Available online at www.sciencedirect.com
Personality and Individual Differences 44 (2008) 712–722
www.elsevier.com/locate/paid
Emotional intelligence and job satisfaction: Testing the
mediatory role of positive and negative affect at work
Konstantinos Kafetsios
a,*
, Leonidas A. Zampetakis
b
a
b
Department of Psychology, University of Crete, Rethimnon 74100, Crete, Greece
Department of Production Engineering and Management, Technical University of Crete, Crete 73100, Greece
Received 25 June 2007; received in revised form 21 September 2007; accepted 3 October 2007
Available online 14 November 2007
Abstract
The study tested the extent to which positive and negative affect at work mediate personality effects (Emotional Intelligence) on job satisfaction. Participants were 523 educators who completed the Wong Law Emotional Intelligence Scale, a version of the Job Affect Scale and the General Index of Job Satisfaction. Results
using structural equation modelling indicated that positive and negative affect at work substantially mediate
the relationship between EI and job satisfaction with positive affect exerting a stronger influence. In males,
affect at work fully mediated the EI effect on job satisfaction. Among the four EI dimensions, use of emotion
and emotion regulation were significant independent predictors of affect at work. The results confirm expectations deriving from Affective Events Theory regarding the role of work affectivity as an interface between
personality and work attitudes and extend the literature on EI effects in organizational settings.
Ó 2007 Elsevier Ltd. All rights reserved.
Keywords: Emotional intelligence; Affect at work; Job satisfaction
1. Introduction
In the context of the emerging ‘affective revolution’ in social and organizational psychology
(Barsade & Gibson, 2007) Emotional intelligence (EI) is proposed as an important predictor of
*
Corresponding author. Tel./fax: +30 28310 77534.
E-mail address: k.kafetsios@psy.soc.uoc.gr (K. Kafetsios).
0191-8869/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.paid.2007.10.004
K. Kafetsios, L.A. Zampetakis / Personality and Individual Differences 44 (2008) 712–722
713
key organizational outcomes including job satisfaction (Daus & Ashkanasy, 2005; Van Rooy &
Viswesvaran, 2004). There is accumulating evidence that EI abilities and traits influence job satisfaction (e.g., Carmeli, 2003; Sy, Tram, & O’Hara, 2006) but to our knowledge, there is no study
testing the extent to which affective components of the work experience may mediate such a
connection. Theories of emotion in organizations (i.e., Affective Events Theory, Weiss & Cropanzano, 1996) suggest that affective states at work are key vehicles of personality and organizational influences on job satisfaction and performance. Moreover, it has been pointed out that
there is little research examining dispositional factors of job satisfaction (especially affect-related;
Connolly & Viswesvaran, 2000). Emotional intelligence thus can constitute a link between trait
level affectivity, work affect and job satisfaction and the present study tested such a model using
a structural equations modeling approach.
1.1. Defining EI
At a theoretical level EI reflects the extent to which a person attends to, processes, and acts
upon information of an emotional nature intra-personally and inter-personally. However, there
are ensuing debates at the operational level mainly, that have led to two distinct approaches:
the ability and trait Emotional Intelligence. The ability approach uses maximum performance
measures such as the Mayer, Salovey & Caruso Emotional Intelligence Test (Mayer, Salovey,
& Caruso, 2002) to assess individual differences in the interface of emotion with cognitive processes (Mayer & Salovey, 1997). Trait EI on the other hand incorporates EI abilities within a more
general framework of individual self-perceived emotionality and emotion efficacy (e.g., Petrides &
Furnham, 2000). Nevertheless, some trait EI approaches and corresponding measures (e.g., EQi,
Bar-On, 1997) diverge from the conceptualization of EI described in the beginning of this section
in that they focus on individual differences in adaptation to environmental demands.
