Examining the structure of emotional intelligence at the item level: new perspectives, new conclusions

Despite twenty years of research, many unknowns remain regarding the Mayer-Salovey (e.g., 1997) model of emotional intelligence (EI) and the validity of tests that have been designed to measure it. Evidence relevant to the internal structure of EI has come mainly from factor-analytic studies of the MSCEIT and the MEIS, utilising parcelled task scores rather than individual test items. This approach has several deficiencies: in addition to the loss of item-level information, it results in an insufficient number of observed variables per factor and an inability to separate structural sources of local item dependence (i.e., method variance) from construct-related variance. The present study (N=707) employed multidimensional item response modelling to investigate the dimensional structure of the MSCEIT, at the item level, for the first time. It is shown that item format and the specific choice of task explain far more of the variance in response patterns than does the hypothesised dimensional structure of EI, to the point that there is no empirical reason to prefer a higher-dimensional model of EI over a unidimensional model. It is argued that the advantage of an item-level perspective can be fundamental, rather than merely incremental.     

Ability EI as an intelligence? Associations of the MSCEIT with performance on emotion processing and social tasks and with cognitive ability

Associations among the MSCEIT, a broad-bandwidth measure of ability emotional intelligence (EI), self-report EI, social perception ability, psychometric intelligence and performance on inspection time (IT) tasks, which assessed the speed of processing of emotional and non-emotional information were investigated in two student samples (N=99, 199). The main findings were that MSCEIT scores were unrelated to fluid ability or speed of non-emotional information processing as assessed by IT, but evidence for associations of MSCEIT scores with crystallised ability was found. Positive associations were found between MSCEIT scores, self-report EI and some emotion/social task scores. The results suggest that EI as assessed by the MSCEIT has some properties of an intelligence and is more closely related to crystallised than to fluid ability. The relatively small MSCEIT/g _c correlations suggest that the MSCEIT is not a pure ability measure, although restriction of range in the samples used may also be relevant. More work, and the development of new measures, is required to determine whether performance EI has a fluid component.     

Differences between multimedia and text-based assessments of emotion management: An exploration with the multimedia emotion management assessment (MEMA)

People process emotional information using visual, vocal, and verbal cues. However, emotion management is typically assessed with text based rather than multimedia stimuli. This study (N?=?427) presents the new multimedia emotion management assessment (MEMA) and compares it to the text-based assessment of emotion management used in the MSCEIT. The text-based and multimedia assessment showed similar levels of cognitive saturation and similar prediction of relevant criteria. Results demonstrate that the MEMA scores have equivalent evidence of validity to the text-based MSCEIT test scores, demonstrating that multimedia assessment of emotion management is viable. Furthermore, our results inform the debate as to whether cognitive saturation in emotional intelligence (EI) measures represents “noise” or “substance”. We find that cognitive ability associations with EI represent substantive variance rather than construct-irrelevant shared variance due to reading comprehension ability required for text-based items.     

The ability model of emotional intelligence: Searching for valid measures

Current measures of ability emotional intelligence (EI) – in particular the well-known Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT) – suffer from several limitations, including low discriminant validity and questionable construct and incremental validity. We show that the MSCEIT is largely predicted by personality dimensions, general intelligence, and demographics having multiple R’s with the MSCEIT branches up to .66; for the general EI factor this relation was even stronger (Multiple R = .76). As concerns the factor structure of the MSCEIT, we found support for four first-order factors, which had differential relations with personality, but no support for a higher-order global EI factor. We discuss implications for employing the MSCEIT, including (a) using the single branches scores rather than the total score, (b) controlling for personality and general intelligence to ensure unbiased parameter estimates in the EI factors, and (c) accounting for measurement error. Failure to correctly model these methodological aspects may severely compromise predictive validity testing. We also discuss potential avenues for the improvement of ability-based tests.     

Convergent, discriminant, and incremental validity of competing measures of emotional intelligence

This study investigated the convergent, discriminant, and incremental validity of one ability test of emotional intelligence (EI)--the Mayer-Salovey-Caruso-Emotional Intelligence Test (MSCEIT)--and two self-report measures of EI--the Emotional Quotient Inventory (EQ-i) and the self-report EI test (SREIT). The MSCEIT showed minimal relations to the EQ-i and SREIT, whereas the latter two measures were moderately interrelated. Among EI measures, the MSCEIT was discriminable from well-studied personality and well-being measures, whereas the EQ-i and SREIT shared considerable variance with these measures. After personality and verbal intelligence were held constant, the MSCEIT was predictive of social deviance, the EQ-i was predictive of alcohol use, and the SREIT was inversely related to academic achievement. In general, results showed that ability EI and self-report EI are weakly related and yield different measurements of the same person.     

