present study, the intelligence scale for figure
detection that did reveal a positive relationship concerned fluid rather than crystallized
intelligence.
The findings with respect to academic intelligence do not warrant conclusion that the
present indicator of emotional intelligence can be considered as a cognitive ability. Not
surprisingly, emotional intelligence was more strongly related to personality than to
academic intelligence (see also Davies et al., 1998; Schutte et al., 1998; Van der Zee et al.,
manuscript submitted for publication). Conceptually, the three factors closely resemble the
descriptions of the Big Five, with empathy corresponding with Agreeableness and
Extraversion, emotional control with Emotional Stability and autonomy with Intellect/
Autonomy. Indeed, examining the correlations between the five dimensions for emotional
intelligence and the Big Five, empathy was best explained by a combination of
extraversion and respect, whereas autonomy was strongly predicted by the fifth factor of
the Big Five. The relationship with the Big Five was not perfect: autonomy was also
strongly related to (self-rated) emotional stability and emotional control was also related to
the remaining Big Five factors. Moreover, although the Big Five explained a significant
amount of variance in emotional intelligence, it seems not warranted to infer from the
present data that emotional intelligence is nothing more than the Big Five in disguise.
Unexpectedly, the relationship between emotional intelligence and personality revealed
relationships with all the Big Five factors. In this regard, Hofstee’s (2001) notion of the p-
factor seems of relevance. Hofstee argues that, departing from personality traits for the Big
Five scored in the positive direction, a first principal component (p-factor) can be derived
that can be interpreted as ‘competence’ or ‘coping’, in the sense of adequacy of reaction to
situations. According to Hofstee, in this p-factor ‘stylistic intelligence and other traits
seem to be united in an evolutionary function: adequate reaction to reality leads to
survival’ (Hofstee, 2001). Although this definition of the p-factor is not restricted to
emotional adaptation, it comes rather close to conceptions of emotional intelligence.
Unfortunately, the present measure of personality did not allow us to perform a
comprehensive test of the relationship between the p-factor and emotional intelligence.
Future studies may depart from the abridged Big Five circumplex model as developed by
De Raad et al. (1992). Although the present instrument for the Big Five was clearly based
on the interpretations of the main factors by De Raad et al. (1992), the item pool did not
comprise a comprehensive representation of the Big Five circumplex model.

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
 Emotional intelligence 121

In addition to relationships that are most commonly reported in the literature

between emotional intelligence and the Big Five factors that are associated with
interpersonal behaviour (extraversion and agreeableness) and emotional stability (see e.g.
Davies et al., 1998; Schutte et al., 1998; Van der Zee et al., manuscript submitted for
publication), the present study also revealed a relationship between autonomy and the
fifth Big Five factor. This finding is in line with our a priori predictions. The fifth
factor, which can be considered as a quasi-ability factor of personality, is sometimes
referred to as Intellect (Goldberg, 1983; Peabody, 1984), and the Intellect interpretation of
Factor V strongly resembles common sense characterizations of intelligent people. Shafer
(1999) studied the relationship between the Big Five and an indicator of Social
Intelligence and found that the fifth factor was a strong and the most potent Big Five
predictor of social intelligence. Shafer concludes that his results provide support for the
prediction that Factor V trait terms summarize behaviours that shape implicit conceptions
of intelligence.
Why does the literature consistently show stronger relationships of emotional
intelligence with personality than with academic intelligence? The difference between
indicators of intelligence and personality is that the former refer to ‘maximum
performance’, whereas the latter refer to ‘typical performance’ (Cronbach, 1949; see
also Hofstee, 2001). Examining the items of the present instrument, it seems that they are
fairly similar to personality items in that they seem to capture stylistic rather than ability-
related aspects of behaviour. Take for example the item ‘Is able to concentrate on work
without being distracted by his or her emotions’ ( þ ) versus ‘Is easily distracted’ (
).
Although the phrasing of the item explicitly refers to an ‘ability’, being easily distracted
seems to refer to typical rather than maximum behaviour. Data showing that performance
measures of emotional intelligence correlate more strongly with traditional cognitive
abilities than self-report scales support this assumption (see e.g. Mayer et al., 2000). These
measures evaluate respondents’ performance against expert or consensus norms.
The present indicator of emotional intelligence equalled related instruments such as the
Emotion Quotient Inventory (EQ-i; Bar-On, 1996) in that items were fairly similar to
personality items. A distinctive feature of the present instrument for Emotional
Intelligence was that respondents had to indicate their position on a scale between two
behavioural descriptions that were assumed to be opposites. It must be noted that there is
no evidence available yet that the descriptions are in fact valid opposites. Future studies
may examine this further, for example by starting from unipolar items and examining their
factor structure.
Finally, the present study supported the incremental validity of both self- and other
ratings of emotional intelligence in predicting success in academic and social life above
traditional measures of academic intelligence and personality. As predicted, emotional
intelligence was found to be able to predict independent variance in both academic and
social success. Academic success was predicted by autonomy; social competence and
social activities were primarily predicted by empathy. The support for the incremental
validity of the emotional intelligence scales was obtained both for self- and for other
ratings, suggesting that the results cannot be attributed to common method variance. The
incremental validity of emotional intelligence was not very strong. Personality appeared to
be a rather strong predictor of both social and academic success. Social competence was
strongly associated with extraversion, engagement in social activities (negatively) with
conscientiousness. The latter personality dimension positively predicted academic
success. Academic intelligence was not a very strong predictor of either social or