1.2. EI and job satisfaction
There are several reasons why workers’ EI may influence job satisfaction. Interpersonally, emotion awareness and regulatory processes associated with EI are expected to benefit peoples’ social
relationships hence affecting the experience of emotion and stress at work. Intrapersonally, use of
emotion and being aware of one’s own emotions can lead to regulating stress and negative emotion so that one can perform better at work.
Research that has empirically examined relationships between EI and job satisfaction has returned mixed findings. A number of studies have observed weak to modest relationships between
trait EI measures (i.e., EQi, Carmeli, 2003; Kafetsios & Loumakou, 2007; a Greek trait EI scale,
Vacola, Tsaousis, & Nikolaou, 2003) and job satisfaction. A recent study of food service workers
and their managers (Sy et al., 2006) observed a positive association between an ability based EI
scale (Wong Law Emotional Intelligence Scale, WLEIS; Wong & Law, 2002) and job satisfaction
in employees and their managers. Finally, a study of a group of managers Lopes, Grewal, Kadis,
Gall, and Salovey (2006) found links between EI abilities and affective proxies of job satisfaction
measured via self and supervisor reports.
Like in Sy et al.’s (2006) study we used a self-report measure of EI (WLEIS) that conceptually
subscribes to the ability approach and has shown to have good discriminant and predictive
714
K. Kafetsios, L.A. Zampetakis / Personality and Individual Differences 44 (2008) 712–722
validity in organizational settings. Research has shown that the scale is distinct from the Big Five
personality factors and has convergent validity with other EI measures such as the Trait Metamood scale (Law, Wong, & Song, 2004). The WLEIS has proven to be a predictor of job satisfaction in different organizational levels (Law et al., 2004; Sy et al., 2006).
1.3. The role of positive and negative affect
Work affect is an important aspect of the work experience. Affective Events Theory (Weiss &
Cropanzano, 1996) proposes that cumulative affective experiences in the work environment together with other factors (including personality) shape workers’ job related attitudes. Affect at
work cannot be equated to job satisfaction since work attitudes also reflect cognitive evaluations
and beliefs (Weiss, 2002). Affect is a subjective feeling state with a positive and negative hedonic
tone and should be distinguished from discrete emotions and moods in that it has a specific contextual element but not a particular target or elicitor (Frijda, 1986). There are debates over
whether the dimensions that underlie positive and negative affectivity are best described in terms
of activation (e.g., Watson, Clark, & Tellegen, 1988) or valence (Barrett & Russell, 1998). Based
on evidence that hedonic tone and not activation is a significant correlate of job satisfaction
(Weiss, Nicholas, & Daus, 1999) the present study distinguished primarily between positive and
negative affectivity. Positive and negative affect as different but correlated dimensions (Russell &
Carroll, 1999).
1.4. Overview of the study’s aims and hypotheses
The study aimed firstly to determine whether, and the extent to which, EI is associated with
affect at work and job satisfaction. Based on recent findings (Lopes et al., 2006; Sy et al., 2006)
we expected that self-reported EI abilities will be associated with positive affect and job satisfaction and inversely related to negative affect at work. Moreover, we wanted to examine which EI
dimensions may be related to positive and negative affect and job satisfaction. Secondly, we aimed
to test whether, and the extent to which, positive and negative affect at work mediate EI effects on
job satisfaction. Based on Affective Events Theory, we expected that work affect would at least
partially mediate EI effects on job satisfaction; we did not have any hypotheses with regards to
the relative strength of positive and negative affect in this relationship. We applied a Structural
Equations Modelling approach in order to clearly distinguish the mediatory power of positive
and negative affect while controlling for common variance between positive and negative affect.
2. Method
2.1. Participants
Participants were 523 teachers, working in primary and secondary education from various regions in Greece. The sample included 155 males and 368 females aged 25–59 years (M = 38.48
years, SD = 8.02). The average length of service as an educator was 12 years (SD = 8.48). Half
of the participants (51%) were involved in teaching in primary education. Administration of
K. Kafetsios, L.A. Zampetakis / Personality and Individual Differences 44 (2008) 712–722
715
the questionnaires was carried out by trained post-graduate students who acted as research assistants and no monetary incentive was provided.
