Testing the incremental validity of Trait Emotional Intelligence: Evidence from an Italian sample of adolescents

Personality and Individual Differences 64 (2014) 24–29

Contents lists available at ScienceDirect

Personality and Individual Differences journal homepage: www.elsevier.com/locate/paid

Testing the incremental validity of Trait Emotional Intelligence: Evidence from an Italian sample of adolescents Federica Andrei a,⇑, Giacomo Mancini a, Elena Trombini a, Bruno Baldaro a, P.M. Russo b a b

Department of Psychology, University of Bologna, Italy Department of Specialised, Experimental, and Diagnostic Medicine, University of Bologna, Italy

a r t i c l e

i n f o

Article history: Received 21 December 2013 Received in revised form 5 February 2014 Accepted 6 February 2014 Available online 4 March 2014 Keywords: Trait emotional self-efficacy TEIQue Incremental validity Adolescence Internalizing symptoms Personality

a b s t r a c t The present study was aimed at validating the Italian version of the Trait Emotional Intelligence Questionnaire–Adolescent Full Form (TEIQue–AFF), as well as at exploring its incremental validity over emotional maladjustment. To this end, a sample of 351 (163 males) Italian adolescents was collected. Overall, the Italian TEIQue–AFF both replicated the original-English four-factor structure (Well-Being, Self-Control, Emotionality, Sociability), and its construct validity was confirmed. However, reliability coefficients for eight facets and two factors were low. In addition, at all levels of analysis (i.e., global, factor, facet) the TEIQue–AFF was found to be a significant incremental predictor of adolescent’s emotional maladjustment, over and above gender, IQ, and the Big Five personality dimensions. At the factor level, significant effects were related to the contribution of the factor Well-Being thus supporting perspectives arguing for a further refinement of trait EI content domain. Implications of the findings are discussed. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Trait EI refers to emotion-related dispositions and self-perceived abilities measured via self-reports. Specifically the construct of trait EI has been formally defined as a constellation of emotional self-perceptions located at the lower levels of personality hierarchies (Petrides, Pita, & Kokkinaki, 2007). Studies on samples of adolescents have shown that trait EI may act as protective agent against emotional maladjustment, as lower trait EI levels have been found to be associated to internalizing symptoms such as depression in clinical (Delhaye, Kempenaers, Stroobants, Goossens, & Linkowski, 2012), as well as in normal samples (Williams, Daley, Burnside, & Hammond-Rowley, 2010). One of the main criticism raised against trait EI refers to its lack of incremental validity, particularly considering the construct’s overlap with the basic personality dimensions (Harms & Credé, 2010). Yet, it has been argued that advancement of the construct depends on whether EI is able to explain a significant proportion of incremental and unique variance in psychological outcomes over and above known predictors (Fiori & Antonakis, 2011). However, it is noteworthy that the construct of trait EI is thought to be related to higher order personality traits rather than being ⇑ Corresponding author. Address: Department of Psychology, University of Bologna, viale Berti Pichat 5, Bologna 40127, Italy. Tel.: +39 0512091845; fax: +39 051243086. E-mail address: [email protected] (F. Andrei). http://dx.doi.org/10.1016/j.paid.2014.02.007 0191-8869/Ó 2014 Elsevier Ltd. All rights reserved.

independent of them, thus justifying a certain degree of overlap between dimensions. Given the relevance of measurement to move towards a science of EI (Zeidner, Roberts, & Matthews, 2004), in the present study the incremental validity of the Trait Emotional Intelligence Questionnaire (TEIQue; Petrides, 2009) was systematically investigated over anxiety and depression symptoms, as indicators of non-clinical internalizing problems during adolescence. The TEIQue was developed in order to cover the 15 facets of the trait EI’s sampling domain comprehensively, which in turn load on four-factors, namely Well-Being, Self-Control, Emotionality, and Sociability. To date, the TEIQue is one of the most widely used and valid measures of trait EI (Martins, Ramalho, & Morin, 2010). As the full adolescent form of the TEIQue (TEIQue–AFF) allows for investigations at the factor as well as the facet level, it appears to be particularly suitable for a complete test of the construct’s incremental validity. Investigations on the issue of incremental validity have been performed mainly in adult samples. These studies have robustly shown that trait EI, as assessed through the TEIQue, might be a valuable explanatory and incremental predictor of criteria such as life-satisfaction (Freudenthaler, Neubauer, Gabler, Scherl, & Rindermann, 2008), and academic achievement (Sanchez-Ruiz, Mavroveli, & Poullis, 2013) over and above the Big Five, perfectionistic concerns beyond Neuroticism (Smith, 2014), and cortisol secretion under pressure after controlling for the effects of age and anxiety (Laborde, Lautenbach, Allen, Herbert, & Achtzehn,

