Journal of Research in Personality 49 (2014) 21–24

Contents lists available at ScienceDirect

Journal of Research in Personality

journal homepage: www.elsevier.com/locate/jrp

Brief Report

When to cooperate and when to compete: Emotional intelligence

in interpersonal decision-making
Pablo Fernández-Berrocal a,⇑, Natalio Extremera a, Paulo N. Lopes b, Desireé Ruiz-Aranda a
a
University of Málaga, Campus de Teatinos, s/n, 29071 Malaga, Spain
b
Catholic University of Portugal, Travessa Palma, 1649-023 Lisboa, Portugal

a r t i c l e i n f o a b s t r a c t

Article history: This study examined the relationship between emotional intelligence (EI), assessed with an ability test,
Available online 29 December 2013 and interpersonal decision-making using the Prisoner’s Dilemma Game (PDG). Previous research found
that individuals who self-report high EI tend to cooperate more than others in the Prisoner’s Dilemma.
Keywords: We relativize these findings by showing that individuals scoring high on an ability measure of EI choose
Emotional intelligence effective strategies to deal with three different PDG conditions during real interactions. This suggests that
Social interaction emotionally intelligent individuals are not rigidly predisposed to cooperate regardless of others’ behavior.
Social dilemmas
Instead, EI is associated with the capacity to respond flexibly to others’ strategies and to the interaction
Prisoner’s dilemma
Cooperative goals
context in order to maximize long-term gains – even when this means competing rather than
cooperating.
Ó 2013 Elsevier Inc. All rights reserved.

1. Introduction as generalized social phobia, schizotypy, and schizophrenia, who

Emotional intelligence (EI) theory proposes that the abilities to
perceive, understand, use and manage emotions in oneself and
others are inter-related and constitute a distinct form of intelli-
gence (Mayer & Salovey, 1997). According to this theory, emotion-
ally intelligent individuals use these skills to navigate the social
world and manage interpersonal interaction and social dilemmas
effectively. EI plays a role in the establishment and maintenance
of interpersonal relationships (Mayer, Roberts, & Barsade, 2008).
Previous research has found that individuals scoring highly on abil-
ity measures of EI (Mayer–Salovey–Caruso Emotional Intelligence
Test, MSCEIT) tend to have more positive and less conflictive rela-
tionships with others in a range of naturalistic interpersonal con-
texts (e.g., Brackett, Warner, & Bosco, 2005; Lopes, Salovey, &
Straus, 2003; Lopes et al., 2004). EI predicts emotional and social
competencies not only in a context of consolidated social relations,
but also in initial encounters with strangers (Brackett, Rivers, Shiff-
man, Lerner, & Salovey, 2006). Recent studies using neuroimaging
techniques provide further evidence that persons with higher
MSCEIT scores solve social problems more quickly and accurately
than their counterparts, while revealing less brain activity, even
after controlling for differences in performance in solving non-
social problems (Reis et al., 2007). Studies with clinical samples
lead to similar conclusions: patients with psychopathologies such
⇑ Corresponding author. Address: University of Málaga, Campus de Teatinos, s/n,
29071 Malaga, Spain. Fax: +34 95 213 26 31.
E-mail address: berrocal@uma.es (P. Fernández-Berrocal).
score lower on the MSCEIT, show poorer interpersonal and social
functioning than control group with higher MSCEIT scores (Aguir-
re, Sergi, & Levy, 2008; Jacobs et al., 2008; Kee et al., 2009).
On the basis of existing theory and research, people with higher
EI would be expected to anticipate others’ actions and therefore
make better interpersonal decisions in the course of social interac-
tions. This hypothesis can be tested using the Prisoner’s Dilemma
Game (PDG; Axelrod, 1984), one of the most widely used para-
digms for simulating the course of social decision-making. Of the
different strategies analyzed in the literature on PDG (e.g., always
cooperate, always defect, tit-for-tat), the most effective for achieving
cooperation based on reciprocity in the medium and long term is
tit-for-tat (Axelrod, 1984). Studies using the PDG have shown that
emotional states and individual differences related to emotional
traits influence cooperation. For example, Nelissen, Dijker, and
De Vries (2007) reported that fear reduces, while guilt increases,
cooperation in PDG; Hirsh and Peterson (2009) showed that higher
scores on the withdrawal facet of neuroticism and the enthusiasm
facet of extraversion independently predict greater likelihood of
cooperation.
But how do people with high EI behave? Schutte et al. (2001;
Study 4) found that participants scoring higher on a self-report
measure of EI cooperated more in PDG. This raises a concern that
high-EI individuals might be inclined to cooperate in PDG even
when they should adopt a different strategy to defend their inter-
ests. Evolutionary game theory (Axelrod, 1984) has shown that cer-
tain strategies, such as always cooperate, are not evolutionarily
stable, because in certain highly competitive contexts they do