2.2. Measures
All scales were translated into Greek (by the first author), blindly back-translated by a Greek
graduate student with some items modified to enhance the naturalism of the translations (Van de
Vijver & Leung, 1997).
2.2.1. Job satisfaction
We adapted into Greek the General Index of Job Satisfaction (Brayfield & Rothe, 1951). The
scale comprises of 18 items (Cronbach’s a = .92). Based on the results of an exploratory factor
analysis we created four parcels of this construct. We assigned items to indicators on the relative
size of their factor loadings in order to evenly distribute items across indicators. Coefficient alphas
for the four scales were: Job1 – .81; Job2 – .81; Job3 – .80; Job4 – .79.
2.2.2. Emotional intelligence
We used the self-report Wong Law Emotional Intelligence Scale (WLEIS, Wong & Law, 2002).
The scale consists of four dimensions that are consistent with Mayer and Salovey’s (1997) definition of EI. The Self-Emotion Appraisal (SEA) dimension assesses an individual’s self-perceived
ability to understand their emotions. The Others’ Emotion Appraisal (OEA) dimension assesses
a person’s tendency to be able to perceive other peoples’ emotions. The Use of Emotion (UOE)
dimension concerns the self-perceived tendency to motivate one self to enhance performance. The
Regulation of Emotion (ROE) dimension concerns individuals’ perceived ability to regulate their
own emotions. Coefficients alphas for the four elements were: SEA: .83; OEA: .77; ROE: .83;
UOE: .79.
We used confirmatory factor analysis (AMOS 7.0; Arbuckle, 2006) to evaluate the factorial
structure of the Greek version of the WLEIS. We compared 2 alternative models: Model 1 specified a single factor behind all the 16 items, while Model 2 specified the four correlated dimensions
from their respective items and then a second-order factor behind the four EI dimensions. Model
2 fitted the data better [v2 (100, N = 523) = 423.43, p = 0.000; RMSEA = 0.079 (90% CI: 0.071–
0.084); GFI = 0.915; CFI = 0.932; and AIC = 495.43] than Model 1 [v2 (104, N = 523) = 1084.16,
p = 0.000; RMSEA = 0.134 (90% CI: 0.127–0.142); GFI = 0.770; CFI = 0.744 and
AIC = 1148.84]. These results are in line with previous studies (Law et al., 2004; Wong & Law,
2002) and indicate that items for EI measurement can serve as a reasonable estimate of their
dimensions, and that the dimensions in turn can represent an underlying multidimensional EI
construct. The internal consistency for all 16 items was a = .90.
2.2.3. Positive and negative affect at work
To assess affect at work we used 12 items of the Job Affect Scale (Brief, Burke, George,
Robinson, & Webster, 1988) that assesses participants’ experience of positive and negative affect at work during the previous week on a 5-point scale. The six positive items (JAS-PA)
were: enthusiastic, elated, active, strong, happy, and excited. The six negative affective states
(JAS-NA) were: hostile, scornfull, fearful, sleepy, placid, and sad. The positive and negative
716
K. Kafetsios, L.A. Zampetakis / Personality and Individual Differences 44 (2008) 712–722
affect parts of the scale had good internal consistency (Alphas .89 and .78, respectively). We
followed the same procedure as with the job satisfaction construct and we created two indicators for both positive and negative parts. Alphas for the scales were: PA1 – .82; PA2 –
.81; NA1 – .68; NA2 – .61.
