Emotional Intelligence and Its Measurement
Summary and Keywords
Emotional intelligence (EI) is used in organizational training, coaching, and graduate schools. Despite its acceptance in practical applications, researchers continue to argue about its validity. EI can be defined “as a constellation of components from within a person that enable self-awareness of and management of his/her emotions, and to be aware of and manage the emotions of others.” EI seems to exist at the performance trait or ability, self-schema and trait, and behavioral levels. Based on this multilevel view, all the conceptualizations of EI and the different measures that result are EI. Research on the behavioral level of EI—its assessment, strengths, psychometric validity, and challenges—complements that on other approaches, which have already been the subject of many academic papers.
Understanding human capability has been a quest for thousands of years (Boyatzis, 1982). The need for leaders or “citizens” to manage their emotions and build better relationships was noted by Plato and ancient Chinese scholars. But it was not until the 1920s that psychologist Edward Thorndike claimed that this complex set of a person’s actions might be one underlying concept. He conceived of it as some form of “social intelligence.” In 1990, Salovey and Mayer (1990) published the first paper using the term emotional intelligence (EI), and Daniel Goleman (1995) popularized the concept worldwide. Since then, a number of different concepts have used the label EI, and these are often anchored in different methods of assessment. Sharma (2008) and Fernandez-Berrocal and Extremera (2006) reviewed the history, beginning with Spinoza and other philosophers, before getting to Thorndike. In this article term emotional intelligence will apply to emotional and social intelligence.
In brief, Howard Gardner (1983) claimed that people had seven intelligences, of which two—intrapersonal and interpersonal—constituted what we now call EI. Salovey and Mayer (1990) described EI as having four components: perceiving, using, understanding, and managing emotions. Other models have claimed the domain of EI with a variety of labels, such as “practical intelligence” and “successful intelligence” (Sternberg, 1996), which seem to integrate what other psychologists call cognitive abilities and to anchor the concepts around the consequence of the person’s behavior, notably success or effectiveness. The behavioral level of EI has been discussed and documented by Boyatzis (1982, 2009, 2017), Cherniss (2010), and Cherniss and Boyatzis (2013). An early discussion of the criticism of EI is found in Matthews, Zeidner, and Roberts (2002). In their critique, they seemed to have conflated issues in theoretical models and their measurement.
A Multilevel Theory of EI
In an effort to explain a multilevel theory of emotional intelligence (EI), prominent published theories of EI and measures are displayed in a multilevel graphic (Figure 1). The levels are range from deeply unconscious to social-context sensitive and value-sensitive levels of self-schema to the behavioral level (i.e., what other people would see).
Before examining specific issues, it is useful to explore why and how these are conceptualizations of an emotional intelligence. Mayer, Salovey, and Caruso (1999) claimed three standards for an intelligence: “(1) it should reflect a “mental performance rather than preferred ways of behaving; (2) tests of it should show positive correlation with other forms of intelligence; and (3) the measures should increase with experience and age” (pp. 269–270).
Different interpretations of “intelligence” are offered in the literature. For example, Petrides and Furnham (2000, 2001) proposed differences between trait EI and ability EI. Because trait EI is measured through self-assessment, it is likely to be closer to the personality realm in terms of the Big Five traits. Boyatzis and Sala (2004) claimed that to be classified as an “intelligence,” the concept should be
1) behaviorally observable; 2) related to biological and in particular neural-endocrine functioning. That is, each cluster should be differentiated as to the type of neural circuitry and endocrine system involved; 3) related to life and job outcomes; 4) sufficiently different from other personality constructs that the concept adds value to understanding the human personality and behavior; and 5) the measures of the concept, as a psychological construct, should satisfy the basic criteria for a sound measure, that is show convergent and discriminant validity.
(Boyatzis & Sala, 2004, p. 148)
Mayer et al.’s criterion no. 3 would likely will be somewhat related to the third, fourth, and fifth criteria of Boyatzis and Sala (2004). But Mayer et al.’s (1999) first and second criteria claim that because EI is an “intelligence,” it should correlate with measures of cognitive intelligence.
The connection to neural functioning is another way to view why EI is an intelligence. Any conceptualization of EI should be linked, theoretically or empirically, to neural networks. In other words, the construct should actually be able to predict neural and endocrine (i.e., hormonal) patterns within the individual.
The need for Boyatzis and Sala’s (2004) criterion no. 2 (i.e., job and life outcomes) is rooted in relevance of research. The American Psychological Association’s Task Force on Intelligence (American Psychological Association Public Affairs Office, 1997) reported that predicting real life outcomes is an important part of the standard against which we should judge an intelligence. They added that there should be a consensus within a field about the definition. Although such a consensus is lacking for emotional intelligence, the link between EI competencies and real-life outcomes is, in fact, testable, as recent meta-analyses have shown (Joseph, Jin, Newman, & O’Boyle, 2014; O’Boyle, Humphrey, Pollack, Hawver, & Story, 2011). This seems a more relevant test of the concept than merely showing a link to experience and age (i.e., as Mayer et al.’s, 1999, third criterion does).
A Review of Measures of the Multiple Levels
Scholars have been writing about Streams 1, 2, 3 and 4 of EI and trying to relate which is more valid for a variety of outcomes and also shows convergent and discriminant validity. According to Ashkanasy and Daus (2005), Stream 1 consists of direct measures of a person’s intellectual facility with emotional information, as exemplified by the Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT). Stream 2 consists of self-assessment measures, based on the MSCEIT model, such as Schutte et al. (1998) and Law, Wong, and Song (2004). Stream 3 consisted of all other measures. Ashkanasy and Daus clustered self-assessment measures such as the Emotional Quotient Inventory 1 (EQ-I; Reuven Bar-On, 1997), the Trait Emotional Intelligence Questionnaire TEIQue (Petrides & Furnham, 2000, 2001), CII coded measures (Ryan, Emmerling, & Spencer, 2009; Spencer & Spencer, 1993), and the ESCI and its earlier versions (Boyatzis, 2017). This confounds the assessment of traits, self-schema, and behavior (McClelland, 1951). Self-assessments may examine trait-like EI or self-image, self-perception EI (called “self-schema” in Boyatzis, 1982).