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
122 K. van der Zee et al.

academic success. It must be noted that the criteria that were used in the present study were
not very strong. Particularly, the social activities scale refers to social norms (Ciarrochi,
Chan and Caputi, 2000; Mayer et al., 2000), engagement rather than success, and was of
only moderate reliability. Future studies may for example relate assessment centre ratings
of social skills to evaluations of emotional intelligence. In addition, because the present
study was performed among students, the outcome measures could not be extended to
relevant work outcomes such as sales figures or profitability. Wagner and Sternberg (1985;
see also 1990), for example, showed a positive relationship between practical intelligence
and performance of managers in a management simulation, merit-based salary increase
among bank employees, and average performance rating with respect to generating new
business for the bank. For the use of scales for emotional intelligence in the context of
personnel selection and assessment their predictive validity has to be established against
criteria of real importance to job success. Nevertheless, the present findings provide
support for the incremental validity of emotional intelligence against all four criteria that
we used, giving us reason to have some fate in the additive value of this ‘elusive
construct’.
An important limitation of the present study concerned the fact that the range of
cognitive abilities may be restricted due to the fact that all respondents were students. This
seems to be an important problem when comparing the predictive value of cognitive
abilities and personality against success criteria, because it can be assumed that the
personality scores are not restricted in their range. However, in terms of external validity,
the EI instrument that was under evaluation in the present study was aimed at the higher
educational levels, making the ‘competition’ of personality and cognitive abilities in terms
of their predictive value rather representative for daily practice. It seems that our findings
generalize to the selection context in which the EI indicators are usually applied rather
than providing an ultimate answer to the question of how well intelligence may be able to
predict daily life success in the general population. Moreover, it was found that, taking
pupils from secondary education as a norm group, decile scores on five of the six GATB
scales were below the mean, with mean scores for four scales below 4, indicating that the
group was clearly not a selective group of highly intelligent individuals. Of course,
students from secondary education score above the general population level, and provide
by no means a representative standard. Future studies may try to replicate the present
findings among samples that are more heterogeneous in nature.

REFERENCES

Archer D. 1980. How to Expand your Social Intelligence Quotient. Evans: New York.
Ashton MC, Lee K, Vernon PA. 1999. Fluid intelligence, crystallized intelligence, and the Openess/
Intellect factor. Journal of Research in Personality 22: 198–207.
Bagby RM, Taylor GJ, Parker JD. 1994. The twenty item Toronto Alexithymia Scale_2. Convergent,
discriminant and concurrent validity. Journal of Psychosomatic Research 38: 33–40.
Bar-On R. 1996. The Emotional Quotient Inventory (EQ-i): A Test of Emotional Intelligence.
Toronto: Multi-Health Systems.
Bar-On R, Brown JM, Kirkcaldy BD, Thomé EP. 2000. Emotional expression and implications for
occupational stress: an application of the Emotional Quotient Inventory (EQ-i). Personality and
Individual Differences 28: 1107–1118.
Borkenau P, Liebler A. 1993. Convergence of stranger ratings of personality and intelligence with
self-ratings, partner ratings and measured intelligence. Journal of Personality and Social
Psychology 65(3): 546–553.

