Investigating the mediating effects of emotional intelligence and coping on problem behaviours in adolescents

Abstract

This study explored the mediating effect of emotional intelligence (EI) and coping strategies on problem behaviours in Australian adolescents. One hundred and forty-five adolescents (60 boys and 85 girls with a mean age of 12.02 years) completed self-report instruments of EI, stress coping strategies, and problem behaviours. The relationships between Emotional Management and Control and engagement in internalising and externalising behaviours were found to be mediated by the use of non-productive coping strategies. Mediation models of the relationship between problem behaviours and the Understanding Emotions and Emotional Recognition and Expression dimensions were found to be only partially mediated by the engagement in problem-focused and non-productive coping strategies. The results are discussed in regards to how coping strategies utilised in adolescence may produce more or less adaptive patterns of coping during adulthood. The development of emotional abilities may be required to improve coping outcomes for adolescents, which in turn may produce better psychological outcomes for adolescents in the long term.

Keywords:

• Adolescence
• coping
• EI
• problem behaviour
• stress

Early adolescence is a time of transition during which the individual experiences not only profound physical, emotional and cognitive changes, but also changes in terms of societal expectations regarding appropriate behaviour patterns and preparation for taking on adult roles (Steinberg & Morris, 2001). This transition from childhood to adolescence can be a challenging time for many young people. While many adolescents navigate this period without experiencing major difficulties, some individuals may experience psychological and behavioural problems (Steinberg & Morris, 2001). As many as 14% of Australian children under the age of 15 have some form of mental or behavioural problem (Australian Bureau of Statistics, 1997), and such problems may have flow-on negative consequences into adulthood. In light of this, the establishment of methods to identify individuals who are at risk for developing problem behaviours and strategies for the amelioration of these behaviours is of critical importance. It has been proposed that the recently emerging concept of emotional intelligence (EI) may offer a measure that has predictive value for identifying those at greatest risk for problem behaviour (Petrides, Frederickson, & Furnham, 2004). It has been further demonstrated that EI is amenable to development through positive intervention strategies (Hansen, Gardner, & Stough, 2007), and might therefore provide a route to diminishing the likelihood of the emergence of problem behaviours.

Research concerning adolescent behaviour has detailed the broad and complex interactions of individual temperament, parenting and environment that contribute to the development of problem behaviours. Studies have explored the relationship between attachment (Allen, Moore, Kuperminc, & Bell, 1998; Gerard & Buehler, 2004), family systems (Hughes & Gullone, 2008), parenting (Zimmer-Gembeck & Locke, 2007), temperament (Austin, Hadzi-Pavlovic, Leader, Saint, & Parker, 2005), resilience, stressful life events, socioeconomic status, ethnicity (Finn & Rock, 1997), self-concept (Ybrant, 2008) and identity consolidation (Besser & Blatt, 2007) to identify the determinants of problem behaviour in adolescents. Research suggests that the origins of problem behaviour may be complex and multifaceted, but this does not presuppose that the identification of antecedents in individual cases are a necessary prerequisite for amelioration.

In terms of defining problem behaviours, they have been broadly categorised into two classes: internalising and externalising behaviours (Achenbach & Rescorla, 2001). Internalising behaviours and disorders are characterised by primary disturbance in mood and emotion, and include depression, anxiety, somatic complaints and withdrawn behaviour (Hughes & Gullone, 2008). In contrast, externalising problems are characterised primarily by disturbances in the regulation of behaviour and acting-out behaviours such as rule breaking, aggression and delinquency (Hughes & Gullone, 2008). Internalising and externalising behaviours occur on a continuum that range from typical or low-level symptoms that change over time, to more severe presentations of symptoms that cause significant impairment and meet the criteria for diagnosable disorders including anxiety and major depression (Broberg et al., 2001; Reitz, Dekovic, & Meijer, 2005). In terms of gender distribution, it is widely accepted that male subjects display more externalising problem behaviours and female subjects display more internalising behaviours (Besser & Blatt, 2007; Broberg et al., 2001; Ciarrochi, Chan, & Bajgar, 2001; Santesso, Reker, Schmidt, & Segalowitz, 2006). These broad categories of behaviour are not mutually exclusive, because an individual may express both internalising and externalising behaviours (Broberg et al., 2001; Reitz et al., 2005).