A general Bayesian multilevel multidimensional IRT model for locally dependent data

Many item response theory (IRT) models take a multidimensional perspective to deal with sources that induce local item dependence (LID), with these models often making an orthogonal assumption about the dimensional structure of the data. One reason for this assumption is because of the indeterminacy issue in estimating the correlations among the dimensions in structures often specified to deal with sources of LID (e.g., bifactor and two-tier structures), and the assumption usually goes untested. Unfortunately, the mere fact that assessing these correlations is a challenge for some estimation methods does not mean that data seen in practice support such orthogonal structure. In this paper, a Bayesian multilevel multidimensional IRT model for locally dependent data is presented. This model can test whether item response data violate the orthogonal assumption that many IRT models make about the dimensional structure of the data when addressing sources of LID, and this test is carried out at the dimensional level while accounting for sampling clusters. Simulations show that the model presented is effective at carrying out this task. The utility of the model is also illustrated on an empirical data set.     

The unbearable articulatory nature of naming: on the reliability of word naming responses at the item level

Single-word naming is one of the most widely used experimental paradigms for studying how we read words. Following the seminal study by Spieler and Balota (Psychological Science 8:411–416, 1997), accounting for variance in item-level naming databases has become a major challenge for computational models of word reading. Using a new large-scale database of naming responses, we first provided a precise estimate of the amount of reproducible variance that models should try to account for with such databases. Second, by using an item-level measure of delayed naming, we showed that it captures not only the variance usually explained by onset phonetic properties, but also an additional part of the variance related to output processes. Finally, by comparing the item means from this new database with the ones reported in a previous study, we found that the two sets of item response times were highly reliable (r = .94) when the variance related to onset phonetic properties and voice-key sensitivity was factored out. Overall, the present results provide new guidelines for testing computational models of word naming with item-level databases.     

Measuring emotional intelligence with the MSCEIT V2.0

Does a recently introduced ability scale adequately measure emotional intelligence (EI) skills? Using the Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT; J. D. Mayer, P. Salovey, & D. R. Caruso, 2002b), the authors examined (a) whether members of a general standardization sample and emotions experts identified the same test answers as correct, (b) the test's reliability, and (c) the possible factor structures of EI. Twenty-one emotions experts endorsed many of the same answers, as did 2,112 members of the standardization sample, and exhibited superior agreement, particularly when research provides clearer answers to test questions (e.g., emotional perception in faces). The MSCEIT achieved reasonable reliability, and confirmatory factor analysis supported theoretical models of EI. These findings help clarify issues raised in earlier articles published in Emotion.     

Examining the dimensional structure and factorial validity of the Bar-On Emotional Quotient Inventory in a sample of male athletes

Objectives

Research examining how emotional intelligence (EI) relates to the performance of athletes has been conducted using various EI measures including the Bar-On Emotional Quotient Inventory (EQ-i; Bar-On, 1997), but no one has investigated the factor structure of the EQ-i in the area of sport psychology. The current study explored the dimensional structure and factorial validity of the EQ-i in a sample of male athletes.

Design

Confirmatory factor analysis was used to examine [Bar-On, 1997] and [Bar-On, 2004] model of emotional-social intelligence and the 1-5-15 dimensional structure which underpins the EQ-i.

Method

A total sample of 706 male athletes from various sports and competing at the national age group level through to the professional level of competition completed the EQ-i.

Results

Confirmatory factor analyses demonstrated that the 1-5-15 dimensional structure was a poor fit for the data. A re-specification of the model representing the best fit for the data was a 1-4-15 dimensional structure. The factorial validity of the individual subscales was also examined at the item level using confirmatory factor analysis. Thirteen of the 15 subscales showed close, reasonably good, or mediocre fit for the data.

Conclusion

Further construct validation of Bar-On’s model and measure is required. Sport psychologists administering the EQ-i in applied practice should consider using the EQ-i subscales rather than referring back to the 1-5-15 dimensional structure.     

The Utility of Item-Level Analyses in Model Evaluation: A Reply to Seidenberg and Plaut

Seidenberg and Plaut (this issue) argue that the implications of our analyses (Spieler & Balota, 1997) for the two extant connectionist models of word naming are limited by two factors. First, variables outside the scope of these models influence naming performance, so it is not surprising that the models do not account for much of the variance at the item level. Second, there is error variance associated with large item-level data sets that obviously should not be captured by these models. We point out that there are a number of variables that have been incorporated within the targeted connectionist models that should provide these models an advantage over the simple predictor variables that we selected as a baseline to evaluate the efficacy of the models (e.g., log frequency, length in letters, and number of orthographic neighbors). We also point out that there is considerable consistency across four large-scale studies of item means. Finally, we provide evidence that even under conditions of a standard word-naming study (with a small set of items), simple word frequency, orthographic neighborhoods, and length accounted for more variance than the extant connectionist models. We conclude that item-level analyses provide an important source of evidence in the evaluation of current models and the development of future models of visual word recognition.     