Stream 4 measures a person’s behavior (Amdurer, Boyatzis, Saatcioglu, Smith, & Taylor, 2014). They are different from the self-perceptions of Stream 3. In Stream 4 measures, EI is defined and operationalized as distinct from other approaches and represents informants’ own observations or direct observations by others of a person’s behavior through 360-degree measures, or coding from audio tapes of critical incidents or videotapes of simulations. Stream 4 measures include 360-degree measures, such as the Emotional and Social Competence Inventory (Boyatzis & Goleman, 1996, 2007); Dulewicz, Higgs, and Slaski (2003) and Bar-On (1007) both have availability of a 360-degree version of their tests and the Wong and Law Emotional Intelligence Scale (WLEIS; Wilderom, Hur, Wiersma, Van Den Berg, & Lee, 2015).
Behavioral measures do not show consistent relationships to personality traits in terms of the Big Five (Boyatzis, 2017). Behavioral EI with the ESCI was tested with the Big Five measures; two studies showed that personality did not predict effectiveness measures (Boyatzis, Good, & Massa, 2012; Boyatzis, Rochford, & Cavanagh, 2017) or unique variance, whereas behavioral EI did. It is also worth noting that these same two studies also tested 360-degree EI with the Ravens Progressive Matrix. In both studies, g was not shown to account for any significant, unique variance, whereas behavioral EI did.
Two prominent meta-analyses of EI research measures have concluded that although all measures are significant, Stream 3 measures do a better job of predicting job performance (Joseph et al., 2014; O’Boyle et al., 2011) and subordinate job satisfaction (Miao, Humphrey, & Qian, 2016). Joseph et al. (2014) criticized Stream 3 measures on outcomes in their own and others’ meta-analyses because of the contamination of self-assessment measures with personality. None of these meta-analyses included Stream 4 as a separate measurement method.
Stream 1: Ability-Based Level of EI
Salovey and Mayer (1990) defined EI as “the subset of social intelligence that involves the ability to monitor one’s own and others’ feelings and emotions, to discriminate among them and to use this information to guide one’s thinking and actions” (p. 190). It is “emotional information processing” (Mayer et al., 1999). They clarified EI as “the ability to perceive and express emotion, assimilate emotion in thought, understand and reason.” In their theory (Mayer, Salovey, Caruso, & Sitarenios, 2003), EI has four dimensions: emotion perception, emotion understanding, emotional facilitation, and emotion regulation.
The MSCEIT is a direct performance assessment of emotional processing with some scenarios testing (Mayer et al., 1999). It uses consensus and expert scoring. As they claim in their theory, it correlates with cognitive ability, and some of the branches show a statistically more significant association (Joseph & Newman, 2010). The MSCEIT has content validity (Cherniss, 2010) and appropriate reliability (Mayer et al.,2003). Test-retest reliability has been estimated as r = .86 (n = 60; Brackett & Mayer, 2003). The overall internal consistency reliabilities are usually above .75, although the reliabilities of the subscales have not been as consistently good (Conte, 2005; Matthews et al., 2002).
A CFA did not support their 4-factor model as proposed (Gignac, 2005; Rossen, Kranzler, & Algina, 2008). That is, the empirical evidence is that in life, the various branches do not correspond to the theoretical model, which, by the way, is not unique to the MSCEIT. The MSCEIT has discriminant validity with tests of personality. MSCEIT scales appear to have the strongest correlations with agreeableness from the Big Five (r = .21 to .28). Correlations with the other four Big Five traits are less than .20 (Mayer, Roberts, & Barsade, 2008).
The low correlations between MSCEIT and relevant constructs threaten the convergent validity. The MSCEIT correlates with measures of verbal intelligence (r = .36) and other kinds of intelligence (r = .10 to .20). This is consistent with a measure that is supposed to be related to but distinct from other types of intelligence (Cherniss, 2010). Roberts et al. (2006) reported that the Japanese and Caucasian Brief Affect Recognition Test (JACBART) had low correlation with the overall MSCEIT and many subscales (r = .20 to .26) but did not correlate with the emotional perception scale at all (r = .00). The MSCEIT subscales had correlations with the Levels of Emotional Awareness Scale (LEAS) from .15 to .20 (Ciarrochi, Caputi, & Mayer, 2003) suggesting that the MSCEIT has weak or little convergent validity with other relevant emotional constructs.
The incremental validity of the MSCEIT is quantitatively supported. In a student sample, Rossen and Kranzler (2009) controlled for general mental ability and personality. They found the MSCEIT explained incremental variance to positive relationships with others and alcohol use.
The scoring system is complicated (MacCann & Roberts, 2008). It is claimed to be difficult to tell whether a test item is appropriately answered (Matthews et al., 2006). To resolve this, the authors took two approaches to scoring: consensus scoring and expert scoring. The two scoring approaches are highly correlated (r = .96 to .98), but “it is unclear whether a person who thinks about the emotional domain differently from experts or from the average of several peers is low on that ability or whether that person simply has a new (and perhaps better) way of thinking” (Murphy, 2006, p. 348). Most published studies use one or the other scoring system, but not both, and the consensus scoring shows the best consistent validity.
A criticism of the response format of the MSCEIT has been compared to knowledge tests of EI, which may not provide an assessment of a person’s actual ability (Cherniss, 2010). “The assessment of knowledge in the abstract does not reflect the live performance of EI in the rich social situation of real life. One might understand that smiling at someone can be an effective means of producing a positive emotional reaction, but recognizing in a live encounter the moment to smile and doing so in a way that does not seem false or insincere may well be a different ability” (Spector & Johnson, 2006, p. 335).