As foreshadowed, assessment of adolescent EI may be able to offer a window to predicting, understanding, and attenuating problem behaviours (Petrides et al., 2004). Levels of EI have previously been found to be associated with subjective wellbeing (e.g. Austin, Saklofske, & Mastoras, 2010; Schutte et al., 2010), and conversely, levels of EI have been successfully used to predict psychological maladjustment. For instance, recent research exploring the relationship between EI and psychological adjustment found that EI uniquely predicted levels of anxiety and depression (Downey et al., 2008), and has been found to be associated with better psychological adjustment independent of the effects of self-esteem or thought suppression (Fernandez-Berrocal, Alcaide, Extremera, & Pizzarro, 2006). EI has been conceptualised as a set of abilities concerned with the regulation, management, control and use of emotions in decision-making (Mayer, Caruso, & Salovey, 2000), particularly in relation to the promotion of healthy and adaptive mental functioning. Several models and measures of EI have been proposed in recent years, with the measures generally falling within one of two conceptions of the construct: ability or trait. In a factor analytic study using the Swinburne University Emotional Intelligence Test (SUEIT) (Palmer & Stough, 2001), a modified version of the SUEIT (Palmer & Stough, 2001) was administered to 1002 Australian adolescents to assess how they generally (a “trait” measure) feel, think and act with emotions and on the basis of emotional information. Luebbers, Downey, and Stough (2007) identified four factors of EI: Emotional Recognition and Expression (ERE; the ability to identify one's own feelings and emotional states and the ability to express those inner feelings to others); Understanding Emotions (UE; the ability to identify and understand the emotions of others); Emotions Direct Cognition (EDC; the extent to which emotions and emotional knowledge are incorporated in decision-making and/or problem solving); and Emotional Management and Control (EMC; the ability to manage positive and negative emotions both within oneself and others and control strong emotional states). Both the ability and trait measures of EI have been shown to have predictive validity in recent research (Matthews, Zeidner, & Roberts, 2002). Other authors have adapted their adult EI measures for use with adolescents, and have linked ability-based adolescent EI with adolescent health risk factors (Trinidad & Johnson, 2002), and trait-based EI with altruism (Charbonneau & Nicol, 2002), mood management behaviour, amount and satisfaction with social support (Ciarrochi et al., 2001), and scholastic achievement (Downey, Mountstephen, Lloyd, Hansen, & Stough, 2007; Hogan et al., 2010). One previous study has examined an aspect of problem behaviour in relation to EI. That study examined relationships between EI, academic performance and truancy and exclusions in 650 English secondary school students (Petrides et al., 2004). Students with low trait EI were significantly more likely to have been expelled from school. The authors suggested that students with poor social and emotional skills were more likely to feel withdrawn and excluded, increasing the likelihood that they will behave in anti-conventional ways (Petrides et al., 2004).

When examining EI as a predictor of problem behaviour in adolescence, another concept of significant interest is that of coping strategies. Coping refers to a variety of cognitive and behavioural strategies that individuals use to manage their stress (Lazarus & Folkman, 1984). Research suggests that a lack of adaptive coping techniques is predictive of problem behaviour in adolescents (Bartek, Krebs, & Taylor, 1993). For example, Compas, Connor-Smith, Saltzman, Harding-Thomson, and Wadsworth (2001) investigated the relationship between coping and measures of psychological adjustment such as internalising behaviour, externalising behaviour, social competence and academic competence. The authors examined a number of categories of coping including problem-focused coping (problem solving, information seeking, and problem-focused support), emotion-focused coping (emotional expression, denial, wishful thinking), engagement coping (problem solving, emotional expression, support seeking), and disengagement coping (problem and cognitive avoidance, social withdrawal). Overall the findings indicated that engagement and problem-focused coping were associated with better psychological adjustment while disengagement coping and emotion-focused coping were associated with poorer psychological adjustment. Thus, adolescents who deal with stress rather than avoid it are likely to experience better psychological adjustment (Compas et al., 2001). Of interest is the extent to which EI facilitates the development of adaptive coping strategies.