F. Andrei et al. / Personality and Individual Differences 64 (2014) 24–29

2014). Even though few studies have explored this issue across samples of adolescents, their results are consistent with those obtained with adults (Davis & Humphrey, 2012; Mavroveli & Sanchez-Ruiz, 2011). In such investigations the short form of the TEIQue devoted to adolescents (TEIQue–ASF) was employed and the incremental validity analyses were focused on the global trait EI score, as was the case for studies run on adult populations. Despite this, those few studies focusing on the four factors of the adult TEIQue versions have shown that the incremental validity of trait EI derives mainly from the contribution of those facets associated with the factors Well-Being and Self-Control (e.g., Mikolajczak, Luminet, & Menil, 2006). In addition, a recent study employing the adult short form of the TEIQue suggests that the construct’s predictive power may be altered by those facets loading onto the factors of Emotionality and Sociability domains (Siegling, Vesely, & Saklofske, 2013), thus suggesting that the current content domain of trait EI may still be too heterogeneous. To our knowledge investigations of this assumption have focused on adult populations only. 1.1. The present study While there is little research on the incremental validity of the TEIQue across adolescence, to our knowledge there is no published investigations using the full adolescent form of the questionnaire, and focusing on the predictive contribution of the lower level components of the construct. In order to fill such gaps, the present study aimed at assessing the structure of the Italian TEIQue–AFF, as well as at examining its construct, criterion, and incremental validity. Consistently with the theoretical expectation of a predictive effect of trait EI over relevant affect-related criteria, and in line with relevance of the Big Five for psychological health (Ozer & Benet-Martinez, 2006), internalizing symptoms, a construct generally referring to symptoms of depression and anxiety (Lonigan, Carey, & Finch, 1994), was selected as indicator of emotional maladjustment during adolescence. Considering the contribution of each trait EI factor in the prediction of emotional maladjustment is an important step to ascertain which element of the construct may compromise the predictive power of trait EI at the global level. This evidence may ultimately provide further directions for a potential refinement of the construct, which in its current definition may not yet reflect the underlying emotion-related personality trait (Siegling et al., 2013). Moreover, showing whether and how trait EI predicts an incremental portion of variance at the lowest level of the construct (i.e., facet level) beyond the Big Five will provide a finer perspective on this issue. Thus, on the basis of the existing literature, it was hypothesized that:

25

needs (n = 8), and those (n = 6) who missed more than 15% of the items on the TEIQue–AFF were excluded from subsequent analyses. The final sample for the Factor Analysis was composed by 351 pupils (163 males), with age ranging from 14 to 18 (M = 15.31, SD = 1.80). Of these, a subsample of 174 (89 males, Mage = 16.54, SD = 1.12) participants completed additional measures for personality, IQ, depression and anxiety. 2.2. Measures 2.2.1. Trait EI Trait EI was appraised through the Italian translation of the TEIQue–AFF (Petrides, 2009). The TEIQue–AFF comprises 153 brief statements responded to on a 7-point scale, ranging from completely disagree to completely agree. The TEIQue–AFF was developed on the full form of the adult TEIQue and is intended to yield scores on the same 15 facets and four factors. The items of the TEIQue–AFF were translated into Italian by one of the authors (Russo, P.M.) and then back-translated into English by an independent English-native speaker. Item order was preserved and the Italian TEIQue–AFF was pretested on a small group of subjects (N = 30) assessing comprehension and ease of answering 2.2.2. Personality The Big Five Questionnaire-2 (BFQ-2; Caprara, Barbaranelli, Borgogni, & Vecchione, 2007) is aimed to assess the Big Five personality dimensions in adolescents and adults through 134 items rated on a 5-point Likert scale. Cronbach’s reliability coefficients for the BFQ-2 scales were very good (.82 for Extraversion, .85 for Agreeableness, .83 for Consciousness, .90 for Emotional Stability, and .84 for Openness). 2.2.3. Cognitive ability Raven’s Standard Progressive Matrices (SPM; Raven, 2008) consist of a measure of pure non-verbal reasoning ability, comprising 60 items presented in five sets of 12 each, and providing a global IQ score. 2.2.4. Emotional maladjustment The SAFA (Cianchetti & Sannio Fancello, 2001) is an Italian selfreport instrument aimed to assess mental health conditions in children and adolescents aged from 8 to 18. The SAFA displayed satisfactory psychometric properties (Franzoni et al., 2009). For the purpose of the present study, a general index of internalizing symptoms was obtained by computing a mean score from the scales assessing depression and anxiety. The correlation between the two scale was .56 (p < .01) 2.3. Procedure