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
 Emotional intelligence 123

Boyle G, Stankov L, Cattell RB. 1995. Measurement and statistical models in the study of
personality and intelligence. In International Handbook of Personality and Intelligence, Sakloske
D, Zeidner M (eds). Plenum: New York; 417–446.
Brown W. 1910. Some experimental results in the correlation of mental abilities. British Journal of
Psychology 3: 296–322.
Cantor N, Harlow R. 1994. Social intelligence and personality: flexible life-task pursuit. In
Personality and Intelligence, Sternberg RJ, Ruzgis P (eds). Cambridge University Press:
Cambridge; 137–168.
Cantor N, Kihlstrom JF. 1987. Personality and Social Intelligence. Prentice-Hall: Englewood Cliffs, NJ.
Cattell RB. 1946. Description and Measurement of Personality. World: New York.
Ciarrochi JV, Chan AYC, Caputi P. 2000. A critical evaluation of the emotional intelligence
construct. Personality and Individual Differences 28: 539–561.
Cronbach LJ. 1949. Essentials of Psychological Testing. Harper: New York.
Cronbach LJ. 1955. Processes affecting scores on ‘understanding of others’ and ‘assumed
similarity’. Psychological Bulletin 52: 177–193.
Davies M, Stankov L, Roberts RD. 1998. Emotional Intelligence: in search of an elusive construct.
Journal of Personality and Social Psychology 75(4): 989–1015.
De Raad B. 1992. The replicability of the big five dimensions in three word-classes of the Dutch
language. European Journal of Personality 6: 15–30.
De Raad B, Hendriks AAJ, Hofstee WKB. 1992. Towards a refined structure of personality traits.
European Journal of Personality 2: 81–96.
Derksen JJL. 1998. EQ in Nederland EQ in the Netherlands. Nijmegen: Pen Test.
Digman JM. 1990. Personality structure: emergence of the five-factor model. Annual Review of
Psychology 41: 417–440.
Eysenck HJ. 1994a. Four ways five factors are not basic. Personality and Individual Differences 13:
667–673.
Eysenck HJ. 1994b. Personality and intelligence: psychometric and experimental approaches.
In Personality and Intelligence, RJ Sternberg (ed.). Cambridge University Press: New York;
3–31.
Feshbach ND, Feshbach S. 1987. Affective processes and academic achievement. Child
Development 58: 1135–1347.
Ford M. 1986. For all practical purposes: criteria for defining and evaluating practical intelligence. In
Practical Intelligence: Nature and Origins of Competence in the Everyday World, Sternberg RJ,
Wagner RK (eds). Cambridge University Press: Cambridge; 183–200.
Ford ME, Tisak MS. 1983. A further search for social intelligence. Journal of Educational
Psychology 75: 196–206.
Funder DC. 1989. Personality psychology and the dancing bear. In Personality Psychology:
Recent Trends and Emerging Directions, Buss DM, Cantor N (eds). Springer: New York; 210–
223.
Funder DC. 1999. Personality Judgment: A Realistic Approach to Person Perception. Academic:
London.
Gardner H. 1983. Inquiry into Human Faculty and its Development. Macmillan: London.
Goff M, Ackerman PL. 1992. Personality–intelligence relations: assessment of typical intellectual
engagement. Journal of Educational Psychology 84: 537–552.
Goldberg LR. 1983. The magical number five, plus or minus two: some conjectures on the
dimensions of personality trait descriptors. Paper presented at a research seminar, Gerontology
Research Center: Baltimore, MA.
Goldberg LR. 1990. An alternative ‘description of personality’: the Big-Five factor structure.
Journal of Personality and Social Psychology 59: 1216–1229.
Goleman D. 1996. Emotional Intelligence. Bantam: New York.
Hofstee WKB. 1994. Who should own the definition of personality? European Journal of Personality
8(3): 149–162.
Hofstee WKB. 2001. Personality and intelligence: do they mix? In Intelligence and Personality:
Bridging the Gap in Theory and Measurement, Messick SJ, Collis JM (eds). Lawrence Erlbaum:
Mahwah, NJ; 43–60.
Hofstee WKB, Ten Berge JMF, Hendriks AJJ. 1998. How to score questionnaire items. Personality
and Individual Differences 25: 897–909.