Various authors have proposed theoretical links between EI and coping. For example, Taylor (2001) argued that the emotionally intelligent can cope better with life's challenges and control their emotions more effectively, both of which contribute to good psychological and physical health. Matthews and Zeidner (2000) claimed that “adaptive coping might be conceptualised as EI in action, supporting mastery of emotions, emotional growth, and both cognitive and emotional differentiation, allowing us to evolve in an ever-changing world” (p. 460). Given these theoretical links, associations between EI and adaptive coping styles have been observed in studies of adult populations (Bastian, Burns, & Nettelbeck, 2005; Goldenberg, Matheson, & Mantler, 2006). Bar-On (1997) also found positive associations between EI and task-focused coping and, further to this, negative associations between EI and emotion-focused coping. Austin et al. (2010) reported that emotion-focussed coping was negatively associated with the Intrapersonal, General Mood and Stress Management EI subscales, while task-focused coping was positively associated with Adaptability, and social diversion coping with Interpersonal EI in a sample of university students. The notion that deficits in EI skills (i.e., the ability to perceive, express and regulate emotions) interfere with the development of complex coping skills requires further investigation.

There is currently little information available on the associations between EI, coping and problem behaviour in adolescents. Levels of EI have previously been linked to the use of more or less adaptive coping strategies (Saklofske, Austin, Galloway, & Davidson, 2007). Moreover a lack of adaptive coping techniques is predictive of problem behaviour in adolescents (Bartek et al., 1993). Lower levels of EI have also been linked to adolescents engaging in problem behaviours (Petrides et al., 2004). As such, EI assessment tools may offer a predictive measure of adolescents' use of maladaptive coping strategies, which in turn may contribute to the emergence of problem behaviours. This hypothesis will be investigated through modelling of these interactions through the mediation analysis as described by Baron and Kenny (1986). Where a priori assumptions of logical priority can be justifiably made, this type of analysis can be used to model the interaction of predictive variables in explaining the outcome measure.

Method

Participants

Participants in this study were 145 students aged between 11 and 13 from a coeducational public school in Melbourne, Australia. The sample consisted of 60 boys and 85 girls with a mean age of 12.02 years. Eighty-nine percent of students identified their background as “Oceanian” (Australia, New Zealand, Melanesian, Micronesian and Polynesian), 5% identified as European, 1% Southern and Central Asian and 1% as North African or middle eastern. Four percent of students failed to identify their background.

Procedure

Written consent was obtained from all participants prior to participation. Because respondents were under the age of 18, written consent was also obtained from a parent or legal guardian. All participants completed the questionnaire booklet in class time under the supervision of their teacher. The questionnaire booklet consisted of self-report measures of EI, problem behaviour and coping strategies and demographic questions related to age, gender and ethnicity. The self-report measures used in the present study are detailed below. This study was approved by the Swinburne University Human Research Ethics Committee.

Measures

The Adolescent SUEIT (Luebbers et al., 2007) was used to assess adolescent's self-perceptions about how effectively they deal with emotions and emotional information. The 57-item self-report questionnaire consists of four subscales: ERE, UE, EDC and EMC. ERE measures the ability to identify and express one's feelings to others (“I find it hard to say how I feel”). UE measures the ability to identify and understand the emotions of others (“I can tell when other people are trying to hide their true feelings”). EDC measures the extent to which emotions and emotional knowledge are incorporated in decision-making and/or problem solving (“I solve my problems using logic and clear thinking instead of feelings”). EMC measures the ability to manage positive and negative emotions both within oneself and others and control strong emotional states (“I find it hard to make others excited about things”). Participants are asked to indicate the accuracy of each statement in regards to the way they typically feel, think and act. Participants respond to statements on a 5-point Likert scale, ranging from 1 (seldom) to 5 (very often). Reliability for total EI and each subscale has been shown to be high (ERE, α = .75; UE, α = .81; EDC, α = .75; EMC, α = .75; and Total EI, α = .85) (Luebbers et al., 2007).