H1: The 15 facets of the Italian TEIQue–AFF will show a four factor structure, thus replicating the one of the original English version of the questionnaire. H2: Trait EI will be distinct from cognitive ability, while it will be associated with higher order personality traits, particularly with the factor Emotional Stability. H3: Trait EI will show incremental validity at all construct levels over and above gender, cognitive ability and the Big Five dimensions. Significant incremental effects will be mainly due to the factors Well-Being and Self-Control. 2. Method 2.1. Participants Participants were 365 adolescents, recruited from secondary schools in major Italian cities. Pupils with special educational

A letter explaining the aims and rationale of the study was sent to the headmasters and teachers in each school. Informed consent was obtained from parents/carers. After brief group explanations on the purpose of the activities, confidentiality, and the answer formats, participants filled out all measures in their classrooms. 2.4. Statistical analysis The reliabilities of TEIQue–AFF were estimated using Cronbach’s alpha, while Student’s t test for independent samples was used for gender differences. Principal Axis Exploratory Factor Analysis was conducted on the 15 facets of the TEIQue–AFF. To determine the number of factors to retain, we combined Parallel Analysis (Horn, 1965), with Velicer’s minimum average partial (MAP) test (Velicer, 1976), as suggested by O’Connor (2000). In the present study, Parallel Analysis was applied using the ‘‘Marley

26

F. Andrei et al. / Personality and Individual Differences 64 (2014) 24–29

Watkins Monte Carlo PCA for Parallel Analysis’’ program (Watkins, 2000), and the MAP test was performed using a SPSS syntax file provided by O’Connor (2000). We used the ‘‘Marley Watkins Monte Carlo PCA for Parallel Analysis’’ program (Watkins, 2000) to obtain the eigenvalues and standard deviations generated from completely random data (and necessary to perform parallel analysis) using the following specifications: 15 variables, 352 participants, and 1000 replications. We then compared our observed eigenvalues to the 95th percentile of the eigenvalues generated from these random data to reject factors that were most certainly obtained by chance (at p = .05). In order to test the incremental validity of trait EI on emotional maladjustment a three-block hierarchical regression was performed at the global, factor and facet level of the TEIQue. All analyses were performed using PASW (SPSS version 19.0 for Windows). 3. Results 3.1. Descriptive statistics and reliability of the TEIQue–AFF Descriptive statistics, number of items, and internal consistencies for the TEIQue–AFF facets, factors and global score, are presented in Table 1. Factors and global score appear to be normally distributed, as indicated by the Kolmogorov–Smirnov (K–S) test (e.g., global score K–S(353) = .66; p > .05). Seven of the 15 facets had solid internal reliabilities (between .72 and .84; Table 1), while others are quite low (alphas below .70). The reliability of the TEIQue–AFF was acceptable for three factors, varying between .67 and .82, with the exception of Self-Control (a = .63). Nevertheless, the reliability of trait EI global score was satisfactory (a = .85). With respect to gender differences, several significant comparisons have been found at the both the facet and factor levels (see Table 1). 3.2. Validation and factor structure of the TEIQue–AFF The theoretical factor structure of the TEIQue–AFF, on which the a priori scoring key is based (see Table 1), was evaluated by a Principal Axis Factor Analysis on the total sample. Based on the Velicer