Stream 2: Self-Assessment Measures of the MSCEIT Model
The Schutte Self-Report Emotional Intelligence Test (SREIT) is a self-report scale to measure “a homogeneous construct of emotional intelligence” based on Mayer-Salovey-Caruso model (Schutte et al., 1998). It has 33 items. The SREIT is a face-valid and content-valid measure of EI (Petrides & Furnham, 2000). It has internal consistency (i.e., Cronbach’s alpha) of .90 for the 33 items (Schutte et al., 1998). A two-week test-retest reliability is sound (r = .78; Conte & Dean, 2006). Discriminant validity has various results. A study with 23 college students showed that SREIT’s correlation with openness was high (r = .54), but the absolute correlations with other Big Five personality traits were lower (r = .21 to .28; Schutte et al., 1998). In another study, the SREIT’s correlations with the Big Five were from .18 (agreeableness) to .51 (extraversion), and it was not related to cognitive ability (Saklofske, Austin, & Minski, 2003). A criticism is that Schutte failed to show that the three subscales of EI were different (Petrides & Furnham, 2000). “It would have been more appropriate to perform a factor analysis on the 62 items, rather than a component analysis” (Gignac, Palmer, Manocha, & Stough, 2005, p. 1030).
The Wong and Law test of EI (WLEIS) is also a Stream 2 measure in that it is based on the four scales of the MSCEIT (Law, Wong, & Song, 2004; Wong & Law, 2002). The WLEIS test has 16 items. In the two samples, the authors showed convergent, discriminant, and incremental validity of this 16-item EI scale (Wong & Law, 2002). They reported incremental predictive power over life satisfaction. After controlling for Big Five traits, the student and the work samples showed additional variance (Law, Wong, & Song, 2004). All of the criticisms of self-assessment explained in the next section, “Stream 3: Self-Schema Self-Perception Level of EI,” also apply to these Stream 2 measures.
Stream 3: Self-Schema Self-Perception Level of EI
Bar-On’s (1997) model is “an array of non-cognitive capabilities, competencies, and skills that influence one’s ability to succeed in coping with environmental demands and pressures” (p. 14). The components included five forms of EI: intrapersonal intelligence, interpersonal intelligence, adaptability, stress management, and general mood (Bar-On, 1997, 2006). The Emotional Quotient Inventory (EQ-i) is composed of 15 scales in four composites: self-perception includes emotional self-awareness, self-regard, and self-actualization; self-expression includes assertiveness, emotional expression, and independence; interpersonal includes empathy, interpersonal relationships, and social responsibility; decision-making includes problem-solving, reality testing, and impulse control; stress management includes optimism, flexibility, and stress tolerance. The original EQ-i was a self-report; the 360-degree version was introduced in 1997 (Bar-On, 1997).
Most if not all of the research reported in journals and book chapters to date is from the self-assessment version. The self-assessment survey asks persons to describe their own thoughts, feelings, values, or behavior. It is self-report and therefore included in this section as a Stream 3 measure. To the extent the published research appears using the 360-degree version or Bar-On changes his model, it could be included in the subsequent Stream 4 behavioral level as well.
The EQ-i has good internal consistency. With 243 university students, the internal reliability ranged from .69 to .96, with an overall estimate of .96 (Dawda & Hart, 2000). Another study showed reliability of .97, and a six-month test-retest reliability of .72 for men (n = 73) and .80 for women (n = 279; Bar-On, 2004). However, the internal structure of EQ-i does not appear to be consistent with the model. An exploratory factor analysis (EFA) with a varimax rotation generated a 13-factor model instead of the 15-factor model in the Bar-On theory. After removing the problematic factors, a confirmatory factor analysis (CFA) generates a 10-factor model (Bar-On, 2006).
Bar-On’s theory was “designed to examine . . . a conceptual model of emotional and social functioning” (Bar-On, 2006, p. 15). Criticism of this measure has been focused on its broad conceptualization (Murphy, 2006). Others have claimed that the EQ-i has a great deal of overlap with personality, and that “predictive validity may simply be a consequence of the EQ-i functioning as a proxy measure of personality” (Matthew et al., 2002, p. 16). The content validity of EQ-i seems less strong than other measures because it includes nonability personality traits and ignores some EI components, such as emotional understanding and emotional perception (Cherniss, 2010).
An empirical study confirmed that the EQ-i has convergent validities “with respect to measures of normal personality, depression, somatic symptomology, intensity of affective experience and alexithymia” (Dawda & Hart, 2000, p. 797). The EQ-i correlated with other self-report EI measures (r = .58 to .69; Bar-On, 2004). The EQ-i does not have good discriminant validity with personality traits. Bar-On (2006, p. 16) mentioned that the EQ-i overlaps with personality “probably no more than 15% based on eight studies in which more than 1,700 individuals participated.” However, one study showed that the Big Five predicts EQ-i scores with a multiple correlation of .79. This suggests that the Big Five accounted for the majority of variance in the EQ-i (Grubb & McDaniel, 2007).
Another Stream 3 measure based on a composite of all of the major models is the TEIQue by Petrides and Furnham (2000, 2001), Petrides, Fredrickson, and Furnham (2004). They claim to assess “trait EI,” which they say is “a constellation of emotion-related self-perceived abilities and dispositions located at the lower levels of personality hierarchies” (Petrides & Furnham, 2000). It is meant to include all “personality facets that are specifically related to affect” (Petrides, Pita, & Kokkinaki, 2007, p. 274). The TEIQue has four subdimensions: emotionality, self-control, sociability, and well-being.
After controlling for the Big Five, the TEIQue has a positive relationship with happiness (Furnham & Petrides, 2003). It is linked with distinctive reactivity to affect-laden information and has incremental validity over the Big Five (Petrides, Frederickson, & Furnham, 2004; Petrides & Furnham, 2003). The TEIQue also has incremental validity over alexithymia and optimism (Mikolajczak, Luminet, Leroy, & Roy, 2007; Mikolajczak, Luminet, & Menil, 2006). Criticisms of the TEIQue are similar to those raised about self-assessment in general.
Stream 4 as Behavioral EI
The behavioral approach to EI began in 1974 with inductive studies of work samples that were selected on the basis of performance differentiations. Previously, assessment centers began coding simulations. The first studies identified both outstanding and “average” performers in a job, often based on nominations (Lewin & Zwany, 1976). The nominations were from bosses, peers, and subordinates (Boyatzis, 1982). Other criteria might be collected, such as climate surveys of subordinates. The outstanding group appeared on multiple lists from each of the sources. The average group was randomly selected from all those with no nominations from any source. Because the inductive approach was used, the cultural and language biases often found in questionnaires were avoided. Then, a variation on the critical-incident interview was used to recreate specific work events (Boyatzis, 1982; Flanagan, 1954; Spencer & Spencer, 1993).