Adolescent problem behaviour was assessed using the Youth Self-Report checklist (YSR) (Achenbach & Rescorla, 2001). The YSR is a widely used standardised instrument for measuring a broad range of self-reported emotional and behavioural problems in adolescents aged 11–18 years. The 112 items measure problem behaviours corresponding to eight subscales: Withdrawn/Depressed, Somatic Complaints, Anxious/Depressed, Social Problems, Thought Problems, Attention Problems, Rule Breaking behaviours, Aggressive behaviours. These subscales can also be scored into two broader dimensions: Internalising behaviours (Withdrawn, Somatic, Anxious/Depressed) and Externalising behaviours (Rule Breaking behaviours, Aggressive behaviours). Participants rate how well each statement (i.e., “I argue a lot”, “I am mean to others”, “I destroy my own things”) describes their behaviour on a 3-point Likert scale (0 = not true; 1 = somewhat or sometimes true; 2 = very true or often true). Reliability for each subscale has been shown to be high (Anxious/Depressed, α = .84; Withdrawn/Depressed, α = .71; Somatic complaints, α = .80; Social problems, α = .74; Thought problems, α = .78; Attention problems, α = .79; Rule-breaking behaviour, α = .81; Aggressive behaviour, α = .86; Internalising, α = .90; Externalising, α = .95).

Adolescent coping styles were assessed using the Adolescent Coping Scale (ACS) (Frydenberg & Lewis, 1990). The 18-item ACS assesses coping strategies that can be grouped into three distinctive coping styles: Solving the problem, Reference to others; and Non-productive coping. “Solving the problem” refers to strategies that remove the problem through personal endeavour with minimal use of others. “Reference to others” refers to coping strategies that draw on the support of others as a resource with a problem-focused orientation. “Non-productive coping” refers to coping strategies that do not lead to a solution but may reduce tension. Participants are asked to respond to statements that describe each coping style (i.e., “Wish a miracle would happen”, “See myself as being at fault”, “Ask for professional help”), indicating how often they would use the behaviour on a 5-point Likert-type scale ranging from 1 (doesn't apply or don't do it) to 5 (used a great deal). Developed with Australian samples, the ACS has demonstrated satisfactory test−retest reliability and high internal consistency with reliability coefficients ranging between .76 and .89 (Frydenberg & Lewis, 1990).

Statistical procedure

To assess the possible mediating effect of coping style and EI on problem behaviours, the technique outline by Baron and Kenny (1986) was utilised. This technique requires that the initial variable, mediator variable, and the outcome variable are all significantly correlated. Further to this, the initial variable and mediator variable need to significantly predict the outcome variable in separate regression equations. The final step involves the assessment of whether the proposed mediator mediates, or partially mediates the relationship between the initial variable and outcome variable.

Results

The means and standard deviations of the Adolescent SUEIT, ACS, and YSR are presented in Table I. The mean scores for the Adolescent SUEIT appear slightly lower than previously reported averages (Downey, Mountstephen, et al., 2007), this may be a consequence of the lower age range sampled in this study. This result is consistent with the suggestion that the abilities assessed by EI measures develop across the lifespan (Luebbers et al., 2007). Significant differences between boys and girls were observed, with girls scoring higher on the ERE, UE and EDC subscales similar to the findings of Luebbers et al. (2007). The means for both the ACS and YSR appeared in the normal range for the sample studied, with male and female adolescents reporting similar use of coping strategies in line with previous findings (Frydenberg & Lewis, 1990). Girls did, however, report greater levels of anxiety, somatic, and internalising behaviours; one-way ANOVAs reflecting these comparisons appear in Table I.