Revisiting the predictive validity of emotional intelligence: self-report versus ability-based measures

In response to general press assertions that training emotionally intelligent children will lead to great rewards, this study examined the relationship between emotional intelligence (EI) and academic achievement in college students, using both self-report and ability-based measures of EI. Specifically, the Mayer, Salovey, Caruso Emotional Intelligence Test (MSCEIT, an ability-based measure) and the Bar-On Emotional Quotient Inventory (EQ-i, a self-report measure) were used to predict academic achievement. Achievement was operationalized as the respondents’ cumulative GPA. Results indicated that EI is not a strong predictor of academic achievement regardless of the type of instrument used to measure it. However, a construct validity examination revealed that the MSCEIT correlated highly with indices of cognitive ability but minimally with personality dimensions. In contrast, the EQ-i failed to correlate with indices of cognitive ability but correlated substantially with numerous personality dimensions.     

Exploring the validity of the Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT) with established emotions measures

Emotions measures represent an important means of obtaining construct validity evidence for emotional intelligence (EI) tests because they have the same theoretical underpinnings. Additionally, the extent to which both emotions and EI measures relate to intelligence is poorly understood. The current study was designed to address these issues. Participants (N = 138) completed the Mayer-Salovey-Caruso Emotional Intelligence Test (MSCEIT), two emotions measures, as well as four intelligence tests. Results provide mixed support for the model hypothesized to underlie the MSCEIT, with emotions research and EI measures failing to load on the same factor. The emotions measures loaded on the same factor as intelligence measures. The validity of certain EI components (in particular, Emotion Perception), as currently assessed, appears equivocal.     

Emotional intelligence and its relation with hedonic and eudaimonic well-being: A prospective study

The primary aim of the current study was to replicate and extend previous findings by examining the relationship between emotional intelligence abilities and levels of hedonic and eudaimonic well-being in a 12-week follow-up study. Participants were 349 undergraduate students who completed an ability measure of emotional intelligence (MSCEIT), personality questionnaires, and hedonic and eudaimonic well-being measures at time 1. After 12 weeks, participants repeated the hedonic and eudaimonic measures at time 2. Focusing first on cross-sectional analysis, we found emotional intelligence scores to be moderately and significantly related to hedonic and eudaimonic well-being measures. Next, we found that ability EI predicted a modest but additional variance in prospective levels of hedonic and eudaimonic well-being over and above personality traits. Also, the explained variance of EI was higher for eudaimonic than for hedonic well-being. These findings provide some preliminary evidence on the prospective value of ability EI in the maintenance of positive mood and better outlook on life (hedonia) and, specifically, in the development of aspects of human functioning (eudaimonia).     

Conceptual knowledge of emotions includes somatosensory component: Evidence from modality-switch cost effect

Processing and storage constitute important aspects of working memory. However, the relationship between them remains unclear. If they constitute separate, independent capacities then manipulations of processing and storage load should have independent effects on task performance. In the present research, university students (N=84) were administered a reasoning task based on the Greco-Latin square task, in which processing (as defined by relational complexity) and storage load were orthogonally manipulated. Theoretical processing and storage load were each able to predict significant unique portions of the variance in both item difficulty and correct response time. However, no interaction term was able to significantly increase predictive power for either item difficulty or correct response time. This suggests that over the range of processing and storage load imposed by the task, processing and storage capacity are largely additive (i.e., independent). Implications for models of working memory are discussed.     

Item factor analysis: current approaches and future directions

The rationale underlying factor analysis applies to continuous and categorical variables alike; however, the models and estimation methods for continuous (i.e., interval or ratio scale) data are not appropriate for item-level data that are categorical in nature. The authors provide a targeted review and synthesis of the item factor analysis (IFA) estimation literature for ordered-categorical data (e.g., Likert-type response scales) with specific attention paid to the problems of estimating models with many items and many factors. Popular IFA models and estimation methods found in the structural equation modeling and item response theory literatures are presented. Following this presentation, recent developments in the estimation of IFA parameters (e.g., Markov chain Monte Carlo) are discussed. The authors conclude with considerations for future research on IFA, simulated examples, and advice for applied researchers.     

Romantic love: What’s emotional intelligence (EI) got to do with it?

The study examined the role of EI in romantic love among newlywed couples. In addition, we set out to assess to what degree outcomes depend on the specific EI measure employed, i.e., maximal performance (ability-based) vs. typical performance (self-report). In the dyadic design implemented in this research, 100 young heterosexual married couples independently completed the Mayer–Salovey–Caruso emotional intelligence test (MSCEIT), the Schutte self-report inventory (SSRI), and a vocabulary subtest of the MILTA group intelligence battery. EI, assessed via an ability and a self-report measure, showed significant ‘actor effects’, but failed to show significant ‘partner effects’, thus only partially supporting the hypothesis that EI predicts romantic love. Whereas ability-based and self-report measures show the same pattern of relationship with romantic love in dyadic assessments, the magnitudes of outcome relationships as well as the correlations of EI with verbal ability are measure-dependent. Marital partners were not correlated on EI, thus providing little evidence for assortative mating for EI.     

