Trait emotional intelligence and behavioral problems among adolescents: A cross-informant design

Highlights

• Self and parental ratings of trait EI relate to those of behavioral problems.

• Results reveal a strong correspondence between father and mother ratings.

• Multisource ratings of trait EI are differentially predictive of external criteria.

Abstract

This study examined how self and parental ratings of trait emotional intelligence (trait EI), relate to self and parental ratings of internalizing and externalizing problems. A total of 263 adolescents between 13 and 17 years and their parents participated in the study, completing two forms of the Trait Emotional Intelligence Questionnaire (TEIQue-AF and TEIQue-360°) and the Child Behavior CheckList (CBCL-YSR and CBCL-parents’ report). Results revealed a strong correspondence between father and mother ratings, especially for externalizing problems. A doubly MANOVA, with gender as the between-subjects variable, rating source (father, mother, and adolescent) as the within-subjects variable, and the four trait EI factors as dependent variables revealed a significant effect of ratings. Both fathers and mothers attributed higher Well-being and Self-control to their children than the children attributed to themselves. The findings showed that while multisource ratings of trait EI generally converge, they are differentially predictive of external criteria.

Introduction

Trait emotional intelligence (trait EI) refers to a constellation of emotional self-perceptions located at the lower levels of personality hierarchies (Petrides, Pita, & Kokkinaki, 2007). The construct concerns people’s perceptions of their emotional abilities, which is why it has also been labelled as “trait emotional self-efficacy”. Trait EI has been negatively associated with depression, anxiety, anger and disruptive behavior (Martins, Ramalho, & Morin, 2010), and positively with adaptive coping styles, peer relations, and socio-emotional competence (Frederickson, Petrides, & Simmonds, 2012). Multiple studies have demonstrated that the construct is implicated in many important life domains, including health (Costa, Petrides, & Tillman, 2014), parenting (Gugliandolo, Costa, Cuzzocrea, & Larcan, 2014), mental well-being (Andrei & Petrides, 2013), university performance (Sanchez-Ruiz, Mavroveli, & Poullis, 2013), leadership and career decision-making (Di Fabio and Saklofske, 2014, Siegling et al., 2014).

Even though personality characteristics are frequently measured by self-report Likert-type rating scales (Vazire & Mehl, 2008), single-informant data are inherently biased by the possibility of response distortion (Connelly & Hulsheger, 2012). For this reason, a robust method to verify the validity of the personality trait model is via an examination of self-other convergence patterns. This method has been widely used as evidence of the validity of the five-factor model, showing consistent correlations between self- and other-reported personality traits (Connelly & Hulsheger, 2012).

Gathering information from a variety of sources in the assessment of behavior (triangulation) increases the reliability and validity of assessments, especially in children and adolescent samples (Achenbach, McConaughy, & Howell, 1987). Vazire and Mehl (2008) have shown that self- and other-judgments have differential relationships and predictive validity in relation to daily behavior. Other meta-analyses that have examined emotional and behavioral problems have shown that the convergence between self-reports provided by children and adolescents and the reports provided by their parents is relatively weak (Achenbach et al., 1987, Renk and Phares, 2004).

Duhig, Renk, Epstein, and Phares (2000), in a meta-analysis of 60 studies, showed that maternal and paternal ratings exhibit weak correspondence in ratings of internalizing behavior problems, but strong correspondence in ratings of externalizing behavior problems. In general, convergence in cross-informant ratings tends to be more consistent for behaviors that are more easily observable and bothersome (Achenbach et al., 1987, De Los Reyes and Kazdin, 2005).

Another variable that can play a role in rating differences is adolescent (target) gender. Penney and Skilling (2012) found that gender moderated the discrepancy between informant reports of young people’s internalizing and externalizing problems, such that only female youths reported more somatic symptoms compared to their caregivers. The investigation of gender differences is hindered by the fact that many studies have examined boys and girls together. One exception is the study by Wang et al. (2014), who studied parental (father–mother) correspondence indices separately for boys and girls and reported larger discrepancies for the former probably because they tend to be less communicative of their feelings and problems at home.

Despite the extensive use of self- and other-ratings in personality and adolescent behavior, only few studies have included other-ratings in the EI field. Petrides, Furnham, and Martin (2004) presented direct estimates of own and parental EI scores, noting significant gender differences, while Petrides, Niven, and Mouskounti (2006) showed that teacher ratings of ballet students’ trait EI converged with students’ trait EI scores (r = .58). In a behavioral genetic study, Vernon, Petrides, Bratko, and Schermer (2008) reported that the median father–offspring and mother–offspring correlations at the facet level of trait EI were .09 and .16, respectively, and that the correlations at the factor (r = .15 and r = .20) and global trait EI (r = .14 and r = .22) levels were somewhat stronger. However, measures of parent–offspring resemblance were calculated jointly for sons and daughters in that research. These studies, therefore, do not provide a complete view of the role of cross-informant ratings in trait EI, since they were conducted either with short forms which do not offer comprehensive coverage of the sampling domain of the construct, or without reference to the gender of the targets.

The use of cross-informant ratings has become an important aspect of research and clinical practice for those working with children and adolescents (Renk & Phares, 2004). Such ratings allow for a comparison of a target individual’s functioning across situations, potentially resulting in a less biased approach in the evaluation process (e.g., Renk & Phares, 2004). Given the relevance that trait EI has for problematic behaviors and the importance of cross-informant ratings especially in children and adolescent measurements, it seems vital to examine the relationship between self and parental trait EI ratings in an adolescent sample. This study will also examine how these ratings relate to self and parental ratings of internalizing and externalizing problems in adolescence.