0092-6566/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jrp.2013.12.005
22 P. Fernández-Berrocal et al. / Journal of Research in Personality 49 (2014) 21–24
not allow individuals to confront others who use the always defect
strategy. In order to examine whether high EI would lead individ-
uals to cooperate regardless of the other player’s behavior, or in-
stead allow them to respond differently to helpers and cheats,
we decided to analyze the strategies used by emotionally intelli-
gent individuals in different PDGs. We carried out our study using
an ability-based measure of EI (MSCEIT) in order to avoid reliance
on self-report measures, which can introduce error due to limited
or inaccurate self-knowledge, self-enhancement bias and other
distortions (Mayer et al., 2008; Schlegel, Grandjean, & Scherer,
2013).
We investigated the relationship between participants’ EI and
their behavior in a PDG in face-to-face interaction with a confeder-
ate. The confederate used a consistent and predetermined interac-
tion strategy (always cooperate, always defect or tit-for-tat), about
which participants were not forewarned. Statistical analyses
controlled for cognitive intelligence and tendency to cooperate.
Our hypothesis was that people with high EI would adapt their
response to the other person’s behavior (always cooperate, always
defect or tit-for-tat), thereby maximizing gains across conditions.
2. Method
2.1. Participants
The sample consisted of 232 university students (42 male and
190 female) with a mean age of 22 years (SD = 4.19), who partici-
pated in the study in return for course credit.
2.2. Procedures and measures
Data were collected in two phases. In the first phase, partici-
pants completed measures of EI, cognitive intelligence, tendency
to cooperate and socio-demographic information – during normal
lecture hours and in the presence of a researcher. In the second
phase, one month later, individual participants played a PDG with
a confederate in a laboratory room.
Emotional intelligence was measured using the Mayer, Salovey,
Caruso Emotional Intelligence Test (MSCEIT v.2.0; Mayer, Salovey,
& Caruso, 2002). This ability- and knowledge-based test of 141
items assesses the four dimensions of EI proposed by Mayer and
Salovey (1997): identifying, using, understanding and managing
emotions. Tasks range from identifying emotion in human faces
to identifying strategies for managing emotions in social situations.
The four dimensions are thought to contribute jointly to the quality
of interpersonal interaction and effective decision-making and
therefore can be subsumed into a general factor of EI (Mayer
et al., 2002). Thus, for the sake of parsimony, we report results
based on total scores rather than sub-dimensions. Mayer et al.
(2002) reported good reliability values for total MSCEIT scores,
including internal consistency (split-half reliability, .91) and test–
retest reliability (.86). We used scores based on consensus norms,
which map closely onto scores based on expert norms. In our study
the split-half reliability was .73, with M = 98.77 (SD = 15.55) for to-
tal MSCEIT score. For additional information on the psychometric
properties of the MSCEIT, see Mayer et al. (2002) and Mayer, Salo-
vey, Caruso, and Sitarenios (2003).
Cognitive intelligence. The D-48 Intelligence Test was used to as-
sess fluid intelligence (Gough & Domino, 1963). This general intel-
ligence test evaluates the capacity to conceptualize and apply
systematic reasoning to new problems. It contains 48 sets of dom-
inoes displayed in logical series. Each set includes a domino with
one blank face that respondents must fill in. We included cognitive
intelligence as a control variable in our analyses to show that the
effects observed are specific to emotional intelligence and cannot
be explained by other intelligence constructs. This control is espe-
cially important because previous research has shown that persons
with higher cognitive ability tend to be more cooperative in both
experimental and real-world Prisoner’s Dilemmas (Jones, 2008).
The D-48 Intelligence Test reveals adequate internal consistency