MAP test, a four factor solution (explaining 39.1% of the variance) would be appropriate. Results supported our theoretical expectations in that four factors were retained and rotated to simple structure via the Promax algorithm (j = 4). The factor pattern matrix is presented in Table 2 and approximate to a simple structure. Despite four cross-loadings, the four factors were substantively identical to the original British structure (Petrides, 2009) and were thus labelled accordingly: Well-Being, Sociability, Self-Control, and Emotionality. This solution accounted for 62.24% of the total variance. The strength of factors intercorrelations was generally above .30, with the exception of the factor Self-Control which correlates less-consistently with Emotionality (.18) and Sociability (.19). 3.3. Intercorrelations among variables Intercorrelations amongst study variables are presented in Table 3. At the global level, the TEIQue–AFF did not correlate significantly with cognitive ability, but it showed significant positive associations with all the Big Five dimensions, and negative with emotional maladjustment. Each factor of the TEIQue–AFF showed significant positive associations with at least three of the Big Five dimensions, and a negative relationship with emotional maladjustment. However, results showed that the factor Self-Control did not relate significantly with emotional maladjustment. Due to the large number of variables involved, bivariate associations at the facet level will be presented succinctly. With two exceptions (i.e., Trait Empathy and Impulsiveness) significant negative intercorrelations were found between each facet and emotional maladjustment, with r values ranging from .548 (p < .001) for Trait Optimism, to .179 (p < .001) for Relationships (see the Supplementary material). It is worth to notice that those facets loading on the factor Well-Being showed stronger correlations with the criterion variable compared to the other facets. Many of these comparisons resulted to have a significant value as confirmed by Fisher’s r-to-z transformations, which were performed to compare correlation coefficients. Particularly, correlation coefficients for Well-Being facets were significantly higher when compared with rs 6 |.395| for Trait Optimism (z = 1.83,

Table 1 Descriptives and gender differences for the TEIQue–AFF facets, factors and global score. Global sample (N = 351)

Males (n = 163)

Cronbach’s a

M

SD

M

SD

M

SD

Self-Esteem Emotion-Expression Self-Motivation Emotion-Control Trait-Happiness Trait-Empathy Social Awareness Impulsiveness (low) Emotion Perception Stress Management Emotion-Management Trait-Optimism Relationships Adaptability Assertiveness

11 10 10 12 8 9 11 9 10 10 9 8 9 9 9

.77 .79 .62 .72 .84 .74 .74 .66 .67 .65 .65 .76 .52 .51 .50

4.68 4.41 4.51 3.89 5.44 4.59 4.74 4.17 4.70 4.11 4.58 4.88 5.19 4.00 4.63

.93 1.07 .79 .87 1.14 .94 .87 .94 .85 .87 .89 1.03 .80 .74 .80

4.59* 4.57* 4.54 3.61*** 5.49 4.72*** 4.74 4.21 4.75 3.95*** 4.57 4.81 5.28* 3.94 4.60

.94 1.11 .82 .86 1.18 .92 .89 .95 .85 .90 .88 1.10 .81 .76 .80

4.78* 4.22* 4.49 4.22*** 5.38 4.43*** 4.74 4.12 4.63 4.31*** 4.60 4.96 5.09* 4.07 4.67

.91 .99 .76 .77 1.09 .94 .85 .92 .84 .78 .90 .94 .75 .71 .81

Well-Being Self-Control Emotionality Sociability

27 31 38 30

.82 .63 .74 .67

5.00 4.06 4.72 4.65

.89 .68 .69 .67

4.96 3.92*** 4.83** 4.64

.92 .68 .68 .67

5.04 4.22*** 4.59** 4.67

.87 .65 .67 .68

153

.85

4.57

.51

4.56

.50

4.58

.53

Facets/factors

Global Trait EI

N of items

Note. Significant two-tailed gender differences are displayed in bold. p < .05. ** p < .01. *** p < .001. *

Females (n = 188)

27

F. Andrei et al. / Personality and Individual Differences 64 (2014) 24–29 Table 2 Factor pattern matrix for the TEIQue–AFF subscales (N = 351).

Table 4 Summary of the hierarchical regressions testing the incremental validity of the TEIQue–AFF at the global, factor, and facet level (N = 174).

