Emotional Intelligence: A Review of the Literature

Abstract

This paper reviews the literature and traces the origins of the Emotional Intelligence (EI) construct, its definition, measurement and its relationship to leadership and job performance, and other related variables. A discussion of various EI instruments, their relative strengths and weaknesses, psychometric properties and utility are presented. Several studies of EI and their relationship to leadership and job performance are examined. The author concludes by suggesting the EI construct be abandoned and that charisma and maturity are more accurate descriptions of the putative EI construct.

Introduction

The purpose of this article is to review the literature and examine the construct known as Emotional Intelligence (EI). The concept of EI has been in the literature for quite some time, although pinpointing its precise origin is somewhat difficult (Mayer, J. D., Salovey, P. & Caruso, D. R., 2004). Locke (2005) suggests that EI was first introduced by Salovey and Mayer in 1990, although others suggest that similarly labeled constructs precede these researchers (Mayer, Salovey, & Caruso 2004, p. 198). Salovey and Mayer (1990) state, “We define emotional intelligence as the subset of social intelligence…” (p. 189). Interestingly, Cronbach (1960, cited in Salovey and Mayer) concluded that, “fifty years of intermittent investigation …social intelligence remains undefined and unmeasured” (p.188). Most would agree that EI entered the popular mainstream culture as a result of Daniel Goleman’s book entitled Emotional Intelligence, first published in 1995 followed by his article What Makes A Leader? first published in 1998. If one assumes that Salovey and Mayer’s 1990 article is indeed the seminal work that started Goleman’s dialogue of EI as the sine qua non of leadership, it quickly becomes apparent that for about the next twenty-five plus years, scholars struggled with both the definition of the EI construct and its measurement. This article will examine the definition of EI, the measurement of EI, the effect of EI on various dependent variables, and finally a discussion of the literature and conclusions.

Review of the Literature

Salovey and Mayer (1990) state, “We define emotional intelligence 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. 189). Their original conceptual model is depicted in Figure 1. From this model, they posit that there are, “conceptually related mental processes involving emotional information. The mental processes include: a) appraising and expressing emotions in the self and others, b) regulating emotion in the self and others, and c) using emotions in adaptive ways” (p. 190). It is especially important to take note of the phrase, using emotions in adaptive ways, since this becomes the platform upon which many, including Goleman, have constructed an argument that EI is a fundamental and necessary ingredient of leadership.

Salovey and Mayer (1990) go on to suggest that appraising emotions in others requires an ability to observe and translate both verbal and nonverbal expressions of emotions. This skill, it is suggested, then becomes part of one’s EI. Concurrently, the ability to express one’s emotions and be self-aware of those expressions, both verbal and non-verbal, becomes a second skill, and again part of one’s EI.

The second component of EI, regulating emotion (in self) and the skills for regulating it, they suggest, is drawn from research on moods. They state, “Moods, although less intense and generally longer lasting than emotions, should be just as effectively regulated and managed by individuals with emotionally intelligent skills” (p. 196). Salovey and Mayer (1990) describe the meta-experiences of mood as essentially that inventory of all the various types of moods, feelings and emotions one has typically experienced: happy, sad, angry, confused, anxious, etc. This inventory becomes the framework for which each individual builds a theory about how his or her own moods and concomitant behaviors work. As an example, they offer, “…if one experiences a pleasant, acceptable mood when dancing, then the cause of the mood (dancing) could be sought after in the future so as to bring about the mood again. In this way, it would serve as a foundation upon which rules could be constructed that would themselves direct behavior to bring about moods” (p. 196).

Regulating emotions in others, or more simply put, managing impressions, is also a learned skill. For example, a job candidate arriving on time. Goffman (1959) wrote extensively on impression management:

When an individual enters the presence of others, they commonly seek to acquire information about him or to bring into play information about him already possessed. They will be interested in his general socio-economic status, his conception of self, his attitude toward them, his competence, his trustworthiness, etc. Although some of this information seems to be sought almost as an end in itself, there are usually quite practical reasons for acquiring it. Information about the individual helps to define the situation, enabling others to know in advance what he will expect of them and what they may expect of him. Informed in these ways, the others will know how best to act in order to call forth a desired response from him (pp. 135,136).

Goffman seems to be suggesting that a job candidate might just be arriving on time not as an end in itself, but as the means to an end, e.g., getting the job. Moreover, the hiring manager, observing the punctuality of the job candidate is using that information to infer future behaviors relating not only to punctuality, but to job performance at a minimum, if not any other number of inferred positive traits.

The third component, using emotions in adaptive ways, is the suggestion that certain moods give rise to certain types of productivity. For example, if one is happy they might be more creative, whereas if they are angry they may enjoy physical labor as a way to work off their anger. Salovey and Mayer (1990) conclude, “When people approach life tasks with emotional intelligence, they should be at an advantage for solving problems adaptively” (p. 200). This concluding remark by Salovey and Mayer lays down the foundation of the suggestion that if one has EI, he/she will perform at a higher level and launched the next twenty-five plus years’ struggle of defining and measuring EI. Indeed, these same authors revisited their original 1990 model, expanding their work (Mayer and Salovey, 1995) and further revising it to the now so-called Four Branch Model wherein (1) perceiving emotion is at the bottom of the model, (2) followed by facilitating thought with emotion; (3) understanding emotion; and (4) finally managing emotion (Salovey, et. al. 1995, p.235).

One of the first teams to take on the challenge of measuring at least a dimension of EI, was Salovey, Mayer, Goldman, et. al., (1995). This team developed an instrument known as the Trait Meta-Mood Scale (TMMS). They write, “The measure is called the Trait Meta-Mood Scale (TMMS) because it was designed to assess relatively stable individual differences in people’s tendency to attend to their moods and emotions, discriminate clearly among them, and regulate them” (pp. 127, 128). Their 48- item self-report, scale (later refined to 30 items) and subsequent factor analysis revealed three factors labeled as (a) attention to feelings, (b) clarity of feelings, and (c) mood repair. The authors argue for the instrument’s reliability and “sufficiently differentiated from related constructs such as neuroticism and repression. Moreover, the clarity scale in particular demonstrated validity in predicting the unpleasant quality of ruminations after a stressful experience” (p. 147). The authors conclude, “we believe that the TMMS is a reasonable operationalization of aspects of emotional intelligence” (p. 147). However, they further stipulate that “future research needs to be focused on the discriminant validity of the TMMS vis-a-vis the constructs noted…” (p. 147).

The team of Schutte, Malouff, Hall et.al. (1998) attempted to construct a measure of EI, also based on Salovey and Mayer’s original model of EI. This instrument was developed from an initial 62-item self-report scale, subsequently reduced to 33 items using factor analysis, resulting in a one factor solution. This solution, the authors suggested, is “a homogenous construct of emotional intelligence” (p. 175). They cautioned, however, that while the instrument held “promise as a reliable valid measure of EI….the emotional intelligence scale, like most self-report measures, seems susceptible to faking good. Thus the emotional intelligence scale should probably not be used as a method for selecting individuals for jobs or other highly desired opportunities” (176). The authors further presented that the 33-item scale did correlate with theoretically related constructs, including alexithymia, attention to feelings, clarity of feelings, mood repair, optimism, and impulse control (p. 167). The authors also posited that the instrument had predictive validity since the EI scores of 31 college freshmen predicted final grade point averages with an r(63) = .032, and p value of <0.01 (pp. 173,174). They also argued discriminant validity of the instrument citing their own study of 42 college freshmen and corresponding SAT scores. They found no relationship between EI scores and SAT scores, r(41) = -.06. In a second study of 23 college students, they again suggested discriminant validity of their EI instrument based on an analysis of a NEO personality Inventory (Costa and McCrae, cited in Schutte, Malouff, Hall et.al 1998), more commonly referred to as the big five personality dimensions. In this study they found that higher EI scores were significantly associated with greater openness to experience, r(22)=0.54, p value ,0.009. The remaining personality dimensions were reported as: neuroticism, – 0.28; extraversion, 0.28; agreeableness, 0.26; conscientiousness, 0.21. (p.175).

The reader should note at this point that the two previous discussions of EI instruments (TMMS and Schutte, Malouff, Hall et.al 33-item scale) were both of the self-report type, with some weaknesses noted in both. Self-report measures of EI have received critical scrutiny by several researchers. Ciarrochi, Chan and Caputi (2000) state the problem as follows:

In a recent set of studies, Davies et al. (1998) uncovered a number of problems with the different self-report EI measures. First, Goleman’s (1995) EQ test and two self-report measures of the Alexithymia Scale have poor reliabilities (cf. Bagby et al., 1994). Second, the more reliable self-report measures have salient loadings on the well-established personality factors of Neuroticism, Extraversions, Psychoticism, Agreeableness and Openness. Third, although there is a factor-analytic evidence supporting discriminant validity of two emotional intelligence factors (emotional awareness and clarity), these factors no longer emerge when unreliable measures are dropped from factor analysis (p. 541).

Ciarrochi, Chan and Caputi (2000) addressed this in their study by using the Multifactor Intelligence Emotional Intelligence Scale (MEIS) developed by Mayer, Caruso and Salovey (1999). The authors describe this instrument, “as an objective measure of EI….the test is objective in the sense that there are better and worse answers, as determined by consensus” (p. 541). In their study, they assessed the reliability and validity of the MEIS by assessing the factor structure and reliability of 11 of the 12 subscales. Their findings confirmed that, “The EI factors were reliable, all the subscales loaded on a single emotional ‘g’ factor, EI related in expected ways to a number of criterion variables and to mood-based processes even after controlling for IQ and well-established personality variables. Taken together, these results make a case for the distinctiveness and usefulness of EI” (p. 556). It is worth noting that their study also confirmed previous research (Mayer & Geher, 1996) which suggests women score higher on EI than men. The issue of gender differences will be addressed further along in this paper.

The team of Law, Wong and Song (2004) also addressed the challenges of the EI construct and its measurement. In their study, they attempted to validate yet another EI instrument, the Wong and Law EI Scale (WLEIS) (Wong and Law, 2002, cited in Law, Wong and Song, 2004). Using confirmatory factor analyses (CFA) and multitrait-multimethod (MTMM) analyses, they found that EI (as measured by the WLEIS) is “related to but distinct from the big five personality dimensions” (p. 494). Much of the controversy around the EI construct is the question of whether EI is an ability or a personality trait. These authors suggest that if EI is in fact a skill, then self-report measures should not be used, despite the fact their study used the WLEIS, a self-report instrument.

The 2001 study of Morand attempts to measure at least one dimension of EI, specifically the ability to, “recognize emotional expressions displayed by others” (p. 21). In this study, respondents are presented with photographs of faces expressing any of the six so-called universal emotions: anger, happiness, surprise, fear, disgust, and sadness (p. 27). Morand hypothesized that women would outperform men. The study results indicated that women did in fact outperform men on the facial recognition task, specifically the author found a statistically significant difference at the p value <0.05 level with women scoring an average of 14.17 versus men, 12.48 (p. 29). Morand posited that, “proficiencies of a human relations leadership style may be bifurcated. That is, a critical distinction may exist between expressing emotion versus perceiving it in others…There exists a viable argument that these dimensions are quite independent of one another, indeed it is possible that being adept in one area may be inversely related to skill in the other” (p. 30). In other words, one might be a poker player who may be good at reading the other players at the table and know who is bluffing, yet himself has a behavior that unintentionally informs others at the table, i.e., a tell. The author concludes, “the results indicate one facet of emotional intelligence, here operationalized as the accurate discernment of other’s emotional states, can be reliably measured at a non-verbal level in a laboratory setting” (p. 30).

With the several aforementioned studies in mind, the following studies attempt to validate or compare and contrast many of the EI instruments. One study in particular by Conte (2005) examines several of the better known EI instruments. In this research, Conte reviewed the (1) Emotional Competence Inventory, (2) Emotional Quotient Inventory, (3) Multifactor Emotional Intelligence Scale, and (4) Mayer-Salovey Caruso Emotional Intelligence Test V.2. Conte held that there was no evidence of predictive and discriminant validity for the Emotional Competence Inventory. Similarly, Conte’s analysis of the Emotional Quotient Inventory suggested that while there was, “adequate reliability and some validity evidence, it is lacking in discriminant validity evidence, and few studies have examined whether it provides incremental predictive validity above the contribution of established predictors such as cognitive ability and big five personality dimensions” (p. 435). Lastly, Conte took the Multifactor Emotional Intelligence Scale and the Mayer-Salovey Caruso Emotional Intelligence Test V.2, discussing them in tandem, since the latter was born of the former. In the final analysis, Conte offered little support for either of these instruments (p. 436). Conte concludes his analysis that, “because few EI researchers are willing to be specific about what they want to measure, it is difficult to examine content validity. Similarly, construct validity in the form of convergent and discriminant validity is lacking” (p. 437). He continues, “in sum, serious concerns remain for all of the EI measures, ranging from scoring concerns for ability-based EI measures to discriminant validity concerns for self-report measures” (p. 438).

A second work to review, by the team of Davies, Stankov and Roberts (1998), is noteworthy for the depth and breadth of their investigation into the EI construct. Over the course of three studies, they investigated “the psychometric status of emotional intelligence and determine whether it serves as a unique type of human ability” (p. 989). Among their several findings, the authors concluded that self-report measures of EI are “problematic for one of the following two reasons. First, some of the self-report measures have poor reliability….Second, all self-report measures having satisfactory reliability tend to load on well-known personality factors” (p. 1012). The authors further suggest that, “determining what is the correct response to an emotional intelligence test item presents practical and theoretical difficulties” (p. 1012). They conclude their findings by stating, “the three studies reported here converge on a conclusion that, as presently postulated, little remains of emotional intelligence that is unique and psychometrically sound” (p. 1013).

Finally, a third article, by the well-known researcher Edwin Locke, offers his review of the EI construct in his paper, Why Emotional Intelligence is an invalid concept (2005). The essence of his objections is rather straightforward. He suggests that emotions are an “automatic process based on subconsciously held knowledge and values” whereas “reason is the means of gaining and validating one’s knowledge.” The point of this distinction he writes is that, “One cannot therefore reason with emotion, one can only reason about it” (p. 427). Locke saves his strongest words for the linkage of EI to leadership, especially by Goleman (discussed earlier in this article). Locke writes:

In addition to making the concept of EI-leadership preposterously all-encompassing, Goleman et al. seriously misconstrue what organizational leadership involves. They claim that ‘The fundamental task of leaders is to create good feelings in those they lead’. This is simply not true. The function of organizations is to attain goals; in the case of private organizations the goal is long term profitability. Organizations, other than psychotherapy clinics, are not in the ‘feel-good’ business. Employee morale is important, but as a means to an end not as an end in itself divorced from effectiveness (p. 428).

The team of Brackett and Salovey (2006) attempted to rescue the EI construct with their validation study of the Mayer-Salovery-Caruso Emotional Intelligence Test (MSCEIT) in 2006. They concluded that “EI, defined and operationalized as a mental ability with the MSCEIT, can be reliably measured and predict important psychological and behavioral outcomes (e.g., depression and peer reports of relationships satisfaction)” (p. 40). These same researchers, however, pointed out that “mixed and perhaps unexpected findings with MSCEIT in organizational settings. Managing emotions scores on the MSCEIT were significantly lower for the highest ranked and highest paid among 59 senior executives in a large international production and service organization” (Collins, 2001 cited in Brackett & Salovey, 2006). These findings seem to question the proponents of EI when one ponders their preference for a higher EI score or a higher salary.

In 2008 the seasoned research team of Mayer, Salovey and Caruso once again tackled the question of EI in a paper entitled, Emotional Intelligence New Ability or Eclectic Traits. They begin their paper affirmatively stating, “Our principal claim is that a valid EI concept can be distinguished from other approaches” (p. 503). By the authors’ own admission and their five concluding recommendations, one may wonder what remains of the EI construct. Their 5 recommendations include:

…urge researchers and practitioners alike to refer to the scientific literature on emotions, intelligence, and emotional intelligence;
…emotional intelligence be limited to abilities at the intersection of emotions and intelligence;
…an EI refocus on research relevant to the ability conception of EI;
….widely studied personality traits…should be called what they are, rather than being mixed together in haphazard-seeming assortments and named emotional intelligence; and
…applications of EI must be conducted with much greater attention to the research literature, be grounded in good theory, and reject outlandish claims. (513,514)

Like so many other EI researchers, Mayer, Salovey and Caruso seem to struggle with the belief that EI exists, yet their belief appears conflicted by their own data. Their article seems to be predicated on flawed models, suspect measures and tenuous relationships between any numbers of variables. At this point, any reasonable reader or researcher would ask “why does the concept persist?” Yet, persist it does, and there have in fact been quite a few studies using a variety of these EI measures, replete with all the aforementioned weaknesses and limitations, all in an attempt to measure the impact of EI on any number of variables, such as performance. Below is a review of some of these studies.

Sy, Tram and O’Hara (2006) examined the relationship of EI to job satisfaction and performance. In this study of food service employees the authors found that employees with a higher EI score had a higher job satisfaction, with a reported R square of .06, p <.0001. EI was also positively correlated with job performance with a reported R square of .03, p <.001. These finding were however limited, as the authors point out, by the fact that the EI measure used was a self-report instrument and the performance measure was a subjective one, specifically performance was measured by managers. An objective measure of performance such as work output or revenue would strengthen these findings.

Marquez, Martin and Brackett (2006) examined the relationship of EI to social competence and academic achievement. Their study examined 77 high school students, using the MSCEIT instrument to measure student’s EI, the big five questionnaire (BFQ), the Factorial General Intelligence test, the Social-Cognitive Attitudes and Strategies test and finally, student’s final grades from science and humanities classes as a measure of academic achievement. This study is noteworthy for its ambition, and interesting to note that the BFQ measure was used as a control variable in their final analyses. This is commendable given the many concerns discussed earlier in this paper regarding the big five personality traits and correlation with various EI instruments. Nevertheless, the authors make their case for the MSCEIT’s discriminant validity since in their study they did not find significant correlation with neuroticism, extraversion and conscientiousness. Furthermore, the authors found that EI was significantly related to nine of the social competencies measures and final grades (i.e., academic achievement). They also found associations between EI and self-confidence, prosocial behavior and academic grades while controlling for general intelligence.

Bastian, Burns and Nettlebeck (2005) examined EI, life skills, personality and cognitive skills. In their study of 246 first-year psychology students, they administered the Trait Meta Mood Scale (TMMS), the Assessing Emotions Scale (AES) and the MSCEIT. Additionally students completed Raven’s Advanced Progressive Matrices, an abstract reasoning measure, the Phonetic Word Association Test, Revised NEO Personality Inventory, the Satisfaction with Life Scale, the Anxious Thoughts Inventory, The Problem Solving Inventory, the COPE and finally the Tertiary Entrance Rank. Once again a very ambitious methodology is deployed, including a three-pronged approach to quite literally triangulate the EI construct. The ambitious attempt at measuring EI may be conflated, however, by the three different methodologies when they get to their statistical analysis. Specifically, the approach resulted in dimensionality of the TMMS and MSCEIT, requiring the study’s authors to cobble together a model using elements of the knowledge and management subscales from the MSCEIT and the TMMS subscales. The authors concluded that EI, measured as such, predicted life skills; however, it did not predict personality or cognition. The reader should take special note of this study as an example of the difficulties in executing a study of EI, no matter how well intentioned the researchers.

The study by Brackett, Mayer and Warner (2004) examined EI and everday behavior. Everyday behavior was measured using the Life Space scales which “assessed an array of self-care behaviors, leisure pursuits, academic activities and interpersonal relations” (p. 1387). EI was measured using MSCEIT. Subjects (n=330 college students) were also measured for their Big Five personality traits using the International Personality Item Pool. Among the study’s findings were that “In contrast to our expectations, EI was not related to many positive aspects of the Life Space…most of the correlations suggested an interaction between gender and EI as predictors of life space criteria” (p.1394). The authors concluded that “more research is needed to understand how EI is expressed in people’s lives” (1400).

Kerr, Gavin, Heaton and Boyle (2006) examined EI and leadership effectiveness. This study used the MSCEIT to measure the EI of supervisors (n = 38) and a second measure, given to the supervisor’s subordinates (n = 1258), to measure leadership effectiveness. The MSCEIT produces a total EI score as well as factor-level area scores for experiential EI and reasoning EI as well as factor-level branch scores for perceiving emotions, using emotions, understanding emotions and managing emotions. The authors noted that, “surprisingly both the understanding and managing emotions branch scores and their corresponding EI domain, did not display a significant relationship with supervisor ratings” (p. 272). They conclude that, “the ability to understand emotions and the ability to act effectually on this understanding may only be marginally related” (p. 275).

Barbuto, Gotfredson and Searle (2014) examined EI as an antecedent of servant leadership. In this study of civic leaders (n = 75) and followers (n= 401), a 30-item, self-report EI instrument was used which was developed by Carson et. Al. (2000 cited in Barbuto, Gotfredson and Searle 2014). Servant leadership was measured using the Servant Leadership Questionnaire. The authors’ findings suggested that EI was, “a good predictor of a servant leader’s servant-leader ideology but may not be a good predictor of servant-leader behaviors as rated by the leader’s followers” (p. 315).

Wong and Law (2002) examined the EI of leaders and the job performance of their subordinates. In this series of three separate and progressive studies, the authors first developed their own EI instrument. Wong and Law argued for their EI instrument’s reliability and validity based on their confirmatory factor analysis. They subsequently used their EI instrument in the two other successive studies, each looking at supervisor-subordinate relationships and its effect on job performance. They concluded that “no relationship between the EI of leaders and the job performance of their followers has been found” (p. 269).

The final two EI articles are framed in tandem, for they both use a meta-analysis approach and in fact build upon one another. The first by Joseph and Newman (2010) suggested that EI should follow a cascading model, wherein there is a set of sequential behaviors which ultimately impact job performance. The authors also excluded from their model, emotion facilitation because they feel, “…empirical research has shown a lack of empirical validity for the emotion facilitation facet” (p. 55). Their model suggests that emotion perception takes place first, followed by emotion understanding, followed by emotion regulation, and concluding with job performance. Because of their meta-analytic approach, the authors were able to examine the three different approaches to measuring EI, including (1) self-report mixed EI, (2) self-report ability EI and (3) performance-based EI. In their study they reported that, “although we report correlations of overall EI with personality, cognitive ability and job performance in Table 2 (omitted here), we warn that the concept of overall EI (averaged across the three construct-method pairings) is of limited conceptual value due to inconsistent and low correlations among some types of EI measures” (p. 65). In terms of their cascading model, they conclude that “much of the relationship between emotion perception and emotion regulation is mediated by emotion understanding” (p. 65). They continue their summary discussion, “Our further decision to omit self-reported mixed EI from the theoretical model is based on the fact that mixed EI is a muddled construct – an ill-defined composite of ability, personality, affect and possibly other poorly specified content” (Murphy, 2006 cited in Joseph and Newman). The authors concluded, “at worst, measures of ability models of EI show only a modicum of incremental validity over cognitive ability and personality traits, again providing evidence against Goleman’s (1998) expansive claims” (p. 69).

As a follow up, or perhaps as a response to the aforementioned meta-analysis, the team of O’Boyle, Humphrey, Pollack, Hawyer and Story (2011) conducted another impressive meta-analysis of the EI construct. While acknowledging the 2010 work of Joseph and Newman, this team set out to improve upon their work through several efforts including more than double the number of studies included in the current meta-analysis, testing for publication bias and using a recently developed statistical procedure, dominance analysis, which attempts to improve on other methods by providing more precision than beta weights in explaining the relative amount of total variance explained by all of a model’s factors.

Similarly to Joseph and Newman, the current team of O’Boyle, et. al. distinguished the various approaches to measuring EI describing them as three streams. Once again, the distinctions in how EI is measured are no small thing. “Ashkanasy and Daus (2005) argued that there are important theoretical and methodological reasons to distinguish among the three streams” (cited in O’Boyle, et. al., p. 792). The three streams are described as follows, “(1) A four-branch abilities test based on the model of EI defined in Mayer and Salovey (1997); (2) self-report instruments based on the Mayer-Salovey model; and (3) commercially available tests that go beyond the Mayer-Salovey definition” (cited in O’Boyle, et. al., p. 792). Among the reasons for concern are again, overlap across all three approaches with personality traits, i.e. the big five or Five Factor Model (FFM).

O’Boyle, et. al. tackled these concerns head on by positing that, “Although there is considerable debate among the advocates of the three different streams of EI research, there is good reason to believe that all three types of measures predict job performance” (p. 793). From this premise that all three streams will predict job performance, using the meta-analytic approach and dominance analysis the authors concluded that, “Overall, the present data did not offer support for claims that EI is the single most important predictor on job performance” (p. 804). They did, however, further conclude that all three streams of EI provided “additional explanatory power above and beyond the FFM and cognitive ability” (p. 803). Specifically, they found the stream 1 approaches explained 6.4 percent of the total explained variance with an R² = 0.03. Additionally, stream 2 approaches explained 13.6 percent of the total explained variance with an R²= 0.065. Finally, stream 3 approaches explained 13.2 percent of the total explained variance also with an R²= 0.065. These findings are discussed further below.

Discussion

The foregoing has traced an approximate 25-plus-year journey in the literature of the many varied and multifaceted attempts at defining, measuring and ultimately evaluating the utility of the EI construct. The results of these studies appear mixed. The just discussed meta-analyses above are illustrative of what seem to be many of the problems with the EI narrative. Both of these meta-analyses are built upon a seemingly shaky platform drawn from the literature, which is filled with qualifying statements, limitations and the usual requisite calls for further research (Cote, S. & Miners, C., 2006; Brackett, M. A. & Mayer, J. D., 2003; Murphy, K., 2009). Yet the researcher teams in both of these examples forge ahead, plugging into their model data gleaned from a virtual hodge-podge of populations and methodologies with no regard for the underlying weaknesses in the original studies.

Moreover, efforts such as the O’Boyle, et. al. teams, while heroic in their scale would seem to fall short of an aha moment. The suggestion that all Three Streams, regardless of the approach to measuring EI, all predict performance is rather misleading. The reader should look at the O’Boyle et. al. team’s results. An R² = 0.03 is typically not considered a profound linkage, especially in the context of the much larger model of explaining performance. While one must first appreciate O’Boyle et. al.’s findings of a total R²for the Streams 1, 2, and 3 are .427; .474; .491, respectively, it also means that over fifty percent of the variance for performance remains unexplained in the current models. It is very possible, and very likely that further unmasking of other determinants of performance, for example, effort, clarity of goals, feedback, etc., would diminish even further any putative effects of EI. This further begs the question, at that level of significance, given all the other things a manager or leader would wish to work on, is EI really worth the effort?

The careful reader should recognize that there has been a constant changing and shifting of the EI definition and its measurement, and in many ways, the various schools of thought undermine each other, yet simultaneously reveal a universal truth: EI has not been defined, has not been measured, and has not shown a significant and meaningful casual effect on performance. It seems that even when a particular aspect of a given EI measure survives rigorous psychometric evaluation, its relationship to a meaningful variable, say for example, job performance, yields very little. On the other hand, the EI approaches that do “show promise,” the many so-called mixed models of EI, while often times showing some power at predicting job performance, quickly fall apart when scrutinized for psychometric validity.

If there is a consistent thread throughout the literature, it is a bias in the direction that there must be something there and that when that something is not detected, it’s always a case of needing more research or better measures or different populations. Rarely is it held out that things like performance are a function of effort, cognitive ability and personality, all of which are indeed quite measurable and known to predict performance.

There seems to be a few major points in the literature which do not appear to be in dispute. The very narrow skill of perceiving what a given emotion being expressed is generally better performed by women than men as pointed out above. There is strong evidence that emotion perception is rooted in one’s neurobiology, meaning it is not a learned skill or behavior (Phillips, Drevets, Rauch & Lane, 2003). Similar to one who is colorblind, a person may or may not have a particular ability to see or not see certain things, including various emotions being expressed. It is a biological characteristic and one’s natural ability varies. There is also a hierarchy in one’s ability to perceive various emotions in others. Humans are hardwired to more accurately perceive anger in someone, than say sadness or happiness, the result of our Darwinian survival skills. Someone charging at us, possibly wanting to harm us, is much more important information than someone with a smile on their face who wants to tell us a joke.

Finally, it is important to note the following unanswered questions (Mayer, et. al., 2001). Does EI and its purported casual effects transcend race? Does EI transcend cultural settings and cultural background? Does EI transcend organizational settings? Why do EI scores decrease as performance scores increase? What role do neurobiological factors have in emotion perception across age, gender, race and family history? These and many other unanswered questions demand that researchers, leadership and management educators, and practicing managers bridle their own enthusiasm over the seductive power and unmet promises made by the likes of Daniel Goleman.

Conclusions

There is a long and very illustrious line of academics with very well thought out arguments in favor of the EI construct. There are also quite a few notable academics who strongly suggest that there are several issues and concerns with EI, especially as it has been operationalized in the applied world of management and leadership. Nevertheless, the term Emotional Intelligence should not be used. Words have meaning. It is very clever, yet also very misleading, to couple a word like intelligence alongside the word emotion as a type of bait. People should not be induced into thinking EI is something they should want and that if indeed they don’t have it, they are suffering some type of defect. The proponents of EI seem to have a psyche about them which blocks out any contradicting evidence. They should be cautioned. Imagine an employer who unwittingly screens out men or minorities because they are using an EI instrument as part of their selection process. If the goal is to teach good manners, teach good manners because it’s the right thing to do, not because it will increase your performance or the performance of your subordinates. Mayer and Cobb (2000) reminded us,

Educational policy related to emotional intelligence is of considerable interest. The policies are well meaning and often executed through promising curricula devoted to socio-emotional learning. An examination of the emotional intelligence concepts in educational policy indicate some weaknesses in how that policy was formulated, however, and some serious lapses in how it is tied to science. Most centrally, the polices are based on popularizations of a very young science that is, at present, still developing support for its central hypothesis that emotional intelligence exists (p. 180,181).

All roads lead to Rome and many roads can lead to one performing at a high level and being an effective leader. Everyone is a unique individual capable of extraordinary things, and who we are and how we accomplish those things will always naturally vary. It is myopic to suggest everyone should fit into a very narrow model as to how to behave and respond to the behavior of others. Moreover, there is absolutely no proof that EI is a necessary pre-requisite to being an effective leader; the suggestion that it is, is patently false. Leading people is not about making them feel good. Leading people is about getting them to perform.

Think about emotions. The terms, ‘tears of joy’ or ‘nervous laughter’ are both oxymorons, yet we immediately see that underneath these expressions is the real possibility that one’s reaction to an event will be very different than another person’s. Conversely, is one’s ability to recognize that tears truly mean sadness anything more than an innate physical ability that is limited in its capacity for improvement? And indeed, do tears always mean sadness? Are tears the singular correct response to a given situation? And is empathy the truly correct response to those tears? Can someone be taught to be a better liar? Can someone learn how to tell who is lying and who is telling the truth? If the goal is to improve performance of one’s self or in one’s subordinates, they should be taught the skills needed for that job. If the goal is to improve performance, the focus should be on effort. How one feels about doing something over the long run may make them feel better or worse about doing the task, but skill is skill and effort is effort. Telling someone it’s ok if they feel bad about being late for work will not solve the problem of tardiness.

If the challenge is to describe the indescribable phenomena of what makes some great leaders great, there is already a word for that: charisma. If the challenge is to describe what it means to respond appropriately to someone else’s emotions, there is a word for that: maturity. However, no amount of maturity and no amount of charisma will replace skill, effort and intelligence in the ability to perform, lead and inspire others to perform.

References

Barbuto, J., Gottfredson, R.& Searle, P. (2014). An Examination of Emotional Intelligence as an Antecedent of Servant Leadership. Journal of Leadership & Organizational Studies, Vol 21 (3), 315-323.
Bastian, V. A., Burns, N. R.& Nettelbeck, T. (2005). Emotional Intelligence predicts life skills, but not as well as personal and cognitive abilities. Personality and Individual Differences, Vol. 39, 1135-1145.
Brackett, M. A.& Mayer, J. D. (2003). Convergent, Discriminant and Incremental Validity of Competing Measures of Emotional Intelligence. Personality and Social Psychology Bulletin Vol. 29, (9), 1147-1158.
Brackett, M. A., Mayer, J. D.& Warner, R. M. (2004). Emotional Intelligence and its relation to everyday behavior. Personality and Individual Differences, Vol. 36, 1387-1402.
Brackett, M. A.,& Salovey, P. (2006). Measuring emotional intelligence with the Mayer-Salovery-Caruso emotional intelligence test (MSCEIT). Psicothema, Vol. 18, suppl., 34-41.
Ciarrochi, J., Chan, A.,& Caputi, P. (2000). A critical evaluation of emotional intelligence construct. Personality and Individual Differences, 28, 539-561.
Conte, J. M. (2005). A review and critique of emotional intelligence measures. Journal of Organizational Behavior, Vol. 26, No. 4, 433-440.
Cote, S.& Miners, C. T. H. (2006). Emotional intelligence, cognitive intelligence, and job performance. Administrative Science Quarterly, Vol. 51, No. 1 1-28.
Davies, M., Stankov, L. & Roberts, R. (1998). Emotional intelligence: In Search of an Elusive Construct. Journal of Personality and Social Psychology, 75 (4), 989-1015.
Friedman, J. (1997). On Bias, Variance, 0/1 – Loss, and the Curse of Dimensionality. Data Mining and Knowledge Discovery, 1, 55-77.
Goffman, E. (1959). The presentation of self in everyday life. Garden City, N.Y.: Doubleday.
Goleman, D. (1998). What makes a leader? Harvard Business Review, Nov-Dec, 1998.
Goleman, D. (1998) Emotional intelligence. Bantam.
Joseph, D. & Newman, D. (2010). Emotional Intelligence: An integrative Meta-Analysis and Cascading Model. Journal of Applied Psychology, 95 (1), 54-78.
Kerr, R., Garvin, J., Heatoon, N & Boyle E. (2006). Emotional Intelligence and leadership effectiveness. Leadership & Organization Development Journal, 27 (4), 265-279.
Law, K., Wong, C., & Song, L. (2004). The Construct and Criterion Validity of Emotional intelligence and Its Potential Utility for Management Studies. Journal of Applied Psychology, 89 (3), 483-496.
Locke, E. A. (2005). Why emotional intelligence is an invalid concept. Journal of Organizational Behavior, Vol. 26, No. 4, 425-431.
Marquez, P. G., Martin, R. P. & Brackett, M. A. (2006). Relating Emotional Intelligence to social competence and academic achievement in high school students. Psicothema, Vol. 18, supl., 118-123.
Mayer, J.D. & Cobb, C. (2000). Educational Policy on Emotional Intelligence: Does it make sense? Educational Psychology Review 12 (2).
Mayer, J. D. & Geher, G. (1996). Emotional intelligence and the identification of emotion. Intelligence, 22, 89-113.
Mayer, J. D. & Salovey, P. (1995). Emotional Intelligence and the construction and regulation of feelings. Applied and Preventive Psychology, 4, 197-208.
Mayer, J., Caruso, D. & Salovey, P. (1999). Emotional intelligence meets traditional standards for an intelligence. Intelligence 27.4 (1999): 267-298.
Mayer, J. D., Salovey, P. & Caruso, D. R. (2004). A Further Consideration of Emotional Intelligence. Psychological Inquiry, Vol. 15, No. 3, 249-255.
Mayer, J. D., Salovey, P. & Caruso, D. R. (2004). Emotional Intelligence: theory, findings and implications. Psychological Inquiry, Vol. 15, No. 3, 197-215.
Mayer, J. D., Salovey, P. & Caruso, D. R. (2008). Emotional Intelligence New Ability or Eclectic Traits? American Psychologist, Vol. 63, No. 6, 503-517.
Mayer, J. D., Salovey, P., Caruso, D. & Sitarenios, G. (2001). Emotional intelligence as standard intelligence. Emotion, 1 (3),
Morand, D. A. (2001). The emotional intelligence of managers: Assessing the construct validity of a nonverbal measure of “People Skills”. Journal of Business and Psychology, Vol. 16, No. 1, 21-33.
Murpy, K. (2009). Emotional Intelligence: A disappointing decade. The American Journal of Psychology, Vol. 122, No. 1, 131-134.
O’Boyle, E., Humphrey, R., Pollack, J., Hawver, T. & Story, P. (2011). The relationship between emotional intelligence and job performance: a meta-analysis. Journal of Organizational Behavior, 32, 788-818.
Parker, J., Taylor, G. & Bagby, R. (2001). The relationship between emotional intelligence and alexithymia. Personality and Individual Differences, 30, 107-115.
Phillips, M., Drevets, W., Rauch, S. & Lane, R. (2003). Neurobiology of Emotion Perception I: The neural Basis of Normal Emotion Perception. Biological Psychiatry, 54, 504-514.
Salovey, P. & Mayer, J. D. (1990). Emotional intelligence. Imagination, Cognition and Personality, Vol. 9, No. 3, 185-211.
Salovey, P., Mayer, J., Goldman, S., Turvey, C. & Palfai,T. (1995). Emotional Attention, Clarity, and Repair: Exploring Emotional Intelligence Using the Trait Meta-Mood Scale. Emotion, Disclosure and Health. American Psychological Association, 125 – 154.
Schutte, N., Malouff, J., Hall, L., Haggerty, D., Cooper, J., Golden, C., & Dornheim, L. (1998). Development and validation of a measure of emotional intelligence. Personality and Individual Differences, 25, 167-177.
Sifneos, P. (1972), Short-Term Psychotherapy and Emotional Crisis, Harvard University Press, Cambridge.
Sy, T., Tram, S. & O’Hara, L. (2006). Relation of Employee and Manager Intelligence to Job Satisfaction and performance. Journal of Vocational Behavior 68, 461-473.
Wong, C. & Law, K. (2002). The effects of leader and follower emotional intelligence on performance and attitude: An exploratory study. The Leadership Quarterly, 13, 243-274.