(.89) and test–retest reliability (.69), and no ethnicity or gender
bias (Domino & Morales, 2000). In the present study, the mean
score was 28.28 (SD = 5.33).
Tendency to cooperate versus compete was measured one month
before the real PDG by describing to participants the classic Pris-
oner’s Dilemma task for two players and showing the payoff matrix
(Axelrod, 1984; see Appendix). Participants were asked ‘‘Imagine
you are playing; what would be your first move: cooperate or com-
pete?’’ (Komorita, Hilty, & Parks, 1991). Responses were coded 0
for cooperate (61.2% of respondents) and 1 for compete. We in-
cluded tendency to cooperate versus compete as a control variable
in our analyses, because previous research in social dilemmas has
found that people with a tendency to cooperate attempt to maxi-
mize the joint outcome for both players (Fehr & Fischbacher,
2003; Nelissen et al., 2007).
Prisoner’s Dilemma Game. We used a two-player version of the
PDG (Axelrod, 1984). In any one round of this game, a participant
scores 5 points in s/he competes and the opponent cooperates, 3
points if both cooperate, 1 point if both compete, and 0 points if
s/he cooperates and the opponent competes (see Appendix). Partic-
ipants were randomly assigned to three experimental conditions
(dilemma type): always cooperate, always defect or tit-for-tat. In
the always cooperate condition, the confederate always cooperated,
regardless of the participant’s behavior, so that his/her first move
was to cooperate. In the always defect condition, the confederate al-
ways defected, regardless of the participant’s behavior, so that his/
her first move was to compete. In the tit-for-tat condition, the con-
federate used the tit -for- tat strategy, and his/her first move was to
cooperate.
Before the PDG began, participants were provided basic instruc-
tions and told that the aim of the game was to obtain as many
points as possible. For each round, the players (participant and
confederate) presented their choices simultaneously. Participants
did not know how many rounds they would play, and were told
that the game would be ended at random. In reality, the game
was always stopped after 20 rounds. The outcome of interest was
the total number of points obtained by each participant in the
PDG. However, the range of possible scores varied across condi-
tions. If the confederate always competed, the minimum score a
participant could obtain was 0 and the maximum 20. In the always
cooperate condition, the minimum was 60 and the maximum 100.
In order to obtain a meaningful common metric to analyze the ef-
fect of EI across conditions, we standardized PDG scores within
condition (M = .09; SD = 1.07).
3. Results
Intercorrelations between key variables, reported in Table 1,
indicate that high-EI individuals tended to score higher on the
PDG than their low-EI counterparts. Women scored slightly higher
than men on the MSCEIT. Other correlations were not significant.
To test our hypothesis, we used a General Linear Model (GLM)
approach for analysis of covariance (ANCOVA), focusing on the
effect of EI, after controlling for cognitive intelligence, tendency
to cooperate, sex, and age, as well as experimental condition.
Although ANCOVA showed a significant main effect of condition
taking into account these covariates (F(2, 222) = 10.77, p < .001,
g2 = .09), we did not examine this effect further because overall
mean differences across conditions had already been removed by
standardizing the DV within condition. In support of our
 P. Fernández-Berrocal et al. / Journal of Research in Personality 49 (2014) 21–24
Table 1
Correlations between measured variables.
Variable Correlations
1 2 3 4 5
1. Gender
2. Age .10
3. Emotional intelligence .15* .03
4. Cognitive intelligence .07 .00 .10
5. Tendency to cooperate .06 .01 .03 .06
6. PDG score .04 .03 .19** .09 .04
Note. N = 232. Gender was coded 0 for male and 1 for female. Tendency to cooperate
was coded 0 for cooperate and 1 for compete. Emotional intelligence = Mayer–
Salovey–Caruso Emotional Intelligence Test; cognitive intelligence = D-48 Intelli-