Factors WellBeing Trait Happiness Trait Optimism Self-Esteem Self-Motivation Emotion-Perception Trait Empathy Emotion Expression Relationships Emotion Control Stress Management Impulsiveness (low) Adaptability Emotion Management Social Awareness Assertiveness

Emotionality

SelfControl

Emotional maladjustment

Sociability Predictor

.86 .80 .61 .58 .05 .15 .10 .39 .08 .19 .20 .04 .10

.07 .01 .19 .01 .67 .66 .57 .49 .10 .13 .15 .18 .15

.02 .07 .07 .05 .08 .12 .15 .08 .91 .62 .30 .19 .07

.11 .05 .36 .11 .14 .05 .12 .17 .02 .06 .18 .16 .63

.07 .40

.30 .14

.03 .07

.59 .47

F

df1 (df2)

DR2

2 (171)

.029

Block 1 Gender & IQ

2.56

Block 2 Big Five

8.44*

10.51*

5 (166)

.23

13.76* 12.16* 7.8*

37.82* 14.03* 5.8*

1 (165) 4 (162) 15 (151)

.14 .19 .27

Block 3 TEIQue–AFF Global Factors Facets *

DF

p < .001.

tested by entering the 15 TEIQue–AFF facets as the third block, accounting for an additional 27% of variance. Precisely, only two facets of Emotionality, namely emotion expression (b = .18, t = 2.45, p < .05) and trait empathy (b = .23, t = 3.26, p = .001), and a facet of the factor Well-Being, Trait Optimism (b = .22, t = 2.33, p < .05), had significant coefficients.

Note: TEIQue–AFF = Trait Emotional Intelligence–Adolescent Full Form. Coefficients that should theoretically define each factor are displayed in bold.

p < .05, one-tailed), rs 6 |.344| for Self-Esteem (z = 1.85, p < .05, one-tailed), and rs 6 |.237| for Trait Happiness (z = 2.39, p < .01, one-tailed).

4. Discussion The present study systematically suggests that the TEIQue–AFF can be considered as a strong incremental predictor of adolescent’s emotional maladjustment at the global, factor, and facet level of the construct, over and above the Big Five, non-verbal cognitive ability and gender. Additionally, besides the English original and a non-Western (i.e., Chinese; Mavroveli & Siu, 2012) versions, to our knowledge this is the first investigation confirming that even the full adolescent form of the TEIQue, in its Italian translation, has a four-factor structure (H1). Despite this, internal consistencies of the Italian TEIQue–AFF’s facets as well as of the factors SelfControl and Sociability should be improved as they were generally not as good as those reported for both the original adult TEIQue, and its translations (e.g., German: Freudenthaler et al., 2008; Greek: Petrides et al., 2007). Moreover, with the exception of the factor Self-Control, which significantly correlated with IQ but not with Emotional Stability, trait EI did not show relevant associations with IQ, whereas it did with both higher order personality dimensions and non-clinical form of emotional maladjustment, at the global, factor and facet levels (H2). These results are analogue to those obtained by several

3.4. Incremental validity Table 4 provides the statistical indices (F, DF, DR2) for each Block of the three regression models. In each model, gender and IQ were entered as Block 1, resulting in a non significant model (b = .14, t = .30, p = n. s., and b = -.08, t = .77, p = n. s., respectively). At Block 2 the Big Five dimensions were added to the equation. Emotional Stability (b = .402, t = 5.38, p < .001), Extraversion (b = .396, t = 4.77, p < .001), and Openness (b = .335, t = 3.21, p < .01), were individually significant predictors and the model gained additional explanatory power. In the first model, global trait EI was entered as the third block, causing a significant incremental contribution, whereby trait EI was a significant predictor (b = .425, t = 6.15, p < .001). Analyses were rerun entering as the last block the four TEIQue factors in place of the global trait EI score. This change resulted in a significant model, with a significant contribution only from the factor Well-Being (b = .451, t = 5.75, p < .001). To further investigate the incremental validity of trait EI, a last model was

Table 3 Descriptive statistics and intercorrelation matrix for study variables. Variable

1

2

3

4

5

6

7

8

9

10

11

12

[1] Emotional Maladjustment [2] IQ [3] Emotional Stability [4] Openness [5] Consciousness [6] Extraversion [7] Agreeableness [8] Well-Being [9] Self-Control [10] Emotionality [11] Sociability [12] TEI M SD