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
124 K. van der Zee et al.

Horn JL, Catell RB. 1966. Refinement and test of the theory of fluid and chrystallized intelligence in
relation to concepts of cognitive psychology and aging in adulthood. In Aging and Cognitive
Processes, Craig FIM, Trehum A (eds). Macmillan: NewYork; 237–278.
John OP. 1990. The ‘Big Five’ factor taxonomy: dimensions of personality in the natural language
and in questionnaires. In Handbook of Personality: Theory and Research, Pervin LA (ed.).
Guilford: New York; 66–100.
John OP, Robins RW. 1993. Determinants of interjudge agreement on persoality traits: the Big Five
domains, observability, evaluativeness, and the unique perspective of the self. Journal of
Personality 61: 521–551.
Jones K, Day JD. 1997. Discrimination of two aspects of cognitive–social intelligence from
academic intelligence. Journal of Educational Psychology 89: 486–497.
Jones EE, Nisbett RE. 1972. The actor and observer: divergent perceptions of the causes of
behaviour. In Attribution: Perceiving the Causes of Behaviour, Jones EE, Kanouse DE, Kelley
HH, Nisbett RE, Valinis S, Weiner B (eds). General Learning: Morristown, NJ; 79–94
Kaiser HF, Caffrey J. 1965. Alpha factor analysis. Psychometrica 30: 1–14.
Kerr, Speroff BJ. 1954. Validation and evaluation of the empathy test. Journal of General
Psychology 50: 269–276.
Lohman D. 1993. Teaching and testing to develop fluid abilities. Educational Researcher 22: 12–23.
Martell J, Mickelvie SJ, Standing L. 1987. Validity of an intuitive personality scale: personal
responsibility as a predictor of academic achievement. Journal of Educational Psychology 47:
1153–1161.
Mayer JD, Cobb CD. 2000. Educational policy on emotional intelligence. Does it make sense?
Educational Psychological Review 12: 163–183.
Mayer JD, Caruso DR, Salovey P. 2000. Emotional intelligence meets traditional standards for an
intelligence. Intelligence 27: 267–298.
Mayer JD, Geher G. 1996. Emotional intelligence and identification of emotion. Intelligence 22:
89–113.
Mayer JD, Salovey P. 1993. The intelligence of emotional intelligence. Intelligence 17: 443–450.
Mayer JD, Salovey P. 1997. What is emotional intelligence? In Emotional Development and
Emotional Intelligence: Educational Implications, Salovey P, Sluyter DJ (ed.). Basic: NewYork;
3–34.
Mayer JD, Salovey P, Caruso D. 2000. Competing mdoels of emotional intelligence. In Handbook of
Human Intelligence, 2nd edn, Sternberg RJ (ed.). Cambridge University Press: New York; 396–420.
McCown W, Johnson J, Austin S. 1986. Inability of delinquents to recognize facial effects. Journal
of Social Behaviour and Personality 1: 489–496.
McCrae RR. 1982. Different points of view: self-reports and ratings in the assessment of personality.
In Recent Advances in Social Psychology: An International Perspective, Forgas JP, Innes MJ (eds).
Elsevier: Amsterdam; 429–439.
McCrae RR. 1992. The five factor model: issues and applications, special issue. Journal of
Personality 60.
McCrae RR, Costa PT. 1984. Emerging Lives, Enduring Positions: Personality in Adulthood. Little
Brown: Boston, MA.
McCrae RR, Costa PT. 1989. Different points of view: self-reports and ratings in the assessment of
personality. In Recent Advances in Social Psychology: An International Perspective, Forgas JP,
Innes MJ (eds). Elsevier: Amsterdam; 429–439.
Neisser U. 1976. General, academic and artificial intelligence. In Human Intelligence: Perspectives
on its Theory and Measurement, Resnick L (ed.). Ablex: Norwood, NJ; 179–189.
Nunnally JC. 1978. Psychometric Theory. McGraw-Hill: New York.
Peabody D. 1984. Selecting representative trait adjectives. Journal of Personality and Social
Pychology 52: 552–567.
Petrides KV, Furnham A. 2000. On the dimensional structure of emotional intelligence. Personality
and Individual Differences 29: 313–320.
Roger D, Najarian B. 1989. The construction and validation of a new scale for measuring emotional
control. Personality and Individual Differences 10: 845–853.
Rothstein MG, Paunonen SV, Rush JC, King AG. 1994. Personality and cognitive ability predictors
of performance in business schools. Journal of Educational Psychology 86: 516–530.