Table I. Variables included in the study (N = 142–144)

	M	SD	F
Emotional Recognition and  Expression (α = .81)	30.93	4.82	4.97*
Boys	29.88	4.40
Girls	31.67	4.99
Understanding of Emotion (α = .78)	64.90	7.10	7.42**
Boys	63.03	6.49
Girls	66.22	7.25
Emotions Direct Cognition (α = .69)	31.62	4.58	9.71**
Boys	30.25	4.28
Girls	32.59	4.57
Emotional Management  and Control (α = .82)	53.89	7.76	0.36
Boys	54.35	7.71
Girls	53.56	7.82
Social Problems (α = .72)	4.00	3.22	0.41
Boys	3.80	2.70
Girls	4.14	3.55
Aggressive Behaviour (α = .79)	6.72	4.96	0.06
Boys	6.59	4.51
Girls	6.80	5.28
Thought Problems (α = .75)	5.29	3.95	0.43
Boys	5.03	3.94
Girls	5.47	3.97
Attention Problems (α = .75)	5.21	3.36	0.14
Boys	5.33	3.18
Girls	5.12	3.50
Anxious/Depressed (α = .78)	5.38	4.01	6.71*
Boys	4.37	3.24
Girls	6.08	4.36
Withdrawn/Depressed (α = .81)	3.67	2.66	0.84
Boys	3.62	2.42
Girls	3.71	2.83
Rule-breaking Behaviour (α = .85)	3.07	2.98	2.46
Boys	3.53	2.61
Girls	2.75	3.19
Somatic Behaviour (α = .76)	3.83	3.16	4.90*
Boys	3.15	2.64
Girls	4.32	3.42
Internalising Behaviours (α = .84)	12.88	8.48	4.43*
Boys	11.13	6.98
Girls	14.11	9.23
Externalising Behaviours (α = .83)	9.79	7.41	0.21
Boys	10.12	6.49
Girls	9.55	8.02
Non-productive Coping (α = .81)	20.81	5.64	0.90
Boys	21.34	5.99
Girls	20.43	5.38
Problem-solving Coping (α = .83)	22.15	3.74	1.78
Boys	21.65	3.49
Girls	22.50	3.89
Refer to Others (α = .66)	9.59	2.86	0.01
Boys	9.56	3.32
Girls	9.61	2.50
Notes: N = 145, N = 143 for coping variables.
*p < .05, **p < .01.

Correlational analysis

Following the directions of Baron and Kenny (1986), cross-correlations were calculated between the EI, stress coping and problem behaviour measures to identify any significant overlap between the initial (EI), mediator (coping) and outcome measures (internalising or externalising behaviours). Table II contains the correlational matrix, whereby the significant overlap between the problem behaviour scales occurred with the ERE, UE, and EMC dimensions of the Adolescent SUEIT in regards to the initial variables and with problem-focused coping and non-productive coping as the proposed mediators.

Table II. Correlations for variables included in the study (N = 140–145)

	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17
Emotional Recognition	1
Understanding Emotions	.295**	1
Emotions Direct Cognition	.143	.230**	1
Emotional Management and Control	.411**	.217**	.146	1
Problem-Solving Coping	.251**	.314**	.059	.324**	1
Reference to Others	.035	−.117	−.161	−.106	.081	1
Non-productive Coping	−.234**	−.083	−.220**	−.343**	−.056	.395**	1
Social Problems	−.121	−.170*	−.074	−.315**	−.361**	.219**	.452**	1
Aggressive Behaviour	−.140	−.168*	−.149	−.343**	−.266**	.185*	.418**	.623**	1
Thought Problems	−.215**	−.063	−.088	−.342**	−.167*	.388**	.532**	.627**	.627**	1
Attention Problems	−.235**	−.209*	−.230**	−.459**	−.264**	.158	.506**	.558**	.623**	.492**	1
Anxious/Depressed	−.205*	−.039	−.075	−.476**	−.255**	.198*	.499**	.710**	.480**	.622**	.502**	1
Withdrawn/Depressed	−.161	−.112	−.117	−.325**	−.331**	.056	.361**	.657**	.513**	.606**	.491**	.682**	1
Rule-breaking Behaviour	−.195*	−.161	−.127	−.140	−.137	.184*	.373**	.571**	.725**	.602**	.404**	.333**	.408**	1
Somatic Behaviour	−.124	−.054	−.077	−.237**	−.160	.225**	.361**	.611**	.542**	.689**	.447**	.596**	.540**	.506**	1
Internalising Behaviour	−.194*	−.074	−.101	−.416**	−.284**	.196*	.484**	.770**	.590**	.741**	.558**	.909**	.837**	.474**	.824**	1
Externalising Behaviour	−.172*	−.177*	−.151	−.286**	−.234**	.198*	.430**	.646**	.961**	.661**	.579**	.455**	.508**	.888**	.566**	.585**	1
Notes: N = 145, N = 143 for coping variables.
1 = Emotional Recognition and Expression; 2 = Understanding Emotions; 3 = Emotions Direct Cognition; 4 = Emotional Management and Control; 5 = Problem-Solving Coping; 6 = Reference to Others; 7 = Non-productive Coping; 8 = Social Problems; 9 = Aggressive Behaviour; 10 = Thought Problems; 11 = Attention Problems; 12 = Anxious/Depressed; 13 = Withdrawn/Depressed; 14 = Rule-breaking Behaviour; 15 = Somatic Complaints; 16 = Internalising Behaviour; 17 = Externalising Behaviour.*p < .05, **p < .01.

Regression analyses were then conducted to confirm that the proposed initial variables and mediators independently predicted significant proportions of variance of the problem behaviours. ERE significantly predicted 6% of problem-focused coping, and 5.5% of non-productive coping; UE dimension predicted 9.9% of problem-focused coping; and EMC significantly predicted 11.5% of the variance associated with non-productive coping and 10.5% of problem-focused coping, satisfying the requirement to act as the initial variable in each case. In regards to proposed mediator variables predicting the problem behaviours, they predicted internalising (problem-focused, 8.1%: non-productive coping, 8.2%), and externalising (problem-focused, 5.5%: non-productive coping, 18.5%) behaviours significantly.

Prediction of the outcome variables by the initial variables was also calculated, with ERE accounting for 3% of externalising and 3.8% of internalising behaviour scores. UE predicted 3.1% of externalising behaviours and EMC predicted 8.2% of externalising and 17.3% of internalising behaviours.

Mediation analyses

The mediation analyses were conducted to assess the mediating relationship of problem-focused and non-productive coping when added to a regression model with ERE, UE and EMC as the initial variables and the problem behaviours as the outcome variable, and the Sobel test (Sobel, 1982) was utilised to assess if the indirect effect was significant. For internalising behaviours, the regression models involving EMC (z = 3.46, p < .001) were significantly reduced with the addition of non-productive coping, confirming that it mediated the relationship. For externalising behaviours, the relationship with EMC (z = 3.63, p < .001) was significantly mediated by the addition of non-productive coping. A representation of these relationships is presented in Figure 1. The remaining mediation analyses indicated that problem-focused and non-productive coping only partially mediated the relationship between the EI dimensions and the problem behaviours assessed (Figure 1). The strength of the relationships were all reduced, but not to the point where the regression models became non-significant. Further to this, the Sobel's tests did not indicate that a significant proportion of variance was carried through the mediators in these models.

Figure 1. Mediation models of the effects of emotional intelligence and coping on problem behaviours.

Discussion

Unsurprisingly, given the empirical and theoretical links between the recognition, expression, management and control of emotions, adaptive and maladaptive coping strategies, and internalising and externalising problem behaviours in adolescents, a large number of significant correlations between variables was observed. An exhaustive enumeration of these would not in and of itself be particularly informative. It is in cases such as this that a more structured analytic approach is most valuable. This discussion will therefore restrict itself to the specific relationships hypothesised in relation to the mediating effects of coping and EI on problem behaviours. Assessment of the mediating effects of EI and stress coping strategies on problem behaviours in an Australian adolescent population showed that adolescents' ability to manage positive and negative emotions and emotional states was predictive of adolescents' use of non-productive coping strategies and engagement in internalising behaviours. This relationship was investigated using mediation analysis, whereby the relationship between EMC and internalising behaviours was fully mediated by the use of non-productive coping strategies (Figure 1). This finding suggests that the ability to manage and control emotions was predictive of the use of maladaptive coping strategies, which in turn predicted adolescents' engagement in internalising problem behaviours. This follows from the Taylor (2001) suggestion that greater levels of EI may enable individuals to cope better with life's challenges by controlling emotions more effectively, and therefore contributing to higher psychological and physical health. In the current study adolescents who reported being able to manage strong emotions were less likely to internalise their emotional experience through anxiety, depression, withdrawing and excluding themselves from social engagement and were less likely to report chronic physical complaints without a known cause. This result is somewhat similar to that of the Petrides et al. (2004) finding that their trait EI measure predicted truancy and exclusions (both forms of withdrawing and exclusion) in a sample of English adolescents.

In regards to the mediation analyses concerning externalising behaviours, the relationship between EMC and externalising behaviours was also significantly mediated by the use of non-productive coping strategies. Adolescents higher in the ability to manage and control emotions used less non-productive coping strategies and as a consequence were less likely to engage in rule-breaking and aggressive behaviours. This result suggests that individuals with poor social and emotional skills are more likely to ineffectively deal with stressful situations, which may increasingly lead them to acting out in an aggressive manner, whether by destroying things, arguing with parents, friends, and peers, or by breaking rules. Overall, the significant mediation analyses suggest that adolescents with greater ability to manage and control emotions were less likely to use non-productive coping to deal with stressful situations common during adolescence, and in turn internalise or externalise these experiences via antisocial or problem behaviours.

While both higher levels of the ability to manage and control emotions and lower levels of the use of non-productive coping strategies predict unique variance in problem behaviours, the relationship described herein between these two constructs offers a more sophisticated understanding of how these variables predict problem behaviours in adolescence. It is, however, important to acknowledge the interrelation of the EI subscales assessed, because the more basic abilities of EI, the experience, expression, and understanding of emotions are prerequisites for the engagement of more complex abilities, such as the management of one's own and others emotions. This interrelation is reflected in the correlational overlap in this study and previous studies (Luebbers et al., 2007) between both the ERE and UE with the EMC dimension of the Adolescent SUEIT. These dimensions also predicted a significant proportion of problem behaviours and use of coping styles, but this did not translate into significant mediation models. Nonetheless, these results point to the utility of identifying EI subscales or subfactors that individually may predict important life outcomes in lieu of using composite Total EI scores, allowing for more meaningful interpretation. The examination of the predictive validity of subscales derived from different models and measures of EI (Gardner & Qualter, 2010) and of group differences in EI subscale profiles (Sánchez-Ruiz, Pérez-González, & Petrides, 2010) are also areas that could be further investigated in future work.

Previous research suggests that adolescent problem behaviours increase in middle to late adolescence (Steinberg & Morris, 2001) and that EI also increases as a function of cognitive maturation (Mayer et al., 2000) from an “ability” point of view, therefore replication of this study with a broader sample age range to include older adolescents and use of an ability-based measure, may further validate these findings and potentially elucidate stronger relationships between the dimensions. Comparison to an at-risk population of adolescents and a general population sample may also contribute further to the understanding of the relationship between EI, coping and problem behaviours. Problem behaviours occur on a continuum of behaviour ranging from typical or low-level symptoms that change over time, to more severe presentations of symptoms that cause significant impairment and meet the criteria for diagnosable disorders, such as anxiety disorder and major depression (Broberg et al., 2001; Reitz et al., 2005).

The current study assessed the mediation effects of coping behaviour on the relationship between EI and problem behaviours. Adolescents who reported greater abilities in managing emotions, were more likely to effectively deal with stressful situations common during adolescence, and in turn were less likely to internalise or externalise these experiences via antisocial or problem behaviours. While individuals may grow out of certain problem behaviours, maladaptive coping strategies learned in adolescence may inhibit one's sense of personal efficacy or confidence in dealing with stress, and therefore limit the development of more adaptive coping.

The way in which individuals cope with stress in adolescence may influence a more or less adaptive developmental trajectory and establish patterns of coping that may continue into adulthood (Compas et al., 2001). This is an important consideration in the development of emotional coping skills, where the development of emotional abilities may be required to improve coping outcomes for adolescents, which in turn are predictive of better psychological adjustment. Further research into the predictive validity of EI in adolescent behaviour could also lead to the development of EI-focused interventions to address social and emotional deficits within adolescents that may limit development of adaptive stress coping strategies and ameliorate problem behaviours.