.087 .192* .001 .53 .129 .093 .364** .096 .296** .245** .365** 62.06 7.76

.231** .322** .291** .144 .205** .023 .219** .066 .043 .062 53.27 4.31

.357** .275** .195** .265** .242** .561** .171* .113 .363** 47.91 11.48

. .639** .516** .651** .109 .153* .107 .196** .207** 48.39 13.89

.529** .602** .166* .230** .174* .149* .266** 47.63 14.68

.332** .338** .043 .141 .559** .213** 53.00 13.26

.083 .007 .367** .108 .378** 52.42 14.32

.296** .427** .606** .829** 4.97 .86

.222** .158** .550** 4.02 .70

.430** .724** 4.70 .64

.749** 4.58 .69

4.53 .50

Note: TEI = Trait Emotional Intelligence. p < .05. ** p < .01. *

28

F. Andrei et al. / Personality and Individual Differences 64 (2014) 24–29

studies using the short adolescent form of the TEIQue (e.g., Davis & Humphrey, 2012), and showing that trait EI has stronger associations with variables pertaining to the realm of personality and emotion, than with those referring to the cognitive domain. Particularly, higher levels of trait EI were positively associated with Emotional Stability and negatively with emotional maladjustment. The confirmation of this hypothesis reflects the theoretical nature of the construct as a personality dimension, and provides evidence of the cross-cultural stability of the TEIQue. Additionally, given the moderate significant effect sizes (Cohen, 1988), it also suggested for an absence of substantial overlap between measures. The incremental validity of trait EI was substantial at all construct levels (H3), thus putting to the test criticisms raised against its utility as a predictor of relevant psychological outcomes (Landy, 2005). Particularly, in line with the existing literature on adults, the relevant role of Well-Being emerged, as it was the only factor of the TEIQue–AFF having significant incremental effects over emotional maladjustment. Despite this, our last hypothesis was partially supported, as a non-significant effect (at both the bivariate and regression level of analysis) of the factor Self-Control was found, thus contrasting results obtained across adult samples as well. Nevertheless, it is worth to notice that the low levels of internal consistencies found in the present analyses, particularly for the factor Self-Control, may have influenced the predictive value of this factor. While at the facet level trait EI still shows incremental validity over the criterion variable, our results are not consistent with findings obtained at the factor level. However, the low levels of reliability (i.e., alpha values) of some facets may have compromised their contribution. Despite this, and in line with our findings at the factor level, bivariate analyses provided support for the crucial role of Well-Being. Such results were further reinforced by comparing the correlation coefficients of the relationship between the three Well-Being facets and emotional maladjustment, with those obtained from relating the other facets with the criterion variable. Although our results provide further evidence for the inclusion of trait EI within the personality hierarchy, as it resulted to be interrelated to higher order level traits but still distinct from them, it appears that construct’s explanatory power may be due mainly to the intrapersonal components of the sampling domain (i.e., Well-Being factor). Such evidence, together with the existing literature on adult samples, provide further support to the notion that the sampling domain of trait EI may be too broad (Siegling et al., 2013) and that additional research efforts are needed in order to further refine it. 4.1. Limitations and directions for future studies This investigation presents several methodological limitations, particularly the mono-method assessment (i.e., self-evaluative) of both predictors and criteria, as well as the cross-sectional nature of the study design, raise concerns about common-method bias. In addition, the small sample size of Italian students may limit the generalizability of results. Consequently, further research using larger samples, triangulation of data and longitudinal designs is needed. Particularly, it would be important to understand how individual differences in such disposition act in the path from emotional difficulties to the development of psychopathology during adolescence. Moreover, in order to further expand these findings it would be beneficial for future studies to control for other potential source of bias, such as social desirability. As mentioned above, the heterogeneity of facets together with the low internal consistencies found for some of them and for two factors, may have compromised the contribution of trait EI at the factor and global composite level, respectively. Indeed, the incremental validity of a measure may be underestimated when