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)
 Emotional intelligence 125

Safer D. 1986. Nonpromotion correlates and outcomes at different grade levels. Journal of Learning
Disabilities 19(8): 500–503.
Salovey P, Mayer JD. 1990. Emotional intelligence. Imagination, Cognition and Personality 9:
185–211.
Salovey P, Mayer JD. 1994. Some final thoughts about personality and intelligence. In Personality
and Intelligence, Sternberg RJ (ed.). Cambridge University Press: New York; 303–318.
Schutte NS, Malouff JM, Hall LE, Haggerty DJ, Cooper JT, Golden CJ, Dorheim L. 1998.
Development and validation of a measure of emotional intelligence. Personality and Individual
Differences 25: 167–177.
Shafer AB. 1999. Relation of the Big Five and factor V subcomponents to social intelligence.
European Journal of Personality 13: 225–240.
Spearman C. 1910. Correlation calculated from faulty data. British Journal of Psychology 3:
271–295.
Stankov L, Crawford JD. 1998. Self-confidence and performance on tests of cognitive abilities.
Intelligence 25: 93–109.
Sternberg RJ. 1985. Beyond IQ: A Triarchic Theory of Human Intelligence. Cambridge University
Press: New York.
Sternberg RJ. 1997. Successful Intelligence. Plenum: New York.
Sternberg RJ, Wagner RK. 1986. Practical Intelligence: Nature and Origins of Competence in the
Everyday World. Cambridge University Press: New York.
Sternberg RJ, Wagner RK. 1993. The geocentric view of intelligence and job performance is wrong.
Current Directions in Psychological Science 2: 1–5.
Sternberg RJ, Wagner RK, Williams WM, Horvath JA. 1995. Testing common sense. American
Psychologist 50: 912–927.
Stricker LJ, Rock DA. 1990. Interpersonal competence, social intelligence, and general ability.
Personality and Individual Differences 11: 833–839.
Thorne A. 1989. Conditional patterns, transference, and the coherence of personality across time. In
Personality Psychology: Recent Trends and Emerging Directions, Buss DM, Cantor N (eds).
Springer: NewYork; 149–159.
’T Mannetje J, Schattenberg A. 1996. Het PCT Project, de Ontwikkeling van een Computeronder-
steunde Testserie: Connector-P en Connector-C (The PCT-Project: Development of a Computer-
Assisted Testseries: Connector-P and Connector-C). Doctoral Dissertation, University of Tilburg.
US Department of Labor. 1970. Manual for the General Aptitude Test-Battery. Author: Washington,
DC.
Van der Flier H, Boomsma-Suerink JL. 1994. Handbook GATB (GATB Manual). Stichting GATB
Research: Amsterdam.
Wagner RK, Sternberg RJ. 1985. Practical intelligence in real-world pursuits: the role of tacit
knowledge. Journal of Personality and Social Psychology 49: 436–458.
Wagner RK, Sternberg RJ. 1986. Tacit knowledge and intelligence in the everyday world. In
Practical Intelligence: Nature and Origins of Competence in the Everyday World, Sternberg RJ,
Wagner RK (eds). Cambridge: New York; 51–83.
Wagner RK, Sternberg RJ. 1990. Street smarts. In Measures of Leadership, Clark KE, Clark MB
(eds). Leadership Library of America: West Orange, NJ; 493–504.
Wong CT, Day JD, Maxwell SE, Meara NM. 1995. A multitrait–multimethod study of academic and
social intelligence in college students. Journal of Developmental Psychology 19: 521–530.

Copyright # 2002 John Wiley & Sons, Ltd. Eur. J. Pers. 16: 103–125 (2002)