Page Title

Emotional Intelligence as a Moderator of the Relationship between Conscientiousness and Performance

July 29, 2004 

"Changes in organizations, as well as the recent scientific developments in the area of positive organizational behavior (Luthans, 2002), have promoted the importance of social effectiveness in work organizations.  One type of social effectiveness construct that has received considerable attention recently is emotional intelligence.  Unfortunately, little support for criterion-related validity has been established to date for emotional intelligence, leaving in question its role in the prediction of important work outcomes (see Fox & Spector, 2000 for an exception).  Furthermore, personality scholars have argued that personality needs to be ignited or brought to life by social effectiveness constructs in order to demonstrate effects on performance and other outcomes (Hogan, 1991; Hogan & Shelton, 1998).

 Social effectiveness refers to the skills of reading and understanding others, and being able to utilize such knowledge to influence others in the pursuit of individual and/or organizational goals.  As noted by Ferris et al., social effectiveness reflects a class of constructs that are dynamic in nature, demonstrate facility in interpersonal interaction, yet are distinct from (although modestly correlated with) personality traits.  Certainly, emotional intelligence is one of the social effectiveness constructs that has received considerable attention recently in both the scientific and applied literatures.  However, little empirical research has been conducted and reported to date concerning the predictive ability of emotional intelligence, as either main effect or moderator.

 Emotional intelligence is a relatively new construct stemming from the increased interest in emotions in the workplace.  Initiated by Salovey and Mayer (1990), who perceived emotional intelligence as a subset of social intelligence, they defined emotional intelligence as 'the ability to monitor one's own and other's feelings and emotions, to discriminate among them, and to use this information to guide one's thinking and actions' (p. 189).

 Personality and social skills are different; personality traits are rather enduring dispositions while social skills are learnable.  Similar to other social effectiveness constructs, emotional intelligence is a hybrid construct touching both domains.  Spurred by Goleman's best-selling works (1995, 1998), emotional intelligence has risen in popularity over the last decade.  Viewed as being independent of IQ, Goleman (1995) has related emotional intelligence to knowing and managing emotions, self-motivation, and handling relationships.  More recently, research on the implications of emotional intelligence in the workplace suggests that emotional intelligence may have a substantial role in organizational activities.

 Also, the contemporary theories of leadership (i.e., transformational, and charismatic), because of their interpersonal connection with followers, are affected by a leader's emotional intelligence level.  Finally, in the work team development context, empirical research suggests that emotional intelligence might influence teamwork.

 Goleman (1998) combined a broad array of mental abilities and personality traits to form twenty-five emotional competencies ranging from political awareness to achievement drive in an attempt to measure emotional intelligence in the workplace. Some researchers believe that emotional intelligence is best captured by the mental 'ability-model,' which measures specific skills like emotional perception, emotional facilitation, emotional understanding and emotional management (Mayer, Salovey, & Caruso, 2000a, 2000b; Mayer & Salovey, 1997). 

 Working from the mixed-model approach, Bar-On (1997) developed the Emotional Quotient Inventory (EQ-i), a self-report measure designed to assess non-cognitive factors that influence the abilities to manage environmental demands.  The EQ-I contains five composite scales: intrapersonal, interpersonal, stress management, adaptability, and general mood.  Within the Bar-ON framework, the intrapersonal dimension related to internal self-awareness and self-management issues, and the interpersonal dimension is focused on interpersonal relationships with social interaction.  The adaptability dimension covers working through changing situations, while stress management centers on one's ability to handle stress, and general mood reflects one's happiness and optimism. 

 We selected the EQ-I measure for the current study for several reasons.  First, it has undergone an extensive validation process, and it takes a broad approach examining both emotional and social intelligence.  Also, it provides several dimensions that are critical to our work, intrapersonal, interpersonal, stress management and adaptability, which tap into the sub-dimensions of assertiveness, empathy, impulse control, and flexibility. 

 Research on the role of personality in the organizational sciences has evolved since the mid 1960s when Guion and Gottier (1965) issued a discouraging statement concerning the predictability of job performance in organizations from personality.  Beginning in the mid-1980s, the field witnessed a re-emergence of personality research in organizations, which has continued to the present time (e.g., George, 1992; Adler & Weiss, 1998).  Research to date has indicated that conscientiousness has been the most consistent predictor of job performance of all the five factors of personality (e.g., Barrick, Mount & Judge, 2001). 