The behavior that differentiated the two groups in the work setting were identified. The related behaviors were organized based on an underlying intent. This became the definition of the competency (Boyatzis, 2009). A set of competencies found to differentiate the outstanding from the average performers was compiled into a codebook. Since the behavior was shown in the person’s actual work situation, the competencies in the codebook were content valid.
Coding was done by reliable coders. Each coder tended to spend 2 to 3 times the running time of audio or videotape. The coders should be checked for their inter-rater reliability regularly. Because the behavioral coding is a highly labor intensive and therefore costly process, it created a desire for a multisource assessment, called 360 degree. As reflective items, the behaviors and underlying intent in each competency were converted to questionnaire items.
The “other” or informant responses to the 360 assessment provides a view of how others see the focal person act in various situations. In some applications, bosses, peers, and subordinates are solicited as informants. In other studies, those sources may be supplemented by clients or customers, spouses or partners, and friends.
The generation of the items in the behavioral approach is important because it explains why there is not the same pattern of statistical relationship to personality traits found in self-assessment (Joseph et al., 2014) or statistical relationship to g or cognitive intelligence measures found with ability assessments, such as the MSCEIT.
The Emotional and Social Competency Inventory (ESCI) was developed from the codebooks from hundreds of behavioral inductive studies, and items were generated for what appeared to be the most generic competencies in EI. The model has 12 competencies in four clusters representing emotional and social intelligence with two clusters each (Boyatzis, 2009; Boyatzis & Goleman, 2007), as shown in Table 1.
Table 1. The Scales and Clusters of the Emotional and Social Competency Inventory (ESCI)
Emotional Intelligence competencies:
Social Intelligence competencies:
Boyatzis (2017, p. 289) reported:
Analysis of 5,761 self-assessments and 62,297 “Other” assessments of the ESCI, as well as 1,629 self-assessments and 21,288 “Other” assessments of the ESCI-U (Boyatzis, Gaskin, & Wei, 2015) has shown internal consistency, reliability, factor structure, and construct validity, all within acceptable parameters. Both the Other versions and the self-assessment versions (i.e., ESCI and ESCI-U) demonstrated appropriate factor loadings for each item on each scale exploratory factor analyses (EFAs), model fit to rigorous standards for each scale in confirmatory factor analyses (CFAs), and convergent and discriminant validity against appropriate criteria for each scale within each version.
Studies reporting significant prediction or association with effectiveness and a variety of desired-outcome measures of leaders, managers, and professionals include (a) Bajaj and Medury (2013), whose work with Indian software managers (n = 156) predicted leadership effectiveness and transformational style using the Multifactor Leadership Questionnaire (MLQ); (b) Pardasani (2016), who showed that subordinates’ views of their leaders’ ESCI predicted that leader’s psychological well-being, engagement, and organizational virtuousness in 222 Indian dyads; (c) Miller (2014), who studied next-generation leaders in family businesses; (d) Mahon, Taylor, and Boyatzis (2014), who showed that behavioral EI among 231 knowledge workers organized in teams predicted their engagement; (e) Babu (2016), who showed that for 100 community college presidents, behavioral ESI predicted cognitive and emotional engagement and effectiveness; and (f) subordinates’ observations of 105 research and development executives from high-tech companies (Kendall, 2016), which predicted exploratory and exploitative product innovation, product innovation success, competitive organizational performance, and relative market share.
One major criticism of EI is the overlap with general mental ability (i.e., g or cognitive intelligence) and personality.
A study of engineer’s effectiveness at the engineering division of a major automotive company showed that ESCI scores as seen by peers in their project teams significantly predicted effectiveness with unique variance (ΔR2 = .31), whereas general mental ability (g) and personality did not (Boyatzis, Rochford, & Cavanaugh, 2017). A study of financial service sales executives showed that ESCI as seen by others significantly predicted leadership effectiveness above the impact of g and personality (ΔR2 = .03; Boyatzis, Good, & Massa, 2012). These studies are important because they show that behavioral ESCI (i.e., as observed by others) is a more powerful predictor of real-world outcomes than g as measured by Ravens Progressive Matrices in these two studies and personality as measured through the NEO-R and NEO-FFI.
(Boyatzis, 2017, p. 290)
A Bayesian analysis by Boyatzis, Batista, Fernandez-Marin, and Truninger (2015) reported that distributions of g (i.e., as assessed through Graduation Management Admission Test, [GMAT] scores) showed that behavioral EI was clearly different from g. Stream 4, behavioral EI is neither consistently related to measures of cognitive intelligence nor a reflection of it.
Self-Assessment Versus Other Reports (i.e., 360)
Problems of self-assessment threaten Stream 2 and 3 measures of EI, but not Streams 1 and 4. Podsakoff and Organ (1986), Taylor (2014), and Tsui and Ohlott (1988) reported that leaders’ self-assessment of their behavior and skills was inadequate. Scholars considered whether there was contamination because of narcissistic delusion. When behavioral measures are based on inductive derivation of the items in a 360, it results in competencies that should be observable by others. This avoids many of the possible biases of self-assessment. These findings raise questions about whether self-assessment and other-assessments relate to similar outcomes.
In a major study of next-generation leaders in family businesses, Miller (2014) examined 100 leaders from 100 family businesses. Among other variables, Miller (2014) collected 360-degree data from 350 family and nonfamily business executives. EI was assessed with the ESCI (Boyatzis & Goleman, 2007). Leadership effectiveness was assessed by others, using the Leadership Effectiveness Scale (Denison, Hooijberg, & Quinn, 1995), and engagement was measured with the Utrecht Work Engagement Scale (Schaufeli & Bakker, 2004). All scales showed appropriate psychometrics in EFAs, CFA model fit, and appropriate indexes of convergent and discriminant validity (Miller, 2014).