the lower scales of such measure suffer from non-satisfactory levels of reliability (Smith, Fischer, & Fister, 2003). Thus, future studies are urged to improve the levels of internal consistencies of such measure and to subsequent attempt to replicate these analyses. Such steps will help to ascertain that low reliabilities found in the present study for the TEIQue–AFF did not bias these findings. Last, future studies may take into consideration the use of Item Response Theory (IRT) to improve and examine self-report measures (e.g., Levine & Leucht, 2012; Levine, Rabinowitz, & Rizopoulos, 2011). In the field of EI, IRT has been already employed to test the psychometric properties of the TEIQue short form (Cooper & Petrides, 2010). 4.2. Conclusions Results from the present study suggest that trait EI as assessed through the TEIQue–AFF is a stronger predictor of emotional maladjustment than the higher order personality traits covered by the Big Five model. This pattern was consistent at all construct levels. Such results challenge once more those claims for a redundancy between the construct of trait EI and the Big Five dimensions. Additionally, these findings add to the growing body of literature demonstrating the crucial role of the emotion-related aspects of personality over adolescent’s emotional health, and to the TEIQue as a potential tool for the assessment of such individual’s dispositions. This evidence should be useful for practitioners working in applied settings (e.g., schools). However, the predictive effect of the TEIQue appears to be due mainly to the factor Well-Being. Thus, more research is needed to further investigate the contribution of the lower levels of the construct, as new consistent findings may ideally lead to a finer specification of the trait EI construct. Acknowledgments We wish to thank Dr. K.V. Petrides for providing us with reliability analyses, and B. Bastelli for her help in data entry. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.paid.2014.02.007. References Caprara, G. V., Barbaranelli, C., Borgogni, L., & Vecchione, M. (2007). Big Five Questionnaire-2: Manuale. Firenze: Organizzazioni Speciali. Cianchetti, C., & Sannio Fancello, G. (2001). Scale Psichiatriche di Autosomministrazione per Fanciulli e Adolescenti (SAFA) [Psychiatric scales of self-administration for children and adolescents]. Firenze, Italy: Organizzazioni Speciali. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, New Jersey: Lawrence Erlbaum Associates. Cooper, A., & Petrides, K. V. (2010). A psychometric analysis of the Trait Emotional Intelligence Questionnaire-Short Form (TEIQue–SF) using item response theory. Journal of Personality Assessment, 92(5), 449–457. Davis, S., & Humphrey, N. (2012). Emotional intelligence predicts adolescent mental health beyond personality and cognitive ability. Personality and Individual Differences, 52, 144–149. http://dx.doi.org/10.1016/j.paid.2011.09.016. Delhaye, D., Kempenaers, C., Stroobants, R., Goossens, L., & Linkowski, P. (2012). Attachment and socio-emotional skills: A comparison of depressed inpatients, institutionalized delinquents and control adolescents. Clinical Psychology and Psychotherapy. http://dx.doi.org/10.1002/cpp.1787. Fiori, M., & Antonakis, J. (2011). A process-oriented approach to emotional intelligence. London, UK: The International Society for the Study of Individual Differences. Franzoni, E., Monti, M., Pellicciari, A., Muratore, C., Verrotti, A., Garone, C., et al. (2009). SAFA: A new measure to evaluate psychiatric symptoms detected in a sample of children and adolescents affected by eating disorders. Neuropsychiatric Disease and Treatment, 5, 207–214. http://dx.doi.org/10.2147/ NDT.S4874. Freudenthaler, H. H., Neubauer, A. C., Gabler, P., Scherl, W. G., & Rindermann, H. (2008). Testing and validating the trait emotional intelligence questionnaire (TEIQue) in a German-speaking sample. Personality and Individual Differences, 45(7), 673–678. http://dx.doi.org/10.1016/j.paid.2008.07.014.