 Recent personality theory has formulated a dual-component perspective on personality which argues for both internal and external components of personality.  The external view of personality is inherently flawed because observers tend to make the observed more responsible for his or her actions.  Conversely, the internal perspective more accurately reflects strategies a person develops to get along or get ahead, which are driven by the situation in which he or she is performing (Hogan, 1991).  The central notion here is that personality needs social skill or effectiveness in order for personality to realize its potential.

 Self-monitoring is a widely recognized social skill construct that has been noted as a factor that influences individual behavior and social effectiveness.  Self-monitoring reflects the extent to which individuals express and control self-presentation behavior (Snyder, 1987).  Previous research found that high self-monitors tend to emerge as group leaders, and may be less committed to current associates (Kilduff & Day, 1994).  Further, low self-monitors are more likely to exhibit consistent behavior, and received higher supervisory rating than high self monitors (Caliguri & Day, 2000). 

 Conscientiousness has demonstrated some evidence of performance prediction, but that evidence has been somewhat inconsistent.  Martocchio and Judge (1997) found conscientiousness to be related to self-deception, which in turn was negatively related to learning in training sessions.  Goleman (19998) indicated that personality traits should not necessarily be expected to lead to effectiveness in work and social settings unless individuals are attuned to the context, and are capable of making situational adjustments based on implicit social expectations.

 In light of the foregoing discussion, it seems reasonable to expect an interaction between conscientiousness and emotional intelligence on performance.  Hypothesis:  The relationship between conscientiousness and performance scores will be positive for employees high in emotional intelligence and negative among those low in emotional intelligence.

 We collected data from 205 students in two principles of management courses at a university located in the South.  Student participants were 57% male, predominantly white (86%) with an average of 21.2 years.  Students participated voluntarily, and earned extra course credit points for the involvement in the research.  The two courses were taught by the same instructor, and similar classroom management methods were used in both.   In the first week of the semester, students signed rosters identifying social group and athletic affiliations (e.g. fraternities and sororities, band, football, etc.).  From these groupings, they were randomly assigned to groups of seven students, first the males and then the females.  This process ensured that all groups had similar proportions of females and that things held in common (i.e., affiliations and friendships) were minimized, which mitigates ingroup-outgroup formation.

 The typical composition of the groups was very similar in terms of proportion of females (mean = 43%) and of those affiliated with social groups.  The students in each course were given a substantial amount of work to accomplish with their group members.  Group activities varied in objectives and nature, and were a consistent feature throughout the courses.  Activities consisted of short cases, quizzes, short (i.e., approximately one to tow weeks) projects, end of chapter questions and other applications of course materials.  Approximately 50% of the total class grad was based on these varied group activities.

 This context closely parallels real-world conditions in which individuals are interdependent for work performance and success, and that people seldom have choices regarding the individuals with whom they work.  In addition to the scores from their individual examinations and group activities, students completed three peer evaluations.  Students distributed an average of 100 points across the other members of their groups, depending on each person's level of contributions to the group's work.  The average score assigned to each student was converted to a percentage, and used as a multiplier for that student's activity scores.

 The psychometric measures were administered just past the midway point of the semester, and performance scores were collected throughout the semester.  To assess emotional intelligence, we used the Bar-On (1997) Emotional Quotient Inventory (EQ-i).  WE chose to focus on the Intrapersonal, Interpersonal, Stress Management, and Adaptability dimension scores of the EQ-i.  The scores on emotional intelligence represent the combined total for the four dimensions, and the higher scores reflect higher levels of emotional intelligences.  The self-report inventory consists of 152 items that ask participants about such things as impulsivity, satisfaction with financial status, satisfaction with personal appearance, anxiety tendencies, maintenance of friendships, self control of emotions, sensitivity to others, and patience, among others.  The internal consistency reliability estimate for this composite measure of the EQ-i scale was.87.

 Conscientiousness was measured using the self-report Conscientiousness scale from the NEO-PI Personality Inventory (McCrae & Costa, 1987).  The coefficient alpha internal consistency reliability estimate was.71.  Several other scales were administered to avoid signaling the intent of the research to the participants, including a locus of control scale, a self-monitoring scale, a work ethic scale, and the NEO-PI.