2.3. Analytic strategy
In the present study we tested whether positive and negative affectivity at work mediated
fully or partially the relationship between trait EI and job satisfaction in the context of Structural Equation Modelling using maximum likelihood estimation method. Prior to the analysis
data screening was performed and data were tested for deviation from normality. Following
Hall, Snell, and Foust (1999) we formed item parcels on the basis of factor analysis in order
to control for inflated measurement errors and improve the psychometric properties of the
variables. We used a two-stage analytic procedure: in stage 1 the four-factor model was fitted
to the data and then a measurement model specifying perfect correlation among all four latent
variables was assessed to test overall discriminability. The one-factor model also provides a
test for common method bias. We used the sequential v2 differential test (SCDT) to assess
nested model comparisons. We employed several model fit statistics (RMSEA: Root Mean
Square Error Approximation; CFI: Comparative Fit Index; GFI: Goodness of Fit Index;
RMR: Root Mean Square Residual; AIC: Akaike Information Criterion; Shook, Ketchen,
Hult, & Kacmar, 2004). In order to select among the competing structural models we applied
model selection for SEM (Raftery, 1993). We employed a stepwise strategy in model selection,
which included forward selection and backward elimination features. We used Steiger’s Power
Analysis (StatSoft, 2001) to estimate SEM model-level power. Finally, we used bootstrapping
procedures (resampled 1000 times and used the percentile method to create 95% confidence
intervals).
3. Results
Table 1 presents means, standard deviations and variable intercorrelations. Trait EI was significantly related to job satisfaction, positive affect and negative affect. Table 1 also revealed that job
satisfaction was positively and significantly related to positive affect and negatively related to negative affect.
3.1. Assessment of measurement model
Table 2 displays the fit statistics for the measurement model. Overall, the hypothesized measurement model fit the data quite well when evaluated in terms of the recommended cutoffs or
the combination cut off approach (Shook et al., 2004). The hypothesized measurement model
fit the data better than a single factor model, both in terms of the fit statistics and when directly
contrasted with a change in v2 test and AIC. In summary, the results suggest that the proposed
factor structure presents a statistically adequate and sufficient fit to the data, indicating the absence of severe common method variance.
K. Kafetsios, L.A. Zampetakis / Personality and Individual Differences 44 (2008) 712–722
717
Table 1
Descriptive statistics and intercorrelations for total sample
1.
2.
3.
4.
5.
6.
7.
M
SD
38.48
11.90
5.28
3.88
3.46
1.41
8.02
8.47
0.81
0.65
0.80
0.55
a
Gender
Age
Years in service
Trait EI
Job satisfaction
Positive affect
Negative affect
1
2
3
4
5
6
7
–
0.13**
0.03
0.04
0.01*
0.06
0.03
–
0.66**
0.00
0.06
0.03
0.14**
–
0.03
0.02
0.01
0.13**
(0.90)
0.43**
0.42**
0.27**
(0.92)
0.64**
0.44**
(0.89)
0.36**
(0.78)
Note: N = 523. Internal reliabilities are in parenthesis.
a
Gender is coded 1 = male 2 = female.
*
p < 0.05.
**
p < 0.01.
Table 2
Measurement models fit statistics
Model
Hypothesized four
factor measurement
model
One factor
measurement model
v2
df
**
230.16
48
706.83**
52
Dv2
545.47**
RMSEA
GFI
CFI
RMR
AIC
0.085 (90% CI: 0.074–0.096)
0.926
0.948
0.032
290.16
0.115 (90% CI: 0.145–0.166)
0.812
0.814
0.171
758.83
Note: v2: chi-square statistic.
**
p < 0.001.
3.2. Assessment of structural models
The next step was to consider comparative models specifying total effects (direct and indirect),
complete mediation and partial mediation. Results of the model selection procedure indicated that
there was 97.9% probability (in terms of Akaike weights) that the best model is the one presented
in Fig. 1. We considered the error terms of positive and negative affect correlated in line with previous research (Barrett & Russell, 1998).