F. Andrei et al. / Personality and Individual Differences 64 (2014) 24–29 Harms, P. D., & Credé, M. (2010). Remaining issues in emotional intelligence research: Construct overlap, method artifacts, and lack of incremental validity. Industrial and Organizational Psychology, 3, 154–158. http://dx.doi.org/10.1111/ j.1754-9434.2010.01217.x. Horn, J. L. (1965). A rationale and test for the number of factors in factor analysis. Psychometrika, 30(2), 179–185. http://dx.doi.org/10.1007/BF02289447. Laborde, S., Lautenbach, F., Allen, M. S., Herbert, C., & Achtzehn, S. (2014). The role of trait emotional intelligence in emotion regulation and performance under pressure. Personality and Individual Differences, 57, 43–47. http://dx.doi.org/ 10.1016/j.paid.2013.09.013. Landy, F. J. (2005). Some historical and scientific issues related to research on emotional intelligence. Journal of Organizational Behavior, 26, 411–424 [dx.doi.org/10.1002/job.317]. Levine, S. Z., & Leucht, S. (2012). Psychometric analysis in support of shortening the Scale for the Assessment of Negative Symptoms. European Neuropsychopharmacology, 23, 1051–1056. http://dx.doi.org/10.1016/j.euroneuro.2012.11.008. Levine, S. Z., Rabinowitz, J., & Rizopoulos, D. (2011). Recommendations to improve the Positive and Negative Syndrome Scale (PANSS) based on item response theory. Psychiatry Research, 188(3), 446–452. Lonigan, C. J., Carey, M. P., & Finch, A. J. (1994). Anxiety and depression in children and adolescents: Negative affectivity and the utility of self-report. Journal of Consulting and Clinical Psychology, 62, 1000–1008. Martins, A., Ramalho, N., & Morin, E. (2010). A comprehensive meta-analysis of the relationship between emotional intelligence and health. Personality and Individual Differences, 49, 554–564. http://dx.doi.org/10.1016/j.paid.2010.05.029. Mavroveli, S., & Sanchez-Ruiz, M. J. (2011). Trait emotional intelligence influences on academic achievement and school behaviour. British Journal of Educational Psychology, 81, 112–134 [dx.doi.org/10.1348/2044-8279.002009]. Mavroveli, S., & Siu, A. F. (2012). The factor structure of Trait Emotional Intelligence in Hong Kong Adolescents. Journal of Psychoeducational Assessment, 30(6), 567–576. Mikolajczak, M., Luminet, O., & Menil, C. (2006). Predicting resistance to stress: Incremental validity of trait emotional intelligence over alexithymia and optimism. Psicothema (18 Suppl.), 79–88. O’Connor, B. P. (2000). SPSS and SAS programs for determining the number of components using parallel analysis and Velicer’s MAP test. Behavior Research Methods, Instruments, & Computers, 32, 396–402.

29

Ozer, D. J., & Benet-Martinez, V. (2006). Personality and the prediction of consequential outcomes. Annual Review of Psychology, 57, 401–421. http:// dx.doi.org/10.1146/annurev.psych.57.102904.190127. Petrides, K. V. (2009). Technical manual for the Trait Emotional Intelligence Questionnaires (TEIQue). London: London Psychometric Laboratory. Petrides, K. V., Pita, R., & Kokkinaki, F. (2007). The location of trait emotional intelligence in personality factor space. British Journal of Psychology, 98, 273–289. http://dx.doi.org/10.1348/000712606X120618. Raven, J. C. (2008). SPM Standard Progressive Matrices (Matrici Progressive di Raven, Serie A, B, C, D, E; Italian translation). Giunti OS: Firenze. Sanchez-Ruiz, M. J., Mavroveli, S., & Poullis, J. (2013). Trait emotional intelligence and its links to university performance: An examination. Personality and Individual Differences, 54, 658–662. http://dx.doi.org/10.1016/j.paid.2012.11.013. Siegling, A. B., Vesely, A. K., & Saklofske, D. H. (2013). Advancing the trait EI content domain: Further evidence for the distinctiveness of interpersonal facets. Personality and Individual Differences, 54, 81–86. http://dx.doi.org/10.1016/ j.paid.2012.08.010. Smith, M. M. et al. (2014). The link between neuroticism and perfectionistic concerns: The mediating effect of trait emotional intelligence. Personality and Individual Differences, 61–62, 97–100. Smith, G. T., Fischer, S., & Fister, S. M. (2003). Incremental validity principles in test construction. Psychological Assessment, 15(4), 467–477. http://dx.doi.org/ 10.1037/1040-3590.15.4.467. Velicer, W. F. (1976). Determining the number of components from the matrix of partial correlations. Psychometrika, 41, 321–327. http://dx.doi.org/10.1007/ BF02293557. Watkins, M. W. (2000). Monte Carlo PCA for parallel analysis. State College, PA: Ed & Psych Associates. Williams, C., Daley, D., Burnside, E., & Hammond-Rowley, S. (2010). Does item overlap account for the relationship between trait emotional intelligence and psychopathology in preadolescents? Personality and Individual Differences, 48, 867–871. http://dx.doi.org/10.1016/j.paid.2010.02.006. Zeidner, M., Roberts, R. D., & Matthews, G. (2004). The emotional intelligence bandwagon: Too fast to live, too young to die? Psychological Inquiry, 15, 239–248. http://dx.doi.org/10.1207/s15327965pli1503_04.