 Student performance was measured by the total points accumulated from exam scores and assigned peer ratings.  During the semester students were given two exams, a mid-term and final, each worth 100 points.   As noted by Ferris et al. (2002), the various social effectiveness constructs probably reflect some degree of overlap or shared criterion domain space. 

 We entered the three control variables (i.e., age, gender, and self-monitoring) at the first step of hierarchical regression analysis.  Next, main effects variables (i.e., conscientiousness and emotional intelligence) were entered in the second step.  IN the third step, we entered the cross-product term of conscientiousness x emotional intelligence to test the hypothesis that the interaction accounted for a statistically significant increment in the proportion of criterion variance explained beyond the control variables and main effects.

 Because self-monitoring and emotional intelligence are closely related constructs an additional step was take to substantiate the use of self-monitoring as a control variable. 

 The emotional intelligence score (shown as a table) is significantly related to conscientiousness.  The moderate level of this coefficient is consistent with what might be expected given convergent and disicriminant validity of these two constructs.  Also, conscientiousness is related to performance scores at a moderate, but significant level.

 Supporting the hypothesis, the addition of the conscientiousness x emotional intelligence cross-product term was significant in the explanation of criterion variance beyond that accounted for by the control variables and main effects. 

Workers who were high in both conscientiousness and emotional intelligence received the highest performance scores.  However, those high in conscientiousness but low in emotional intelligence received the lowest performance scores.  Therefore, in this study conscientiousness predicted increased performance for individuals scoring high on emotional intelligence, but it predicted lower performance scores for individuals with low EQ-i scores.

 Based on the notion advanced by Goleman (1998), that high conscientious workers without emotional intelligence may demonstrate ineffective work performance, we found that among highly conscientious workers, those high in emotional intelligence had higher performance scores than did those low in emotional intelligence, thus supporting the hypothesis.  IN contrast, highly conscientious workers high in emotional intelligence, who were able to properly calibrate their fastidious attention to detail at work with the savvy to know when and how to do so, had higher performance scores.  For individuals low in emotional intelligence, conscientiousness was associated with decreases in performance scores.  This suggests that conscientiousness, without the more proximal emotional intelligence savvy and skill needed to bring it to life, and regulate and channel it in appropriate ways in order to realize its potential, is not sufficient for a high level of performance.

 This study has both strengths and limitations that need to be noted.  We felt that one strength was that the interaction effect was significant even after controlling for both self-monitoring and demographic variables.  Second, we collected data from tow sources (i.e., workers and their co-workers), using two data collection techniques thereby reducing the probability that common method variance might confound the results (Podsakoff & Organ, 1986).  A third strength is that the use of a classroom environment provides more tightly controlled circumstances and more closely resembles quasi-experimental conditions, if not methods thereby reducing unexplained variance.

 However, there appear to be some limitations to the study as well.  Perhaps the one most in the forefront is the nature of the sample.  Using university students typically begs questions of generalizabilty, and the issue warrants discussion. 

 If social effectiveness constructs (e.g., emotional intelligence) can serve to properly regulate and calibrate personality traits, thereby activating them into productive action, then perhaps future research should investigate the potential moderating role of social effectiveness on the relationships between other personality characteristics and job performance.  Future research needs to move toward more precise articulation of the conditions under which specific dimensions of emotional intelligence would be expected to affect the relationships between personality traits and work outcomes (Douglas, Frink, & Ferris, 2004, p. 2-14)."

Reference:  Douglas, C., Frink, D.D., & Ferris, G.R. (2004).  Emotional intelligence as a moderator of the relationship between conscientiousness and performance.  Journal of Leadership and Organizational Studies, 2004, 10, (3).

-----------------------------------------------------------------------------------------------------------------

The Journal of Leadership and Organizational Studies is available on loan from the Ohio State University Leadership Center.  To borrow this resource or any other resource please go to the resource search page http://140.254.85.212/winnebago/search/search.asp?lib

If this is the first time you have borrowed resources from the OSU Leadership Center, please contact us at 614-292-3114 and we will register you in our library system.  Once you have been added to the system, you may request resources via the website.

Learn how the Ohio State University Leadership Center is strengthening tomorrow's leaders today at http://leadershipcenter.osu.edu

Back to Top >

Created: 2007-11-13, Updated: 2009-02-17