gence Test; PDG score = score on the Prisoner’s Dilemma Game (standardized
within experimental condition to yield a common metric).
*
p < .05.
**
p < .01.
hypothesis, we found a significant main effect of EI, indicating that
EI was positively related to PDG scores across conditions:
F(1, 222) = 5.75, p < .05, g2 = .02. No main effects were found for
cognitive intelligence, tendency to cooperate, sex or age.
An interaction effect was found for dilemma type and EI:
F(2, 222) = 10.17, p < .001, g2 = .08. To examine this interaction fur-
ther and identify the range of values for which the moderator has a
significant effect on the predictor, we applied the Johnson–Ney-
man technique, following the procedure of Hayes and Matthes
(2009). In the always defect condition, participants obtained higher
PDG scores if they had EI scores above 111 (b = 0.20, p < .05), and
lower PDG scores if they had EI scores below 95 (b = 0.18,
p < .05; Fig. 1).1
4. Discussion
A significant main effect of emotional intelligence indicates
that, overall, high-EI individuals did better on the PDG than their
low-EI counterparts, after controlling for cognitive intelligence
and tendency to cooperate. Moderation analysis indicates that this
performance advantage occurred only in the condition where the
confederate always competed. In this setting, PDG scores corre-
spond exactly to the number of times that a participant chose to
compete (because participants earned 1 point if they competed
and 0 if they cooperated in this game). Thus, our results show that,
in a highly competitive context, emotionally intelligent individuals
were more willing to compete in order to attain their goals and ad-
vance their interests than their low-EI counterparts. In other
words, high-EI individuals were less likely to be abused by
‘‘cheats’’.
In the condition where the confederate always cooperated, par-
ticipants could also earn more points by competing instead of
cooperating. In this condition, however, high-EI participants did
not score higher than their low-EI counterparts, indicating that
they were not more likely to compete. In other words, they did
not take advantage of someone who was consistently cooperative.
Previous research has suggested a linear relationship between
EI and cooperation: people with higher self-reported EI tend to
cooperate more (Schutte et al., 2001). However, cooperation is
not always the best strategy. Adapting one’s behavior in response
to others’ is arguably a better approach. The present findings show
1
We obtained similar results when we repeated the analysis without controlling
for cognitive intelligence and tendency to cooperate. Significant main effects were
found for dilemma type [F(2, 224) = 10.64, p < .001, g2 = .09] and EI [F(1, 224) = 6.38,
p < .05, g2 = .03], as well as an interaction effect for dilemma type and EI
[F(2, 224) = 9.95, p < .001, g2 = .08]. The pattern of interactions was similar to that
shown in Fig. 1.
23
Fig. 1. Interaction of dilemma type (AC, always cooperate; AD, always defect; TFT, tit-
for-tat) and emotional intelligence in predicting PDG scores.
that EI, assessed through an ability- and knowledge-based test
rather than by self-report, does not rigidly predispose people to-
ward cooperation regardless of others’ behavior (Côté, DeCelles,
McCarthy, Van Kleef, & Hideg, 2011; Hein & Singer, 2008). Instead,
EI appears to be associated with the capacity to flexibly adapt to
the other’s strategy and to the interaction context in order to attain
one’s goals (Tamir & Ford, 2012). This means competing when nec-
essary to establish or improve one’s position.
Why are people with low EI reluctant to compete when the
other person systematically refuses to cooperate (see Fig. 1)? They
may find it difficult to identify and interpret their feelings in emo-
tionally challenging situations; to use their feelings in order to
guide effective decision making; and to manage their own and oth-
ers’ emotions. Dealing with someone who systematically refuses to
collaborate is likely to arouse negative emotions such as anger, fear