This model revealed a good fit to the data: v2 (48, N = 523) = 230.16, p = 0.000;
RMSEA = 0.085 (90% CI: 0.074–0.096); GFI = 0.926; CFI = 0.948; RMR = 0.032 and
AIC = 290.16. The model postulated that the effects of trait EI on job satisfaction were partially
mediated by positive and negative affect (see Table 3). Power analysis (e1 = 0.08, a = 0.05,
N = 523, df = 48) suggested a very low probability to reject the model (.5%). The standardized
direct effect of EI on job satisfaction was 0.14 (p < 0.01). EI had direct effects on positive affect
(0.49, p < 0.01) and negative affect ( 0.36, p = 0.002) along with indirect effects on job satisfaction (0.24, p < 0.01). In sum, the standardized total effect of EI on job satisfaction was 0.49
(95% percentile confidence interval: 0.40–0.58, p < 0.01).
Positive affect had a statistically significant direct effect on job satisfaction (0.56, 95% percentile
confidence interval (CI): 0.43–0.66, p < 0.01) Negative affect had a direct negative effect on job
718
K. Kafetsios, L.A. Zampetakis / Personality and Individual Differences 44 (2008) 712–722
Fig. 1. Standardized results of the structural model assessment.
Table 3
Standardized direct and indirect effects and the associated 95% confidence intervals
Predictor
Trait EI
Positive affect
Negative affect
Outcome
Positive affect
Negative affect
Direct
Direct
*
0.49
[0.39–0.58]
–
–
Indirect
*
0.36 [( 0.47)–( 0.26)]
–
–
–
–
–
Job satisfaction
Indirect
Direct
*
–
0.14 (0.04–0.25)
–
–
0.56* (0.43–0.66)
0.19* [( 0.30)–( 0.075)]
Indirect
0.24*
(0.18–0.32)
–
–
Note: The upper and lower bounds of the 95% confidence interval (shown in parentheses) were based on the findings
from a bootstrapping analysis using the percentile method.
*
p < 0.01.
satisfaction ( 0.19, 95% CI: ( 0.30)–( 0.075), p < 0.01). The correlation between the error terms
of positive and negative effect was significant ( 0.36, p < 0.01, 95% CI: ( 0.50)–( 0.22),
p < 0.01). The proportion of variance in job satisfaction explained by the collective set of predictors was 57%.
K. Kafetsios, L.A. Zampetakis / Personality and Individual Differences 44 (2008) 712–722
719
3.3. Gender differences
We found no statistically significant gender differences in EI and work affect but women scored
higher than men on job satisfaction at a statistically significant level. Interestingly, job satisfaction
development followed different patterns between genders. For male teachers the most probable
model in terms of Akaike weights (92% probability) was the one where positive and negative affect
fully mediate trait EI effects on job satisfaction. The standardized direct effect of trait EI on positive affect was 0.62 (95% CI: 0.43–0.77, p < 0.01), and on negative effect was ( 0.40), (95% CI:
( 0.56)–( 0.21), p < 0.01). Positive affect had a strong direct effect on job satisfaction 0.77, (95%
CI: 0.56–0.98, p < 0.01), while the direct effect of negative affect on job satisfaction was not significantly different from zero 0.17, (95% CI: ( 0.39)–0.02, p = 0.09). The correlation between
the error terms of positive and negative affect were not significantly different from zero at the
0.05 level: 0.24, (95% CI: ( 0.5)–0.002, p = 0.053). The aforementioned relationships among variables explained 71% of the variance in job satisfaction.
On the other hand, for females the most likely model (99%) was one where trait EI had a
direct and indirect effect on job satisfaction, that is to say, the relationship between EI and job
satisfaction is partially mediated by positive and negative affect. The standardized direct effect
of trait EI on job satisfaction was 0.13, (95% CI: 0.06–0.23, p = 0.002); on positive affect was
0.35 (95% CI: 0.24–0.46, p = 0.002), and on negative affect was ( 0.19), (95% CI: ( 0.32)–
( 0.09), p = 0.002). There was a strong direct effect from positive affect to job satisfaction
0.40, (95% CI: 0.27–0.55, p = 0.002), while the direct effect of negative affect on job satisfaction is 0.24, [95% CI: ( 0.39)–( 0.03), p = 0.007]. The total standardized effect of trait EI on
job satisfaction was 0.47, (95% CI: 0.35–0.57, p = 0.002). The correlation between the error
terms of positive and negative affect was significantly different from zero at the 0.05 level:
0.42, (95% CI: ( 0.58)–( 0.23), p = 0.02). The aforementioned relationships among variables
explained 51% of the variance in job satisfaction.