The Structural Equation Model (SEM) showed different impact of the behavioral and self-assessed EI. Behavioral EI had a highly significant relationship (β = .64***, n = 100) with perceived leadership effectiveness and no relationship with engagement (Miller, 2014). Self-assessed EI showed a significant relationship with engagement (β = .48***, n = 100) but showed no relationship to perceived leadership effectiveness. When Miller calculated a measure of self-awareness as self-assessed EI minus the other-assessed EI, it showed no relationship to perceived leadership effectiveness but showed a significant, negative relationship with engagement (β = -.21**, n = 100). One possible problem is source bias because the data collection of the behavioral EI and effectiveness measures were both from others, whereas the self-assessed EI and engagement of the leader were from the leader him- or herself. This would not have the same bias on the self-awareness variable.
Miller (2014) entered a common latent factor in the SEMs. He then compared the SEM loadings with and without the common latent factor to test common source bias (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). No differences beyond the accepted indicator of .20 loadings were found. So common source bias was not a likely influence on these results. It seems that behavioral EI measures would have a stronger and more consistent relationship to life and work outcomes than self-assessment measures.
EI expresses itself through our relationships. Relationships are the contexts in which EI emerges, and they help us understand why a person may act somewhat differently in different situations—different people are present, and the person has different relationships with them. In one application, the quality of our relationship with someone trying to help us would make a difference in the utility of EI (Rogers, 1961). Ellen Brooks Van Oosten (2013) found this effect in her study of 85 top executives in a major bank. She found that two clusters of behavioral EI predicted three different performance measures. She called one cluster “Emotional Acumen”; it consisted of accurate self-assessment, empathy, emotional self-awareness, emotional self-control, teamwork, transparency, and optimism. It predicted the executive’s degree of engagement and career satisfaction. She called the other cluster “Change Leadership”; it consisted of achievement orientation, change catalyst, initiative, inspirational leadership, and self-confidence. It predicted the performance ratings of that executive as a boss, but negatively predicted the executive’s own career satisfaction.
She also reported that the relationships executives had with their coaches in terms of shared vision, compassion, and positive mood significantly influenced leadership effectiveness. Specifically, the quality of the relationship with the coach moderated the impact of EI on effectiveness. The same characteristic of relationships within the family business had a moderating effect on leadership effectiveness and engagement for next-generation family business leaders (Miller, 2014).
A study of knowledge workers in a manufacturing and a consulting company showed that the degree of perceived shared vision in their relationships moderated the impact of behavioral EI on engagement (Mahon, Taylor, & Boyatzis, 2014). Babu’s (2016) study of community-college presidents showed that shared vision and compassion partially mediated the effect on engagement of a president’s EI.
Applications of EI at the Behavior Level and Other Levels
The popularity of EI in applied settings has driven a great deal of the research interest. Fortunately, the research has supported the validity of the measures. Careful research on the efficacy of applications is emerging, but it is slow.
Streams 2 and 3, self-assessment measures, have been used in many training programs and with coaching. They are convenient to assess, face valid to the client or learner, and do not require any processing. Meanwhile, planning the logistics of behavioral measures, whether coded audio- and videotapes or 360-degree measures requires more time and effort. Nonetheless, the behavioral EI measures, Stream 4, are used in many training programs and coaching, as well as in higher education.
The behavioral-level measures provide clients or learners with feedback and guidance about what they are doing and ideas for other forms of behavior (Cherniss, 2010). Evidence suggests that some personality traits are malleable with specific interventions (Roberts et al., 2017), but changing one’s traits seems to be a long-term prospect. Psychologists contend that one can much more easily change one’s behavior (Kanfer & Goldstein, 1991). This was shown in the therapeutic field with cognitive behavioral therapy (i.e., focusing on specific behavior changes) versus other approaches to psychotherapy (Barlow, 1988; Hubble, Duncan, & Miller, 1999).
In many countries, the behavioral-level measures are a key element in the development-training programs of many companies and government agencies and are used as well by coaches. In a review of studies of changes in behavioral EI worldwide, published between 1950 and 1996, and conducted with scientific rigor, the Consortium for Research on Emotional Intelligence in Organizations reported finding 15 model programs (Cherniss & Adler, 2000). Of those, only five continued to be in use in 1996. The two were the programs for EI competency development at Alverno College in Milwaukee, Wisconsin (Mentkowski et al., 2000), and Case Western Reserve University (CWRU, Cleveland, Ohio, USE) (Boyatzis, Stubbs, & Taylor, 2002). Behavioral EI measures are used in many colleges and universities in many countries in undergraduate and graduate courses on leadership in management, medicine, law, dentistry, engineering, arts and sciences, social work, and other programs. Management schools also use behavioral EI in executive-education and continuing-education programs.
Statistically significant improvement of students’ behavior has been shown in many articles in a variety of these fields since 1996 (Boyatzis et al., 2002). A summary of 32 longitudinal cohort studies conducted with full-time MBA candidates from just one school appears in Boyatzis and Cavanagh (2018).These studies reveal a dramatic increase in positive impact compared to other corporate training or management education, and the changes appear to be sustained as far out as five to seven years (Boyatzis, 2008). The EQ-I (Stream 3) is also used in many government and corporate training programs. The MSCEIT is also used in programs designed to improve functioning in educational institutions (Nathanson, Rivers, Flynn, & Brackett, 2016).
Although some practitioners seek a way to use EI measures in selecting among applicants for jobs that require EI, such as management and service jobs, most consultants avoid such applications for legal reasons. Any test or measure that can withstand being contested in lawsuits as discriminatory would have to have been reviewed research showing no discriminatory bias. Streams 2 and 3, as self-assessment measures, are perceived to be subject to intentional manipulation or falsification by job applicants, whether because the applicants want to show their social desirability and falsify their responses or manage their self-presentation to the prospective employer; this results in less confidence in such measures. At the same time, the history of behavioral EI includes assessment centers (Thornton & Byham, 1982), which have been used heavily in selection applications. But this has been known to result in lawsuits and lengthy legal processes when used with public servants or unionized jobs.
A comprehensive view of emotional intelligence (EI) includes multiple levels. Theories and research using EI would be enhanced by including the behavioral level as a distinct approach to assessing EI. Using all levels would offer stronger unique variance in predicting job and life outcomes, performance, engagement, citizenship, and innovation.