or sadness (Tamir & Ford, 2012). An interesting avenue for future
research would be to examine participants’ emotional states while
they play the game, as a possible mediator of the relationship be-
tween EI and PDG scores.
Why are people with high EI not more likely to take advantage
of someone who is consistently cooperative? Perhaps they view
building a relationship and preserving their reputation as a fair
negotiator as more important – a view that is likely to enhance
benefits over the long term, at the expense of short-term gains.
Once again, future research could examine whether such goals ex-
plain the relationship between EI and performance in dealing with
a cooperative partner. This admittedly speculative interpretation
does not preclude the possibility that individuals who are emo-
tional intelligent and Machiavellian might take advantage of coop-
erative and credulous people (e.g., Côté et al., 2011).
In the condition where the confederate played tit-for-tat, PDG
scores are not directly proportional to the number of times that
participants chose to cooperate or compete. Here, the results of
the game depend on complex series of reactions and counter-reac-
tions. Overall, however, high-EI individuals tended to do well
across experimental conditions, whereas low-EI individuals did
poorly when faced with a fiercely competitive confederate. This
suggests that emotionally intelligent individuals can adapt to and
hold their ground in different contexts, adjusting their response
to the other person’s strategy. Interestingly, adaptability may de-
rive from the consistent use of a tit-for-tat strategy, in line with
the social norm of reciprocity – i.e., competing when others behave
competitively and cooperating when others behave cooperatively.
This response pattern could also be viewed as an adaptive facet of
personality, reflecting stable ‘‘if-then’’ situation-contingent re-
sponses (Mischel & Shoda, 1995).
24 P. Fernández-Berrocal et al. / Journal of Research in Personality 49 (2014) 21–24
Several limitations of our study should be noted. First, partici-
pants were university students, so we must be cautious about gen-
eralizing the results to the general population. Second, we did not
validate the D-48 test for our study population, given that we only
collected the number of total correct responses for each partici-
pant. Nevertheless, the D-48 has been validated in university-age
populations (Domino & Morales, 2000), so we consider it a reliable
assessment tool in this study. Third, while the PDG is one of the
most frequently used paradigms for simulating social decision-
making, its ecological validity is limited insofar as it involves an
artificial situation without serious personal consequences. Future
research should verify and explore our results in the context of
mixed-motive interpersonal negotiations with relevant conse-
quences, requiring both cooperation and competition.
Despite these limitations, the present study suggests that EI is
associated with the ability to both cooperate and compete, adjust-
ing one’s response to different social contexts flexibly so as to max-
imize individual and joint gains. This also implicates EI in the
ability to anticipate others’ decisions and emotions during deci-
sion-making in social contexts.
Acknowledgments
This research was supported in part by projects SEJ-07325 and
PSI2012-37490 (Spain).
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jrp.2013.12.005.
References
Aguirre, F., Sergi, M. J., & Levy, C. A. (2008). Emotional intelligence and social
functioning in persons with schizotypy. Schizophrenia Research, 104, 255–264.
Axelrod, R. (1984). The evolution of cooperation. New York: Basic Books.
Brackett, M. A., Rivers, S. E., Shiffman, S., Lerner, N., & Salovey, P. (2006). Relating
emotional abilities to social functioning: A comparison of self-report and
performance measures of emotional intelligence. Journal of Personality and
Social Psychology, 91, 780–795.
Brackett, M. A., Warner, R. M., & Bosco, J. (2005). Emotional intelligence and
relationship quality among couples. Personal Relationships, 12, 197–212.