3.4. Separating the effects of the EI dimensions
Finally we performed three multiple regression analyses in order to determine the relative effects
of the four EI branches on affect at work and job satisfaction. The four EI branches explained
21% in the positive emotion variance and this model was significant (F(4, 518) = 33.96,
p < .001). Among the four EI branches only use of emotion and regulation of emotion were significant predictors of positive affect at work (b = .14, p < .01 and b = .36, p < .001, respectively).
The four EI branches explained 8% in the negative emotion variance (F(4, 518) = 17.7, p < .001)
and again, use of emotion and emotion regulation were significant independent predictors
(b = .12, p < .01) and (b = .21, p < .001, respectively). Finally, three of the four EI branches
were significant independent predictors of job satisfaction (other appraisal of emotion, b = .13,
p < .05, use of emotion, b = .16, p < .01, regulation of emotion, b = .28, p < .001, R2 = 20%,
F(4, 518) = 32.03, p < .001). Entering positive and negative affect in a second step rendered the
partial correlations of emotion and regulation of emotion with job satisfaction non-significant.
Understanding of emotion however, retained a significant unique prediction on job satisfaction
(b = .09, p < .05), suggesting that direct effects from trait EI on job satisfaction may owe this
to the EI dimension.
720
K. Kafetsios, L.A. Zampetakis / Personality and Individual Differences 44 (2008) 712–722
4. Discussion
The present study extends an emerging body of research on affectivity in the workplace by testing for links between trait level emotionality (EI), affect at work and job satisfaction. The results
underline the important role of positive and negative affect at work in this relationship.
In keeping with recent studies (Lopes et al., 2006; Sy et al., 2006) the results demonstrated convincingly that EI is an important personality-level predictor of work affectivity and job satisfaction. Subsequent regression analyses indicated that use of emotion and emotion regulation are
two EI dimensions predictive of positive and negative affect at work and perceiving others’ emotions was uniquely associated with job satisfaction a finding that primarily concerned women.
Notably, these findings are at odds with studies showing weak relationships between some trait
EI measures (i.e., Kafetsios & Loumakou, 2007) and job satisfaction. Given that some of this research evidence concerned the same population as in this study one may exclude organizational
level variables as possible moderators and focus on the measurement instruments being used
(Brackett & Mayer, 2003). In the present study, we used a self-report measure of EI (WLEIS) that
adheres to the ability model of EI and confirmed its psychometric properties and utility for a nonEnglish speaking culture.
The results from the SEM suggested that in absolute terms the direct effect of EI is stronger for
positive affect compared to negative affect. This is an important finding supporting accumulating
evidence for the primacy of positive over negative affect as a predictor of work outcomes (Thoresen, Kaplan, Barsky, & de Chermont, 2003). Positive affect is a source of human strength (Isen,
2003) and positive affect predisposes people to cognitions, feelings and actions that promote the
building of personal and social resources (Fredrickson, 2001; Lyubomirsky, King, & Diener,
2005). Persons high on EI seem to be best suited to follow these ‘broadening and build’ strategies
also in the work environment.
There were some gender differences in the pattern of results. Male workers’ positive and negative
affect at work fully mediated EI effects on job satisfaction whereas in female workers some direct effects were observed at a level comparable with the final model presented in Fig. 1. The results from
subsequent regression analyses indicated that when controlling for positive and negative affect at
work perceiving others’ emotion was uniquely associated with job satisfaction a finding that may reflect female work related gender role-characteristics (Petrides & Furnham, 2006).