Measurement of ability emotional intelligence: Results for two new tests

Emotional intelligence (EI) has attracted considerable interest amongst both individual differences researchers and those in other areas of psychology who are interested in how EI relates to criteria such as well‐being and career success. Both trait (self‐report) and ability EI measures have been developed; the focus of this paper is on ability EI. The associations of two new ability EI tests with psychometric intelligence, emotion perception, and the Mayer–Salovey–Caruso EI test (MSCEIT) were examined. The new EI tests were the Situational Test of Emotion Management (STEM) and the Situational Test of Emotional Understanding (STEU). Only the STEU and the MSCEIT Understanding Emotions branch were significantly correlated with psychometric intelligence, suggesting that only understanding emotions can be regarded as a candidate new intelligence component. These understanding emotions tests were also positively correlated with emotion perception tests, and STEM and STEU scores were positively correlated with MSCEIT total score and most branch scores. Neither the STEM nor the STEU were significantly correlated with trait EI tests, confirming the distinctness of trait and ability EI. Taking the present results as a starting‐point, approaches to the development of new ability EI tests and models of EI are suggested.     

Item-Level and Scale-Level Factor Structures of the MMPI-A

The Minnesota Multiphasic Personality Inventory-Adolescent (MMPI-A) is a 478-item test that represents a substantial revision of the original form of the MMPI. This investigation sought to identify the item-level factor structure of the MMPI-A and also examined the scale-level factor structure of the 69 scales and subscales of this instrument. The study utilized the 1,620 normal adolescents (805 boys and 815 girls) of the normative sample for the MMPI-A. These adolescents ranged in age from 14 to 18 inclusive, with a mean age of 15.54 for boys and 15.60 for girls. A principal factor analysis of item-level responses resulted in extraction of 14 factors that were subjected to promax (oblique) rotation procedures. These 14 factors incorporated 81% of the total MMPI-A item pool and accounted for 44% of the total item-level response variance. For the scale-level analysis, 8 factors were selected for extraction and submitted to promax rotation procedures. These eight factors accounted for a total of 93.5% of the total variance in MMPI-A scale and subscale raw scores, Item-level results were discussed in terms of areas of congruence and dissimilarities from previously reported MMPI factor analyses in adolescent and adult samples, and scale-level factor results were presented in terms of clinical implications for the interpretation of MMPI-A scales and subscales.     

使用题组反应模型缓解局部题目依赖性对多阶段测验的危害

尽管多阶段测验(MST)在保持自适应测验优点的同时允许测验编制者按照一定的约束条件去建构每一个模块和题板,但建构测验时若因忽视某些潜在的因素而导致题目之间出现局部题目依赖性(LID)时,也会对MST测验结果带来一定的危害。为探究"LID对MST的危害"这一问题,本研究首先介绍了MST和LID等相关概念;然后通过模拟研究比较探讨该问题,结果表明LID的存在会影响被试能力估计的精度但仍为估计偏差较小,且该危害不限于某一特定的路由规则;之后为消除该危害,使用了题组反应模型作为MST施测过程中的分析模型,结果表明尽管该方法能够消除部分危害但效果有限。这一方面表明LID对MST中被试能力估计精度所带来的危害确实值得关注,另一方面也表明在今后关于如何消除MST中由LID造成危害的方法仍值得进一步探究的。     

Emotional Intelligence and cognitive abilities – associations and sex differences

In order to expand on previous research, this cross-sectional study investigated the relationship between Emotional Intelligence (EI) and cognitive abilities in healthy adults with a special focus on potential sex differences. EI was assessed by means of the Mayer-Salovey-Caruso-Emotional-Intelligence Test (MSCEIT), whereas cognitive abilities were investigated using the Brief Assessment of Cognition in Schizophrenia (BACS), which measures key aspects of cognitive functioning, i.e. verbal memory, working memory, motor speed, verbal fluency, attention and processing speed, and reasoning and problem solving. 137 subjects (65% female) with a mean age of 38.7 ± 11.8 years were included into the study. While males and females were comparable with regard to EI, men achieved significantly higher BACS composite scores and outperformed women in the BACS subscales motor speed, attention and processing speed, and reasoning and problem solving. Verbal fluency significantly predicted EI, whereas the MSCEIT subscale understanding emotions significantly predicted the BACS composite score. Our findings support previous research and emphasize the relevance of considering cognitive abilities when assessing ability EI in healthy individuals.