Côté, S., DeCelles, K. A., McCarthy, J. M., Van Kleef, G. A., & Hideg, I. (2011). The Jekyll
and Hyde of emotional intelligence: Emotion-regulation knowledge facilitates
prosocial and interpersonally deviant behavior. Psychological Science, 22,
1073–1080.
Domino, G., & Morales, A. (2000). Reliability and validity of the D-48 with Mexican
American college students. Hispanic Journal of Behavioral Sciences, 22, 382–389.
Fehr, E., & Fischbacher, U. (2003). The nature of human altruism. Nature, 425,
785–791.
Gough, H. G., & Domino, G. (1963). The D-48 test as a measure of general ability
among grade school children. Journal of Consulting Psychology, 27, 344–349.
Hayes, A. F., & Matthes, J. (2009). Computational procedure for probing interactions
in OLS and logistic regression: SPSS and SAS implementations. Behavior Research
Methods, 41, 924–936.
Hein, G., & Singer, T. (2008). I feel how you feel but not always: The empathic brain
and its modulation. Current Opinion in Neurobiology, 18, 153–158.
Hirsh, J. B., & Peterson, J. B. (2009). Extraversion, neuroticism, and the prisoner’s
dilemma. Personality and Individual Differences, 46, 254–256.
Jacobs, M., Snow, J., Geraci, M., Vythilingam, M., Blair, R. J. R., Charney, D. S., et al.
(2008). Association between level of emotional intelligence and severity of
anxiety in generalized social phobia. Journal of Anxiety Disorders, 22, 1487–1495.
Jones, G. (2008). Are smarter groups more cooperative? Evidence from prisoner’s
dilemma experiments, 1959–2003. Journal of Economic Behavior & Organization,
68, 489–497.
Kee, K. S., Horan, W. P., Salovey, P., Kern, R. S., Sergi, M. J., Fiske, A. P., et al. (2009).
Emotional intelligence in schizophrenia. Schizophrenia Research, 107, 61–68.
Komorita, S. S., Hilty, J. A., & Parks, C. D. (1991). Reciprocity and cooperation in social
dilemmas. Journal of Conflict Management, 35, 494–518.
Lopes, P. N., Brackett, M. A., Nezlek, J., Schütz, A., Sellin, I., & Salovey, P. (2004).
Emotional intelligence and social interaction. Personality and Social Psychology
Bulletin, 30, 1018–1034.
Lopes, P. N., Salovey, P., & Straus, R. (2003). Emotional intelligence, personality and
the perceived quality of social relationships. Personality and Individual
Differences, 35, 641–658.
Mayer, J. D., Roberts, R. D., & Barsade, S. G. (2008). Human abilities: Emotional
intelligence. Annual Review of Psychology, 59, 507–536.
Mayer, J. D., Salovey, P., Caruso, D. R., & Sitarenios, G. (2003). Measuring emotional
intelligence with the MSCEIT V2.0. Emotion, 3, 97–105.
Mayer, J. D., Salovey, P., & Caruso, D. R. (2002). Mayer–Salovey–Caruso Emotional
Intelligence Test (MSCEIT) user’s manual. Toronto, ON: MHS.
Mayer, J. D., & Salovey, P. (1997). What is emotional intelligence? In P. Salovey & D.
J. Sluyter (Eds.), Emotional development and emotional intelligence: Educational
implications (pp. 3–31). New York: Basic Books.
Mischel, W., & Shoda, Y. (1995). A cognitive-affective system theory of personality:
Reconceptualizing situations, dispositions, dynamics, and invariance in
personality structure. Psychological Review, 102, 246–268.
Nelissen, R. M. A., Dijker, A. J., & De Vries, N. H. (2007). How to turn a hawk into a
dove and vice versa: Interactions between emotions and goals in a give-some
dilemma game. Journal of Experimental Social Psychology, 43, 280–286.
Reis, D. L., Brackett, M. A., Shamosh, N. A., Kiehl, K. A., Salovey, P., & Gray, J. R. (2007).
Emotional intelligence predicts individual differences in social exchange
reasoning. NeuroImage, 35, 1385–1391.
Schlegel, K., Grandjean, D., & Scherer, K. R. (2013). Constructs of social and
emotional effectiveness: Different labels, same content? Journal of Research in
Personality, 47, 249–253.
Schutte, N. S., Malouff, J., Bobik, C., Coston, T., Greeson, C., Jedlicka, C., et al. (2001).
Emotional intelligence and interpersonal relations. Journal of Social Psychology,
141, 523–536.
Tamir, M., & Ford, B. Q. (2012). When feeling bad is expected to be good: Emotion
regulation and outcome expectancies in social conflicts. Emotion, 12, 807–816.