Importantly, the results from the current confirmatory model support key assumptions of
Affective Events Theory (Weiss & Cropanzano, 1996) regarding the role of positive and negative
affect at work. The theory promotes a multifaceted view of job satisfaction incorporating affective,
cognitive and personality elements (Weiss, 2002). Clearly, positive and negative affectivity at work
explained a significant part in job satisfaction variance as expected by AET suggesting that events
taking place at work influence employees’ attitudes towards their work. The results point to the
personality antecedents to affect at work and further research could strive to distinguish those effects from organizational level variables (e.g., organizational climate, managerial structure, etc.).
4.1. Limitations and further research
One of the limitations of the current study was the adherence to a dimensional model of work
emotionality and future research could test whether different facets of EI are associated with
K. Kafetsios, L.A. Zampetakis / Personality and Individual Differences 44 (2008) 712–722
721
discrete emotions at work. Furthermore, the study used a retrospective measure of work affectivity and future research should expand the study of affective phenomena at work using measures of
‘on-line’ affect such event sampling methodologies or observational methods. Finally, in line with
AET, future research could identify the work events that give rise to positive and negative emotion at work and for which EI acts as a moderator. It would also be interesting to examine the
extent to which EI and affect at work interface to influence work attitudes in occupations with
job characteristics different than educators’.
References
Arbuckle, J. L. (2006). AMOS 7.0. User guide. Chicago: SmallWaters Corporation.
Bar-On, R. (1997). Bar-on emotional quotient inventory technical manual. Toronto: Multi-Health Systems.
Barrett, L., & Russell, J. A. (1998). Independence and bipolarity in the structure of current affect. Journal of Personality
and Social Psychology, 74, 967–984.
Barsade, S. G., & Gibson, D. E. (2007). Why does affect matter in organizations? Academy of Management Perspectives,
21(1), 36–57.
Brackett, M. A., & Mayer, J. D. (2003). Convergent, discriminant, and incremental validity of competing measures of
emotional intelligence. Personality and Social Psychology Bulletin, 29, 1147–1158.
Brayfield, A., & Rothe, H. (1951). An index of job satisfaction. Journal of Applied Psychology, 35, 301–311.
Brief, A. P., Burke, M. J., George, J. M., Robinson, B. S., & Webster, J. (1988). Should negative affectivity remain an
unmeasured variable in the study of job stress? Journal of Applied Psychology, 73, 193–198.
Carmeli, A. (2003). The relationship between emotional intelligence and work attitudes, behavior and outcomes.
Journal of Managerial Psychology, 18, 788–813.
Connolly, J. J., & Viswesvaran, C. (2000). The role of affectivity in job satisfaction: A meta-analysis. Personality and
Individual Differences, 29, 265–281.
Daus, C. S., & Ashkanasy, N. M. (2005). The case for the ability-based model of emotional intelligence in
organizational behaviour. Journal of Organizational Behavior, 26, 453–466.
Fredrickson, B. L. (2001). The role of positive emotions in positive psychology: The broaden-and-build theory of
positive emotions. American Psychologist, 56, 218–226.
Frijda, N. (1986). The emotions. Cambridge: Cambridge University Press.
Hall, R. J., Snell, A. F., & Foust, M. S. (1999). Item parceling strategies in SEM: Investigating the subtle effects of
unmodeled secondary constructs. Organizational Research Methods, 2(3), 233–256.
Isen, A. (2003). Positive affect as a source of human strength. In L. Aspinwall & U. Staudinger (Eds.), A psychology of
human strength (pp. 179–195). Washington, DC: APA.
Kafetsios, K., & Loumakou, M. (2007). A comparative evaluation of the effects of trait emotional intelligence and
emotion regulation on affect at work and job satisfaction. International Journal of Work Organization and Emotion,
2(1), 71–87.
Law, K. S., Wong, C., & Song, L. J. (2004). The construct and criterion validity of emotional intelligence and its
potential utility for management studies. Journal of Applied Psychology, 89, 483–496.