Amdurer, E., Boyatzis, R. E., Saatcioglu, A., Smith, M., & Taylor, S. N. (2014). Longitudinal impact of emotional, social and cognitive intelligence competencies and GMAT on career and life satisfaction and career success. Frontiers in Psychology, 5, article 1447, 1–15.Find this resource:
American Psychological Association Public Affairs Office. (1997). Intelligence: Knowns and unknowns. Washington, DC: American Psychological Association.Find this resource:
Ashkanasy, N. M., & Daus, C. S. (2005). Rumors of the death of emotional intelligence in organizational behavior are vastly exaggerated. Journal of Organizational Behavior, 26, 441–452.Find this resource:
Babu, M. (2016). Characteristic of effective leadership of community college presidents. (Unpublished doctoral dissertation). Cleveland, OH: Case Western Reserve University.Find this resource:
Bajaj, B., & Medury, Y. (2013). Relationship between emotional and social competencies and transformational leadership styles. International Journal of Research in Commerce and Management, 3(2), 56–59.Find this resource:
Barlow, D. H. (1988). Anxiety and disorders: The nature and treatment of anxiety and panic. New York, NY: Guilford.Find this resource:
Bar-On, R. (1997). Bar-On Emotional Quotient Inventory: Technical manual. Toronto, Canada: Multi-Health Systems.Find this resource:
Bar-On, R. (2004). The Bar-On Emotional Quotient Inventory (EQ-i): Rationale, description and summary of psychometric properties. In G. Geher (Ed.), Measuring emotional intelligence: Common ground and controversy (pp. 111–142). Hauppauge, NY: Nova Science.Find this resource:
Bar-On, R. (2006). The Bar-On model of emotional-social intelligence (ESI) 1. Psicothema, 18(Suppl.), 13–25.Find this resource:
Boyatzis, R. E. (1982). The competent manager: A model for effective performance. New York, NY: John Wiley & Sons.Find this resource:
Boyatzis, R. E. (2008). Leadership development from a complexity perspective. Consulting Psychology Journal, 60(4), 298–313.Find this resource:
Boyatzis, R. E. (2009). A behavioral approach to emotional intelligence. Journal of Management Development, 28, 749–770.Find this resource:
Boyatzis, R. E. (2017). Commentary on Ackley (2016): Updates on the ESCI as the behavioral level of emotional intelligence. Consulting Psychology Journal, 68(4), 287–293.Find this resource:
Boyatzis, R. E., Batista, J. M., Fernandez-Marin, X., & Truninger, M. (2015). EI competencies as a related but different characteristic than intelligence. Frontiers in Psychology, 6(Art. 72), 1–14.Find this resource:
Boyatzis, R. E., & Cavanagh, K. V. (2018). Leading change: Developing emotional, social, and cognitive intelligence competencies in managers during an MBA program. In K. V. Keefer, J. D. A. Parker, & D. H. Saklofske (Eds.), Handbook of emotional intelligence in education. Springer Series on Human Exceptionality. New York, NY: Springer.Find this resource:
Boyatzis, R. E., Gaskin, J., & Wei, H. (2015). Chapter 17: Emotional and social intelligence and behavior. In D. Princiotta, S. Goldstein, & J. Naglieri (Eds.), Handbook of intelligence: Evolutionary theory, historical perspective, and current concepts (pp. 243–262). New York, NY: Springer.Find this resource:
Boyatzis, R. E., & Goleman, D. (2007). Emotional and social competency inventory. Boston, MA: Hay Group.Find this resource:
Boyatzis R. E., Goleman D. (1996). Emotional Competency Inventory. Boston, MA: The Hay Group.Find this resource:
Boyatzis R. E., Goleman D. (1999). Emotional Competency Inventory. Boston, MA: The Hay Group.Find this resource:
Boyatzis R. E., Goleman D., Hay Acquisition (2002). Emotional and Social Competency Inventory. Boston, MA: The Hay Group.Find this resource:
Boyatzis R. E., Goleman D., Hay Acquisition (2007). Emotional and Social Competency Inventory. Boston, MA: The Hay Group.Find this resource:
Boyatzis, R. E., Good, D., & Massa, R. (2012). Emotional, social and cognitive intelligence as predictors of sales leadership performance. Journal of Leadership and Organizational Studies, 19, 191–201.Find this resource:
Boyatzis, R. E., Rochford, K., & Cavanagh, K. (2017). The role of emotional and social intelligence competencies in engineer’s effectiveness and engagement. Career Development International, 22(1), 70–86.Find this resource:
Boyatzis, R. E., & Sala, F. (2004). Assessing emotional intelligence competencies. In Glenn Geher (Ed.), The Measurement of Emotional Intelligence (pp. 147–180). Hauppauge, NY: Nova Science.Find this resource:
Boyatzis, R. E., Stubbs, L., & Taylor, S. (2002). Learning cognitive and emotional intelligence competencies through graduate management education. Academy of Management Journal on Learning and Education, 1(2), 150–162.Find this resource:
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.Find this resource:
Campbell, D. T., & Fiske, D. W. (1959) Convergent and discriminant validation by the multitrait multimethod matrix. Psychological Bulletin, 56, 81–105.Find this resource:
Cherniss, C. (2010). Emotional intelligence: Towards clarification of a concept. Industrial and Organizational Psychology: Perspectives on Science and Practice, 3, 110–126.Find this resource:
Cherniss, C., & Adler, M. (2000). Promoting emotional intelligence in organizations: Make training in emotional intelligence effective. Washington, DC: American Society for Training and Development.Find this resource:
Cherniss, C., & Boyatzis, R. E. (2013). Using a multilevel theory of performance based on emotional intelligence to conceptualize and develop “soft” leader skills. In R. Riggio & S. Tan (Eds.), Building interpersonal skills in management programs (pp. 53–72). New York, NY: Routledge.Find this resource:
Ciarrochi, J. V., Caputi, P., & Mayer, J. D. (2003). The distinctiveness and utility of a measure of trait emotional awareness. Personality and Individual Difference, 34, 1477–1490.Find this resource:
Conte, J. M. (2005). A review and critique of emotional intelligence measures. Journal of Organizational Behavior, 26, 433–440.Find this resource:
Conte, J. M., & Dean, M. A. (2006). Can emotional intelligence be measured? In K. R. Murphy (Ed.), A critique of emotional intelligence: What are the problems and how can they be fixed? (pp. 59–77). Mahwah, NJ: Lawrence Erlbaum.Find this resource:
Dawda, D., & Hart, S. D. (2000). Assessing emotional intelligence: Reliability and validity of the Bar-On Emotional Quotient Inventory (EQ-i) in university students. Personality and Individual Differences, 28, 797–812.Find this resource:
Denison, D. R., Hooijberg, R., & Quinn, R. E. (1995). Paradox and performance: Toward a theory of behavioral complexity in managerial leadership. Organization Science, 6, 524–540.Find this resource:
Dulewicz, V., Higgs, M., & Slaski, M. (2003). Measuring emotional intelligence: Content, construct and criterion-related validity. Journal of Managerial Psychology, 18, 405–420.Find this resource:
Fernández-Berrocal, P., & Extremera, N. (2006). Emotional intelligence: A theoretical and empirical review of its first 15 years of history. Psicothema, 18, 7–12.Find this resource:
Flanagan, J. C. (1954). The critical incident technique. Psychological Bulletin, 51, 327–335.Find this resource:
Furnham, A., & Petrides, K. V. (2003). Trait emotional intelligence and happiness. Social Behavior and Personality: An International Journal, 31, 815–823.Find this resource:
Gardner, H. (1983). Frames of mind: The theory of multiple intelligences. New York, NY: Basic.Find this resource:
Gignac, G. E. (2005). Evaluating the MSCEIT V2. 0 via CFA: Comment on Mayer et al. (2003). Emotion, 5, 233–235.Find this resource:
Gignac, G. E., Palmer, B. R., Manocha, R., & Stough, C. (2005). An examination of the factor structure of the Schutte Self-Report Emotional Intelligence (SSREI) scale via confirmatory factor analysis. Personality and Individual Differences, 39, 1029–1042.Find this resource:
Goleman, D. (1995). Emotional intelligence. New York, NY: Bantam.Find this resource:
Grubb III, W. L., & McDaniel, M. A. (2007). The fakability of Bar-On’s emotional quotient inventory short form: Catch me if you can. Human Performance, 20(1), 43–59.Find this resource:
Hubble, M. A., Duncan, B. L., & Miller, S. D. (Eds.). (1999). The heart and soul of change: What works in therapy. Washington, DC: American Psychological Association.Find this resource:
Joseph, D. L., & Newman, D. A. (2010). Emotional intelligence: An integrative meta- analysis and cascading model. Journal of Applied Psychology, 95, 54–78.Find this resource:
Joseph, D., Jin, J., Newman, D., & O’Boyle, E. H. (2014). Why does self-reported emotional intelligence predict job performance? A meta-analytic investigation of mixed EI. Journal of Applied Psychology, 100, 298–342.Find this resource:
Kanfer, F. H., & Goldstein, A. P. (Eds.). (1991). Helping people change: A textbook of methods (4th ed.). Boston, MA: Allyn and Bacon.Find this resource:
Kendall, L. (2016). A theory of micro-level dynamic capabilities: How technology leaders innovate with human connection (Unpublished doctoral dissertation). Cleveland, OH: Case Western Reserve University.Find this resource:
Law, K. S., Wong, C., & Song, L. J. (2004). The construct and criterion validity of emotional intelligence and its potential utility in management research. Journal of Applied Psychology, 87, 483–496.Find this resource:
Lewin, A. Y., & Zwany, A. (1976). Peer nominations: A model, literature critique and a paradigm for research. Springfield, VA: National Technical Information Service.Find this resource:
MacCann, C., & Roberts, R. D. (2008). New paradigms for assessing emotional intelligence: Theory and data. Emotion, 8, 540–551.Find this resource:
Mahon, E. G., Taylor, S. N., & Boyatzis, R. E. (2014). Antecedents of organizational and job engagement: Exploring emotional and social intelligence as moderators. Frontiers in Psychology, 5, 129–139.Find this resource:
Matthews, G., Emo, A. K., Funke, G., Zeidner, M., Roberts, R. D., Costa, P. T., Jr., & Schulze, R. (2006). Emotional intelligence, personality, and task-induced stress. Journal of Experimental Psychology: Applied, 12(2), 96–107.Find this resource:
Matthews, G., Zeidner, M., & Roberts, R. D. (2002). Emotional intelligence. Science and myth. Cambridge, MA: MIT Press.Find this resource:
Mayer, J. D., Roberts, R. D., & Barsade, S. G. (2008). Human abilities: Emotional intelligence. Annual Review of Psychology, 59, 507–536.Find this resource:
Mayer, J. D., Salovey, P., & Caruso, D. R. (1999). Emotional intelligence meets traditional standards for an intelligence. Intelligence, 2, 267–298.Find this resource:
Mayer, J. D., Salovey, P., Caruso, D. R., & Sitarenios, G. (2003). Measuring emotional intelligence with the MSCEIT V2.0. Emotion, 3, 97–105.Find this resource:
McClelland, D. C. (1951). Personality. New York, NY: William Sloane.Find this resource:
Mentkowski, M., et al. (2000). Learning that lasts: Integrating learning, development, and performance in college and beyond. San Francisco, CA: Jossey-Bass.Find this resource:
Miao, C., Humphrey, R., & Qian, S. (2016). Leader emotional intelligence and subordinate job satisfaction: A meta-analysis of main, mediator, and moderator effects. Personality and Individual Differences, 102, 13–24.Find this resource:
Mikolajczak, M., Luminet, O., Leroy, C., & Roy, E. (2007). Psychometric properties of the Trait Emotional Intelligence Questionnaire: Factor structure, reliability, construct, and incremental validity in a French-speaking population. Journal of Personality Assessment, 88(3), 338–353.Find this resource:
Mikolajczak, M., Luminet, O., & Menil, C. (2006). Predicting resistance to stress: Incremental validity of trait emotional intelligence over alexithymia and optimism. Psicothema, 18(Suppl.), 79–88.Find this resource:
Miller, S. P. (2014). Developing next generation leadership talent in family businesses: The family effect (Unpublished doctoral dissertation). Cleveland, OH: Case Western Reserve University.Find this resource:
Murphy, K. R. (2006). Four conclusions about emotional intelligence. In K. R. Murphy (Ed.), A critique of emotional intelligence: What are the problems and how can they be fixed? (pp. 345–354). Mahwah, NJ: Lawrence Erlbaum.Find this resource:
Nathanson, L., Rivers, S. E., Flynn, L. M., & Brackett, M. A. (2016). Creating emotionally intelligent schools with RULER. Emotion Review, 8(4), 1–6.Find this resource:
O’Boyle, E. H., Jr., Humphrey, R. H., Pollack, J. M., Hawver, T. H., & Story, P. A. (2011). The relation between emotional intelligence and job performance: A meta-analysis. Journal of Organizational Behavior, 32, 788–818.Find this resource:
Pardasani, R. (2016). Resonant leadership and its individual and organizational outcomes: A test of mediation by emotional attractors and resilience (Unpublished doctoral dissertation). Gurgaon, India: Management Development Institute.Find this resource:
Petrides, K. V., Frederickson, N., & Furnham, A. (2004). The role of trait emotional intelligence in academic performance and deviant behavior at school. Personality and Individual Differences, 36, 277–293.Find this resource:
Petrides, K. V., & Furnham, A. (2000). On the dimensional structure of emotional intelligence. Personality and Individual Differences, 29, 313–320.Find this resource:
Petrides, K. V., & Furnham, A. (2001). Trait emotional intelligence: Psychometric investigation with reference to established trait taxonomies. European Journal of Personality, 17, 425–448.Find this resource:
Petrides, K. V., Pita, R., & Kokkinaki, F. (2007). The location of trait emotional intelligence in personality factor space. British Journal of Psychology, 98, 273–289.Find this resource:
Podsakoff, P. M., MacKenzie, S. B., Lee, J.‑Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879–903.Find this resource:
Podsakoff, P. M., & Organ, D. W. (1986). Self-reports in organizational research: Problems and prospects. Journal of Management, 12, 531–544.Find this resource:
Roberts, B. W., Luo, J., Briley, D. A., Chow, P. I., Su, R., & Hill, P. L. (2017, January 5). A systematic review of personality trait change through intervention. Psychological Bulletin. Advance online publication.Find this resource:
Roberts, R. D., Schulze, R., O’Brien, K., MacCann, C., Reid, J., & Maul, A. (2006). Exploring the validity of the Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT) with established emotions measures. Emotion, 6, 663–669.Find this resource:
Rogers, C. (1961). On becoming a person: A therapist’s view of psychotherapy. London, U.K.: Constable.Find this resource:
Rossen, E., & Kranzler, J. H. (2009). Incremental validity of the Mayer-Salovey-Caruso Emotional Intelligence Test Version 2.0 (MSCEIT) after controlling for personality and intelligence. Journal of Research in Personality, 43(1), 60–65.Find this resource:
Rossen, E., Kranzler, J. H., & Algina, J. (2008). Confirmatory factor analysis of the Mayer-Salovey-Caruso emotional intelligence test V 2.0 (MSCEIT). Personality and Individual Differences, 44, 1258–1269.Find this resource:
Ryan, G., Emmerling, R. J., & Spencer, L.M. (2009). Distinguishing high-performing European executives: The role of emotional, social, and cognitive competencies. Journal of Management Development, 28, 859–875.Find this resource:
Saklofske, D. H., Austin, E. J., & Minski, P. S. (2003). Factor structure and validity of a trait emotional intelligence measure. Personality and Individual Differences, 34, 707–721.Find this resource:
Salovey, P., & Mayer, J. D. (1990). Emotional intelligence. Imagination, Cognition and Personality, 9, 185–211.Find this resource:
Schaufeli, W. B., & Bakker, A. B. (2004). Job demands, job resources, and their relationship with burnout and engagement: A multi‐sample study. Journal of Organizational Behavior, 25, 293–315.Find this resource:
Schutte, N. S., Malouff, J. M., Hall, L. E., Haggerty, D., Cooper, J. T., Golden, C. J., & Dornheim, L. (1998). Development and validation of a measure of emotional intelligence. Personality and Individual Differences, 25, 167–177.Find this resource:
Sharma, R. (2008). Emotional intelligence from the 17th century to the 21st century: Perspectives and directions for future research. Vision: The Journal of Business Perspective, 12(1), 59–66.Find this resource:
Spector, P. E., & Johnson, H. M. (2006). Improving the definition, measurement, and application of emotional intelligence. In K. R. Murphy (Ed.), A critique of emotional intelligence: What are the problems and how can they be fixed? (pp. 325–344). Mahwah, NJ: Lawrence Erlbaum.Find this resource:
Spencer, L. M., Jr., & Spencer, S. M. (1993). Competence at work: Models for superior performance. New York, NY: Wiley.Find this resource:
Sternberg, R. J. (1996). Successful intelligence: How practical and creative intelligence determine success in life. New York, NY: Simon and Shuster.Find this resource:
Taylor, S. N. (2014). Student self-assessment and multisource feedback assessment: Exploring benefits, limitations, and remedies. Journal of Management Education, 38, 359–383.Find this resource:
Thornton, G. C., III, & Byham, W. C. (1982). Assessment centers and managerial performance. New York, NY: Academic Press.Find this resource:
Tsui, A. S., & Ohlott, P. (1988). Multiple assessment of managerial effectiveness: Interrater agreement and consensus in effectiveness models. Personnel Psychology, 41, 779–803.Find this resource:
Van Oosten, E. (2013). The impact of emotional intelligence and executive e coaching on leader effectiveness (Unpublished doctoral dissertation). Cleveland, OH: Case Western Reserve University.Find this resource:
Wilderom, C. P. M., Hur, Y., Wiersma, U. J., Van Den Berg, P. T., & Lee, J. (2015). From manager’s emotional intelligence to objective store performance: Through store cohesiveness and sales-directed employee behavior. Journal of Organizational Behavior, 36, 825–844.Find this resource:
Wolff, S. B. (2008). Emotional and social competency inventory: Technical Manual Up-Dated ESCI Research Titles and Abstracts. Boston, MA: Hay Group.Find this resource:
Wong, C. S., & Law, K. S. (2002). The effect of leader and follower emotional intelligence on performance and attitude: An exploratory study. Leadership Quarterly, 13, 243–274.Find this resource: