The contribution of emotional intelligence to the social and academic success of gifted adolescents as measured by the multifactor emotional intelligence scale--adolescent version. (Social and Emotional Development.

The concept of emotional intelligence became popularized by the publication of Daniel Goleman's (1995) best-selling book on this construct and many subsequent magazine and newspaper articles (e.g., Henig, 1996; Peterson, 1997). These publications captured the imagination of the media and the general population by claiming extraordinary predictive validity for emotional intelligence. Goleman argued that general intelligence (IQ) typically predicts only about 20% (r = .45) of the variance that determines various domains of life success, leaving about 80% to other factors. Moreover, he professed his belief that many of these "other" factors are related to a construct that has been termed emotional intelligence. Goleman also detailed his belief that emotional intelligence can sometimes be more powerful than IQ-that it can contribute greatly to several important life outcomes including improved learning, less aggression, better decision making, and many other characteristics that imply successful living. In fact, he claimed that increasing emotional intelligence leads to "advantage in any domain in life" (p. 36).

Although Goleman's (1995) work has brought a great deal of attention to the construct of emotional intelligence, the grandiosity of his claims has been questioned by the originators of the term "emotional intelligence" (Mayer & Salovey, 1997; Mayer, Salovey, & Caruso, 2000). These researchers have addressed several potential problems.

First, Mayer et al. (2000) highlighted how Goleman's (1995) model includes several personality and motivational characteristics (e.g., empathy, persistence) in addition to intelligence. They argued that referring to these multiple variables as one entity--emotional intelligence-is both impractical and misleading. Second, they argued that Goleman's arguments about how much variance emotional intelligence can predict are misleading. Goleman maintained that emotional intelligence can account for more variance than general intelligence (IQ), when studying factors such as academic, occupational, and social success. If this were true, emotional intelligence would predict these outcomes at about r = .45 or better, a finding that Mayer et al. suggested would be extraordinary.

Despite their concerns with Goleman's (1995) concept of emotional intelligence, Mayer et al. (2000) reaffirmed that there is reason to be excited about emotional intelligence. However, they emphasized the need to view emotional intelligence as a mental ability, not as a conglomeration of various cognitive, motivational, and personality characteristics. Additionally, they have offered a more conservative estimate of the predictive ability of emotional intelligence, suggesting that satisfactory new variables should add between 1% and 5% additional variance when predicting outcomes related to success. Findings like this could be significant and meaningful, even though far less extraordinary than Goleman's predictions for emotional intelligence.

Modern Models of Emotional Intelligence Defined

Several models of emotional intelligence have emerged in recent years. As indicated previously, in one model Goleman (1995) includes a blend of several characteristics. Included are: (a) knowing one's emotions, (b) managing emotions, (c) motivating oneself, (d) recognizing emotions in others, and (e) handling relationships. Goleman suggested that a wide array of specific qualities such as impulse control, persistence, empathy, good moods, hope, and optimism are subsumed within these broader components and are characteristic of emotionally intelligent individuals. As a whole, Goleman conceived emotional intelligence to be "a master aptitude, a capacity that profoundly affects all other abilities, either facilitating or interfering with them" (p. 80). This statement reflects his belief that emotional intelligence is extremely powerful in how well people perform in life.

Similar to Goleman (1995), Bar-On (1997) includes a wide range of social and personality characteristics in his model of emotional intelligence (e.g., intrapersonal skills, interpersonal skills, adaptability, stress-management, general mood). Also like Goleman, Bar-On agreed that emotional intelligence has predictive ability, specifically suggesting that it can help optimize academic potential and life success. In this second model of emotional intelligence, Bar-On went one step beyond Goleman and developed instruments to measure these components, the BarOn EQ-i (Bar-On, 1997) and the BarOn EQ-i-Youth Version (Bar-On & Parker, 2000).

In 1990, Salovey and Mayer published their first conceptualizations of emotional intelligence, which have since evolved into a third model of emotional intelligence (Mayer & Salovey, 1997). Within this model attempts have been made to avoid nonability factors (e.g., persistence, warmth) completely. Although these characteristics are clearly important components of the human experience, Mayer and Salovey believe they are separate from emotional intelligence. They suggest that nonability qualities be studied independently from emotional intelligence. Their recent conceptualization of emotional intelligence is as follows:

   Emotional intelligence involves the ability to perceive
   accurately, appraise, and express emotion; the ability
   to access and/or generate feelings when they facilitate
   thought; the ability to understand emotion and emotional
   knowledge; and the ability to regulate emotions
   to promote emotional and intellectual growth. (Mayer
   & Salovey, 1997, p. 10)

Within this definition, the included abilities have been arranged into four "branches, with the lower branches consisting of psychological processes that are more basic in nature (e.g., the perception, appraisal, and expression of emotion), and higher branches involving more complex abilities such as the understanding and reflective regulation of emotions. Like Bar-On (1997), Mayer and Salovey developed scales reflecting their model of emotional intelligence-the Mulitfactor Emotional Intelligence Scale (MEIS) (Mayer, Salovey, & Caruso, 1997) and an adolescent version, the MEIS-A (Mayer, Salovey, & Caruso, 1996). Both were unpublished at the time of data collection for this study. The instrumentation section of this article contains additional details on the Mayer and Salovey model and the branches it uses.

Emotional Intelligence and Adolescent Success

Goleman (1995) and Bar-On (1997) have theorized that emotional intelligence is highly related to a variety of social, behavioral, and academic benefits. In actuality, however, only a limited amount of empirical evidence exists indicating that emotional intelligence contributes to any form of successful living.

The results of three recent studies have provided some indication of the predictive nature of emotional intelligence in adolescents. Utilizing a group of 52 junior high school students from an urban setting, Rubin (1999) found that high MEIS-A emotional intelligence scores were inversely related to aggression. Likewise, Trinidad and Johnson (2002) found that higher emotional intelligence scores (also utilizing the MEIS-A) were related to lower admissions of smoking and alcohol use in a group of high school students. In a third study that utilized the MEIS-A, with a small pilot study sample of gifted adolescents (N = 11), Mayer (2001) suggested that emotional intelligence was seemingly related to the ability to organize emotions that can occur in peer relationships. Whether emotional intelligence predicted success per se in each of these studies is debatable. There is growing evidence, however, that emotional intelligence is related to positive behavioral and social outcomes.

In one additional study completed during the Wechsler Intelligence Scale for Children-Fourth Edition (WISC-IV) validation process (Wechsler, 2003), the BarOn EQ-i-Youth Version (BarOn & Parker, 2000) was utilized along with the WISC-IV and the Wechsler Individual Achievement Test--Second Edition (WIAT-II) in a clear attempt to predict a more traditional form of success-academic achievement. Each instrument was administered to a group of 141 students from a variety of backgrounds, all between the ages of 11 and 17. When controlling for the WISC-IV Full Scale IQ (FSIQ), a multiple regression analysis revealed that the BarOn Total EQ did not add significantly to knowledge of WIAT-II total academic achievement. Certain BarOn subscales (Intrapersonal, Stress Management, General Mood, Positive Impression) did appear to predict achievement beyond the WISC-IV- FSIQ. The authors concluded that additional research would be needed to sort out this matter.

Emotional Characteristics of the Gifted

Although little is known specifically about the emotional intelligence of gifted and talented adolescents, much is known about characteristics similar to emotional intelligence in gifted and talented adolescents. For example, gifted youth are often found to have increased sensitivity and exceptional empathy (Davis & Rimm, 1998; Piechowski, 1997), and Roeper (1982) stated that emotionally gifted people often have "the capacity to integrate emotions, intellect, and creativity against enormous odds" (p. 24). These qualities sound strikingly similar to aspects of the Mayer and Salovey (1997) model of emotional intelligence. It appears likely that many gifted young people are able to utilize this integration of emotional and cognitive abilities in a manner that enables more successful interaction with the environment.

In contrast, gifted and talented students can also be extraordinarily challenged by their emotional states. For example, guilt, depression, feelings of inadequacy, and fearfulness (Davis & Rimm, 1998), as well as difficulty adjusting to new environments and loneliness (Falk, Piechowski, & Lind, 1994) have all been identified as being common among adolescents with high general cognitive ability (high IQ scores). Additionally, Lure (1988) found that despite adequate social skills, many gifted children were at-risk for interpersonal problems; and Osborn (1996) suggested that exceptionally gifted children are often at-risk for social isolation. Given findings like these, it is clear that not all gifted adolescents have the same capacity to process their highly variable emotional experiences in an efficient and effective manner.

Purpose

Despite superior general cognitive ability as measured by standardized tests, the population of gifted and talented students can experience as much variation in their successes and failures as other adolescents (Hallahan & Kaufmann, 1994; Rimm, 1996). Certainly personality and environmental factors play a part in this variability, but it is hypothesized that knowledge of cognitive abilities other than general intelligence (IQ) may also help predict successful outcomes for gifted and talented individuals. Given the mixed results within available studies of emotional intelligence in adolescents, it is clear that additional studies are needed to better understand the predictive ability of this construct. Larger studies that include gifted adolescent samples are particularly in need.

The purpose of this study was to measure emotional intelligence, as conceptualized by Mayer and Salovey (1997), with an attempt to better understand how it can contribute to the success of gifted and talented adolescents. Measures of interpersonal relations, social stress, and grade point average were utilized to gauge success. The following hypotheses were formulated:

Hypothesis #1: In gifted and talented adolescents, knowledge of emotional intelligence will add significantly to the explanation of social success (interpersonal relations and social stress) once general intelligence (IQ) has been controlled statistically.

Hypothesis #2: In gifted and talented adolescents, knowledge of emotional intelligence will add significantly to the explanation of academic achievement (grade point average) once general intelligence (IQ) has been controlled statistically. Controlling for IQ was done to further address current questions about the predictive nature of emotional intelligence above and beyond IQ.

Method

Participants

The total subject pool (N = 39) was comprised of a sample of adolescents enrolled in a Midwestern residential high school designed for 11th and 12th grade gifted adolescents (composite IQ standard score M = 129). Although all participants were considered students with high academic ability, the state test scores and grades of the students varied significantly (GPA range for semester of study = 2.0 to 4.0; M = 3.27). According to school administrators, the purpose of this school is to provide a physical, intellectual, and social environment in which students with exceptional ability can thrive in an appropriately exceptional learning community. Participants ranged in age from 15 to 18, with a mean of 16 years 6 months. Fifty-nine percent of the participants were female and 41% were male. Eighty-seven percent of the participants described themselves as White, 5% as African American, 3% as Asian American, 3% as Hispanic American, and 3% as Other. Participants were also questioned about extracurricular strengths and activities. Seventy-four percent rated themselves as being "very talented" in at least one of the following areas: play a musical instrument, sing, visual arts, athletics, dance or ballet, theatre, or debate.

Instrumentation

Multifactor Emotional Intelligence Scale--Adolescent Version (MEIS-A).

The Multifactor Emotional Intelligence Scale--Adolescent Version (MEIS-A) was used to measure emotional intelligence. This unpublished instrument was developed by Mayer, Salovey, and Caruso (1996) and further described by Caruso, Van Buren, Mayer, and Salovey (1998). The complete MEIS-A consists of eight separate tasks that are similar to the adult version, the Multifactor Emotional Intelligence Scale (Mayer, Salovey, & Caruso, 1997). Both measures consist of subtests that correspond with the four branches of emotional intelligence as defined by Mayer and Salovey (1997), and are forerunners to the recently published Mayer-Salovey-Caruso Emotional Intelligence Test (Mayer, Salovey, & Caruso, 2002). For the current study, a short-form version of the MEIS-A was used. The four subtests in this version (one from each branch of the Mayer and Salovey definition) yield a reliable measure of adolescent emotional intelligence (Alpha reliability = .91) as indicated by the authors (D. R. Caruso, personal communication, June 16, 1999). The MEIS-A short-form tasks and their descriptions are presented below under each appropriate branch of the Mayer and Salovey model.

Branch one-Perceiving emotions. Branch one tasks measure one's ability to perceive and identify emotional content in other people and other objects (faces, graphic designs, written stories). Mayer and Salovey (1997) argued that this is an essential component of a test of emotional intelligence as "the developing person begins to evaluate emotion wherever it might be expressed" (p. 12). The Faces task was utilized to measure this branch of adolescent emotional intelligence. The participants were asked to look at eight color photographs of human faces and then indicate the emotion in the face by rating their perception of various feelings (anger, sadness, happiness, disgust, fear, surprise).

Branch two-Assimilating emotions. Branch two tasks assess the participants' ability to bring emotion and cognitive processes together. The Synesthesia task from the MEIS-A was utilized to measure this branch. The written Synesthesia items required participants to think about six different future events that could make them feel a certain way (e.g., happy, jealous). After imagining this event, and while feeling how it might feel, subjects were asked to endorse 10 different scales ranging from one to five. (e.g., 1 = warm to 5 = cold, 1 = sharp to 5 = dull). In other words, participants were required to process emotional and cognitive/perceptual material simultaneously, a concept that is key to the Mayer and Salovey model of emotional intelligence.

Branch three-Understanding emotions. Within this branch, close analysis of emotions is required, much more so than within the other branches. The Perspectives task was used to measure this ability. Each item includes a brief written scenario involving a potentially emotional event. After reading the scenario, participants were asked to rate the probability of several different feelings occurring within two main characters in each scenario.

Branch four-Managing emotions. Branch four involves the ability to manage emotions in oneself and others. Although many people may be able to perceive, assimilate, and even fully understand emotions, proper regulation of emotional experiences may prove more difficult. The Managing Emotions task was utilized to measure this skill. This task includes six items that require participants to read a short paragraph about a situation in which they or others might be involved (e.g., receiving a bad grade after studying). Participants were asked to rate how effective a given response to the situation might be.

MEIS-A scoring and technical information. Although various scoring methods have been developed for the MEIS-A, the consensus scoring method was used in the current study. The authors of the MEIS-A have concluded that the consensus scoring option is the most reliable (Caruso, Van Buren, Mayer, & Salovey, 1998). Consensus scoring involves determining the percentage of participants who endorsed each response, with the participants from each individual study in question being used. For example, on any given test item, if 80% of the participants chose the third of five possible responses, then any participant in the sample who chose that third response would receive a raw score of .80 for that item. A total raw consensus score is calculated by totaling the raw scores from all test items. Z-scores can then be calculated for a standard score representation of emotional intelligence For a more detailed review of all scoring methods, refer to Caruso et al.

Behavior Assessment System for Children (BASC)

The Behavior Assessment System for Children (BASC) (Reynolds & Kamphaus, 1992) is a measure of emotional difficulties, behavioral difficulties, and adaptive skills. The adolescent version-self-report (BASC-SRP-A) was utilized for the current study. Of the many scales included in the BASC-SRP-A, only the Social Stress and Interpersonal Relations scales were utilized for the current study. The Interpersonal Relations scale assesses participants' "success at relating to others and the degree of enjoyment derived from this interaction" (Reynolds & Kamphaus, p. 61). The Social Stress scale measures participants' "feelings of stress and tension in personal relationships; a feeling of being excluded from social activities" (Reynolds & Kamphaus, p. 58).

Test of Cognitive Skills/Second Edition (TCS/2)

The Test of Cognitive Skills/Second Edition (TCS/2) (1993) is a measure of general intelligence. It is a group-administered test that is part of the statewide tests of educational progress administered within schools in several states. Each of the participants in this study completed the TCS/2 in their 10th grade year of high school.

The TCS/2 is a described as a reliable measure of "verbal, nonverbal, and memory abilities that can contribute greatly to students' success in an educational program" (Test of Cognitive Skills--Second Edition, 1993, p. 1). It utilizes a standardization sample of 7,300 students from 42 schools across the United States. The TCS/2 includes 80 test items and four subtests: Sequences, Analogies, Memory, and Verbal Reasoning. The entire TCS/2 takes 45 minutes to complete and yields an overall score called the Cognitive Skills Index (CSI). The CSI is a normalized standard score (M = 100, SD = 16) that "has the same statistical properties as the traditional Intelligence Quotient (IQ)" (p. 28).

Procedures

Arrangements were made for students to choose from one of several group-administered test sessions to complete a demographic sheet, the MEIS-A, and the BASC-SRP-A. All testing was completed within the school building of the student participants. The participants were able to choose from several late afternoon test times in order to avoid interference with classes or other extracurricular activities. All students involved were cooperative and completed all testing within 60 minutes. The CSI scores from the Test of Cognitive Skills and student grades were collected separately on different days from student academic files. Grade point average was calculated from grades received at the end of the semester during which the current study occurred.

Data Analysis

Because "success" was measured in three different ways for this study, three separate multiple regression analyses were conducted. Hierarchical regression analyses were used to determine the influence of emotional intelligence above and beyond general intelligence (IQ). Each multiple regression procedure included the MEIS-A overall z-score (emotional intelligence) and the TCS/2 Cognitive Skills Index (a standard score measure of general intelligence) as predictor variables. Each procedure then utilized one of the three criterion variables of interest, each of which reflected adolescent success in the social or academic realm: social stress (SS) as measured by the BASC (T-Score), interpersonal relations (IR) as measured by the BASC (T-Score), or student grade point average (GPA, based on a four-point scale). For statistical comparisons, an alpha level of .05 was used.

Results

Descriptive statistics for each variable are outlined in Table 1. The CSI mean standard score of 128.7 (96th percentile) was consistent with what is commonly used as a rough measure of giftedness. This was expected and reflects the high level of general cognitive ability among the participants in this study. The range of Grade Point Averages (2.0 to 4.0) was greater than might be expected and the mean (3.27) was lower than expected for a gifted population. However, this variability is understandable given the high level of competition and the caliber of the instruction and curriculum at this residential school for gifted students. The MEIS-A z-scores ranged from nearly one standard deviation below the mean (-.87) to nearly one standard deviation above the mean (.88), suggesting that all of the participants in the current study were within or near the average range of emotional intelligence as measured by the MEIS-A.

A significant negative correlation (p <.01) was obtained between the BASC Social Stress scale and the BASC Interpersonal Relations scale (see Table 2). This moderate correlation was expected, as it would be in any sample. As social stress decreases, the quality of interpersonal relations increases, and vice versa. Additionally, a significant positive correlation (p < .05) was observed between the CSI and GPA. This was also expected given that IQ and academic achievement are almost always found to be highly correlated (Sattler, 1992).

Unexpectedly, a significant correlation was not observed between the CSI and the MEIS-A. Mayer and Caruso (1999) suggested that moderate correlations are expected between all types of intelligences. In this case, there was virtually no correlation between emotional intelligence, as measured by the MEIS-A, and general intelligence (IQ), as measured by the CSI (see Table 2). Moreover, there was virtually no correlation between the MEIS-A and any of the other variables.

In order to better understand how emotional intelligence contributes to interpersonal relations once IQ has been controlled for, a multiple regression analysis was conducted. Table 3 includes data related to this first multiple regression. The results revealed a multiple [R.sup.2] of .020, indicating that about 2% of the variation in interpersonal relations is accounted for by a linear composite of the CSI and MEIS-A predictor variables. This multiple R was not found to be statistically significant, F(2, 36) = .370, p = .693. Low correlations between the CS1, MEIS-A, and the Interpersonal Relations scale were responsible for these insignificant multiple regression results. It was hypothesized that knowledge of the ME1S-A would contribute significantly to predicting interpersonal relations once the CSI was accounted for statistically. However, the squared semipartial correlation was .0086, indicating that 0% of interpersonal relations was predicted from the MEIS-A once the CSI was removed from the equation.

A second multiple regression analysis was carried out using the BASC Social Stress Scale. It was hypothesized that this would augment the study of social success within this sample, adding additional information to what was learned when using the BASC Interpersonal Relations scale in the first multiple regression. Table 4 includes data related to this second multiple regression. The results revealed a multiple [R.sup.2] of .023, indicating that about 2% of the variation in social stress is accounted for by a linear composite of the CSI and MEIS-A predictor variables. This multiple R was not statistically significant, F(2, 36) = .425, p = .657. Similar to the first multiple regression, low correlations between the CSI, MEIS-A, and the Social Stress scale were responsible for these insignificant results. The squared semipartial correlation was .0001, indicating that 0% of social stress was predicted from the MEIS-A once the CSI was removed from the equation.

A third multiple regression analysis was used to better understand the influence of emotional intelligence on academic achievement, with student grade point average being utilized as the criterion variable. Table 5 includes data related to this multiple regression. The results revealed a multiple [R.sup.2] of .141, indicating that about 14% of the variation in GPA is accounted for by a linear composite of the CSI and MEIS-A variables. This multiple R was not statistically significant (F(2, 36) = 2.955, p = .065). The CSI alone shared 14% of the variance with GPA, which was statistically significant (F(1,37) = 5.914, p = .020). After removing the CSI from the equation, however, the MEIS-A was found to contribute insignificantly to the prediction of GPA. Indeed, with a squared semipartial correlation of .0037, the results indicate that 0% of GPA is predicted from the MEIS-A after the CSI is statistically removed.

Discussion

The results suggest that the social and academic success of the gifted adolescent participants in this study were essentially independent of the overall emotional intelligence level of these students. In light of these findings, the view that emotional intelligence is a critical component of successful outcomes could not be supported within this study. Goleman (1995) claimed that emotional intelligence is at least as important as IQ in predicting various forms of success and in some cases more important. Even the conservative estimate that new ability variables should predict between 1% and 5% additional variance beyond general intelligence (Mayer, Salovey, & Caruso, 2000) was not observed in the current results. Consequently, the predictive power of emotional intelligence remains in question, at least when studying gifted adolescents in a residential setting.

Several factors likely influenced the results. First, there was less variability than expected in the emotional intelligence of this sample. Although some variability was observed, all the MEIS-A scores for this sample remained in the average range normatively. Given the literature concerning the emotional and cognitive abilities of gifted adolescents (e.g., Falk, Piechowski, & Lind, 1994; Roeper, 1982), it was expected that the emotional intelligence levels of the gifted and talented participants in this study could vary appreciably. Instead, it appears possible that the residential school setting influenced the results. Struggling students may have honed emotional problem solving skills over time in a supportive environment that encourages exploration of emotional and social dilemmas. However, this conclusion does not explain why none of the gifted students studied demonstrated a higher than average level of emotional intelligence. With this issue in mind, the authors question the ability of the MEIS-A short form to measure the emotional intelligence of gifted students in a valid manner. Suggestions for addressing these issues are proposed in the next section.

Recommendations for Future Research

Despite the results of the current study, the findings of other similar studies (e.g., Maree & Ebersohn, 2002; Mayer, 2001; Rubin, 1999; Trinidad & Johnson, 2002) suggest that additional emotional intelligence research or related studies may still be useful for understanding variations in the success of gifted adolescents. We suggest several possibilities for future studies.

First, rather than study an overall level of emotional intelligence, various personality or temperamental traits that are related to Goleman's (1995) model of emotional intelligence (e.g., persistence, optimism) could be more practical and meaningful. Although not specifically part of the Mayer and Salovey (1997) model of emotional intelligence, and probably better underscored as personality traits, characteristics like persistence and optimism do appear to require the cognitive processing of emotional information. Researchers who take this approach could help identify the specific constructs--whether they are considered cognitive abilities or otherwise-that contribute to successful outcomes. One hypothesis, for example, might be that persistence contributes significantly to the social and academic success of gifted adolescents once IQ is controlled for statistically.

A second suggestion for future research concerns the setting for studies involving gilled and talented adolescents, it appears likely that future research would be more sensitive to the social and academic needs of the majority of gifted adolescents if conducted in regular school settings. The needs of gifted students would likely be more highlighted in regular school settings, where necessary supports and challenges are less likely to be consistent. The social and academic experiences of the gifted adolescents in the current study were likely influenced in a positive way by their residential school environment. Students in this setting may often feel more appropriately challenged as they interact with other similar students. Moreover, they likely feel more socially accepted and more interested in social relationships in this setting. Indeed, the majority of the participants in this study reported social stress and interpersonal relations that were within normal limits, seemingly reflecting this support, and possibly reducing the likelihood of significant findings.

Finally, future research that specifically addresses the emotional intelligence construct must involve improved measurement of this variable. Fortunately, advancements have been made since the data was collected for the current study. Mayer, Salovey, and Caruso (2002) recently published an updated "ability-based" measure of their four-branch model of emotional intelligence, the Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT). Because their goal has always involved the development of an ability-based instrument, Mayer, Salovey, and Caruso attempted to exclude the personality components that have consistently confounded other models and test instruments. The four branches of this measure continue to reflect factors ranging from the perception of emotion to the management of emotion. The MSCEIT was developed for educational and research purposes and was intended for adolescents 17 and older. As another option, Bar-On and Parker (2000) have added a youth version emotional intelligence test (BarOn EQ-i--YV) to the mix of measurement options (ages 718). These measurement advances are clearly an improvement over the unpublished instrument used in the current study.

In summary, careful consideration of the research setting, the changing measurement options, and the model connected to each measurement option will be essential in future studies involving emotional intelligence and gifted students. When the goal is a more applicable understanding of the variables that contribute to success in gifted adolescents, researchers should give consideration to analyzing more specific related constructs --constructs that may or may not be directly subsumed under a model of emotional intelligence. Similar to current concerns involving the measurement of IQ in children and adolescents, the measurement of a total emotional intelligence score may be impractical for those who work directly with students. The results of studies involving specific emotional processing factors instead of one total emotional intelligence score will likely be more valuable for the helping professionals who work directly with gifted and talented adolescents.

Table 1
Descriptive Statistics For
All Variables Measured (N= 39)

Variable    Minimum   Maximum     M        SD

1. CSI       109.00   145.00    128.69    10.39
2. MEIS-A     -0.87     0.88      0.01     0.46
3. IR         31.00    57.00     52.10     7.21
4. SS         38.00    68.00     50.05     9.47
5. GPA         2.00     4.00      3.28     0.56
6. AGE       191.00   221.00    202.87     6.87

Note: CSI = Cognitive Skills Index (Standard Scores); MEIS-A =
            Multifactor Emotional Intelligence Scale--Adolescent
            Version (z-Scores); IR = Interpersonal Relations
            (T-Scores); SS = Social Stress (T-Scores); GPA = Grade
            Point Average (Four-Point Scale); AGE = Age in months.

Table 2
Pearson Correlations Between
All Variables Measured (N=39)

Variable       1           2            3         4          5
1. CSI        --
2. MEIS-A   -.029          --
3. IR       -.107        -.090         --
4. SS        .151        -.016       -.453 **     --
5. GPA       .371 *       .046       -.240       .116        --

* p < .05 (2-tailed).
** p < .01 (2-tailed).

Note: CSI = Cognitive Skills Index; MEIS-A = Multifactor Emotional
            Intelligence Scale--Adolescent Version; IR = Interpersonal
            Relations; SS = Social Stress; GPA = Grade Point Average.

Table 3
Summary of Hierarchical Regression
Analysis for Variables Predicting
Interpersonal Relations in Gifted Adolescents
(N=39)

Variable       df    [R.sup.2]   MS        F

Step 1
CS1             1      .012    22.780   .431 NS
MEIS-A          2      .020    19.906   .370 NS

Note: NS = Nonsignificant (p > .05)

Table 4
Summary of Hierarchical Regression
Analysis for Variables Predicting
Social Stress in Gifted Adolescents
(N= 39)

Variable   df   [R.sup.2]     MS        F

CSI        1     .023       78.223    .869 NS
MEIS-A     2     .023       39.330    .425 NS

Note: NS = Nonsignificant (p > .05)

Table 5
Summary of Hierarchical Regression
Analysis for Variables Predicting
Social Stress in Gifted Adolescents
(N=39)

Variable   df   [R.sup.2]      MS        F

CSI        1     .138        1.663    5.914 *
MEIS-A     2     .141        .851     2.955 NS

* p < .05
Note: NS = Nonsignificant (p > .05)

REFERENCES

Bar-On, R. (1997). The emotional quotient inventory (EQ-i): Technical manual. North Tonawanda. NY: Multi-Health Systems.

Bar-On. R., & Parker. J. D. (2000). The emotional quotient inventory: Youth version (EQ-i-YV), North Tonawanda, NY: Multi-Health Systems.

Caruso, D. R., Van Buren, A., Mayer, J. D., & Salovey, P. (1998). The measurement of emotional intelligence in adolescents. Unpublished manuscript.

Davis. G. A., & Rimm, S. B. (1998). Education d[ the gifted and talented (4th ed,). Boston: Allyn and Bacon.

Falk, R. F.. Piechowski, M. M.. & Lind, S. (1994). Criteria for rating intensity of overeexcitabilities (Manual). Unpublished manuscript, Northland College in Ashland, WI.

Goleman, D. (1995). Emotional intelligence. New York: Bantam Books.

Hallahan, D. P., & Kaufman, J. M. (Eds.). (1994). Exceptional children. Boston: Allyn and Bacon.

Henig, R. M. (1996, June). Are you smarter than you think'? McCalls, 123, 84-91.

Lum. M. S. (1988). Gifted population and adjustment: A literature review, of giftedness on conduct, family adjustment, emotional functioning, social functioning, and perceived competence. Unpublished doctoral dissertation, Biola University, La Mirada, CA.

Maree, J. G., & Ebersohn, L. (2002) Emotional intelligence and achievement: Redefining giftedness? Gifted Education International, 16(3). 261-273.

Mayer, J. D. (2001). Emotional intelligence and giftedness. Roeper Review, 23(3), 131-137.

Mayer. J. D.. & Caruso. D. R. (1999). Emotional intelligence meets traditional standards for an intelligence. Intelligence, 27(4). 267-299.

Mayer. J. D., & Salovey, P. (1997). What is emotional intelligence? In P. Salovey & D. J. Sluyter (Eds.), Emotional development and emotional intelligence (pp. 3-31). New York: Basic Books.

Mayer, J. D., Salovey, P., & Caruso, D. R. (1996), Multifactor emotional intelligence scale--Adolescent version (MEIS-A). Unpublished instrument. University of New Hampshire.

Mayer, J. D., Salovey, P., & Caruso, D. R. (1997). Multifactor Emotional Intelligence Scale (MEIS). Unpublished instrument-Item and answer booklet. University of New Hampshire.

Mayer, J, D., Salovey, P., & Caruso, D. R. (2000). Models of emotional intelligence. In R. J. Sternberg (Ed.), Handbook of Human Intelligence (pp. 396-420). New York: Cambridge.

Mayer, J. D., Salovey, P,. & Caruso, D. R. (2002). The Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT)--User's Manual. North Tonawanda. NY: Multi-Health Systems.

Osborn, J. (1996). Special education needs of gifted and talented children. Youth Mental Health Update, 8(4), 7-14.

Peterson, K. S. (1997, February 18). Signs of intelligence: Do new definitions of smart dilute meaning? USA Today, p. D1.

Piechowski, M. M. (1997). Emotional giftedness: The measure of intrapersonal intelligence. In N. Colangelo & G. A. Davis (Eds.). Handbook of gifted education [pp. 366- 381). Boston: Allyn and Bacon.

Reynolds, C. R., & Kamphaus, R. W. (1992). Behavior Assessment System for Children. Circle Pines, MN: American Guidance Services.

Rimm, S. B. (1996). Why bright kids get poor grades: And what you can do about it. New York: Crown,

Roeper, A. (1982). How the gifted cope with their emotions. Roeper Review, 5, 21-24.

Rubin. M. M. (1999). Emotional intelligence and its role in mitigating aggression: A correlational study of the relationship between emotional intelligence and aggression in urban adolescents. Unpublished manuscript, Immaculata College, Immaculata, PA.

Salovey, P., & Mayer, J. D. (1990). Emotional intelligence. Imagination, Cognition and Personality, 9(3). 185-211.

Sattler, J. M. (1992). Assessment of children, San Diego: Jerome M. Sattler.

Test of Cognitive Skills (2nd ed.). Manual. (1993). Los Angeles: CTB Macmillan/McGraw-Hill.

Trinidad, D. R., & Johnson, C. A. (2002). The association between emotional intelligence and early adolescent tobacco and alcohol use. Personality and individual Differences, 32(1).95-105.

Wechsler, D. (2003). The Wechsler Intelligence Scale for Children--Fourth Edition (WISC-IV)--Technical and Interpretive Manual. San Antonio, TX: Psychological Corporation.

Scott Woitaszewski is currently an assistant professor in the Department of Counseling and School Psychology at the University of Wisconsin--River Falls. He has taught courses in child and adolescent assessment, intervention strategies, and school consultation in addition to directing student research and supervising school psychology interns. His current research interests include the identification and application of resiliency factors in students with disabilities. E-mail: [e-mail]

Matthew Aalsma is currently an assistant professor of pediatrics and psychology at the Indiana University School of Medicine. His research focuses on adolescent health behavior. Current research projects include assessing the role of adolescent romantic relationships in how health behavior is expressed in adolescence as well as the development of antisocial behavior in adolescence. E-mail: [e-mail]