Lopes, P. N., Grewal, D., Kadis, J., Gall, M., & Salovey, P. (2006). Evidence that emotional intelligence is related to job
performance and affect and attitudes at work. Psicothema, 18(1), 132–138.
Lyubomirsky, S., King, L., & Diener, E. (2005). The benefits of frequent positive affect: Does happiness lead to success?
Psychological Bulletin, 131(6), 803–855.
Mayer, J. D., & Salovey, P. (1997). What is an emotional intelligence? In P. Salovey & D. Sluyter (Eds.), Emotional
development and emotional intelligence: Implications for educators. New York: Basic Books.
Mayer, J. D., Salovey, P., & Caruso, D. R. (2002). MSCEIT user’s manual. Toronto, Ontario: Multi-Health Systems
Inc.
Petrides, K. V., & Furnham, A. (2000). On the dimensional structure of emotional intelligence. Personality and
Individual Differences, 29, 313–320.
722
K. Kafetsios, L.A. Zampetakis / Personality and Individual Differences 44 (2008) 712–722
Petrides, K. V., & Furnham, A. (2006). The role of trait emotional intelligence in a gender-specific model of
organizational variables. Journal of Applied Social Psychology, 36(2), 552–569.
Raftery, A. E. (1993). Bayesian model selection in structural equation models. In K. A. Bollen & J. S. Long (Eds.),
Testing structural equation models (pp. 163–180). Newbury Park, CA: Sage Publications.
Russell, J. A., & Carroll, J. M. (1999). On the bipolarity of positive and negative affect. Psychological Bulletin, 125,
3–30.
Shook, C. L., Ketchen, D. J. J., Hult, G. T. M., & Kacmar, K. M. (2004). An assessment of the use of structural
equation models in strategic management research. Strategic Management Journal, 25, 397–404.
StatSoft, (2001). Statistica for Windows v. 6.0: Data analysis software system. Tulsa, OK: USA.
Sy, T., Tram, S., & O’Hara, L. A. (2006). Relation of employee and manager emotional intelligence to job satisfaction
and performance. Journal of Vocational Behavior, 68, 461–473.
Thoresen, C. J., Kaplan, S. A., Barsky, A. P., & de Chermont, K. (2003). The affective underpinningas of job
perceptions and attitudes: A meta-analytic review and integration. Psychological Bulletin, 129(6), 914–945.
Vacola, M., Tsaousis, I., & Nikolaou, I. (2003). The role of emotional intelligence and personality variables on attitudes
toward organizational change. Journal of Managerial Psychology, 19(2), 88–110.
Van de Vijver, V., & Leung, K. (1997). Methods and data analysis for cross-cultural research. Thousand Oaks, CA: Sage.
Van Rooy, D. L., & Viswesvaran, C. (2004). Emotional intelligence: A meta-analytic investigation of predictive validity
and nomological net. Journal of Vocational Behavior, 65(1), 71–95.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative
affect: The PANAS scales. Journal of Personality and Social Psychology, 54, 1063–1070.
Weiss, H. M. (2002). Deconstructing job satisfaction: Separating evaluations, beliefs and affective experiences. Human
Resource Management Review, 12, 173–194.
Weiss, H. M., & Cropanzano, R. (1996). Affective events theory: A theoretical discussion of the structure, causes and
consequences of affective experiences at work. In B. M. Staw & L. L. Cummings (Eds.). Research in organizational
behavior (Vol. 19, pp. 1–74). Greenwich, CT: JAI.
Weiss, H. M., Nicholas, J. P., & Daus, C. S. (1999). An examination of the joint effects of affective experiences and job
beliefs on job satisfaction and variations in affective experiences over time. Organizational Behavior and Human
Decision Processes, 78(1), 24.
Wong, C. S., & Law, K. S. (2002). The effects of leader and follower emotional intelligence on performance and
attitude: An exploratory study. The Leadership Quarterly, 13, 243–274.