Introduction

It can be asserted that from a traditional standpoint, psychology has been mainly focusing in those negative aspects of human behaviour, concentrating in individuals' suffering, personal distressors and psychological disorders (Salanova, & López-Zafra, 2014). This traditional paradigm, although still very present and necessary in the literature, has been partially offset by the emergence of an empirical research trend which envisages psychology from a positive outlook, which is considered as a complementary and needed new approach (^{Snyder & Lopez, 2002}).

Based on this premise, Peterson and Seligman (2004) proposed a classification of character strengths and virtues. Character strengths are positive traits that are reflected at a cognitive, emotional and behavioural level (Park, Peterson, & Seligman, 2004). It is important to note, however, that although they bear certain similarities to personality traits, character strengths are considered to be different due to the moral and cultural value placed on them (Peterson & Seligman, 2004). Character strengths are grouped in a total of six virtue categories, which constitute the core characteristics of character: wisdom, courage, humanity, justice, temperance, and transcendence. Each virtue comprises a number of strengths, up to a total of 24 strengths, such as love or kindness in the virtue of humanity, while the virtue of transcendence includes strengths such as hope or humor (^{Martínez-Martí & Ruch, 2014}).

Evidence shows that some virtues come more naturally to some people than others, but there is the general agreement that strengths and virtues must be practiced in order to become embedded as a natural response (^{Devettere, 2002}). In other words, virtues are proactive by nature, which means that they require active development, practice and habituation (^{Arjoon, 2008}). Although very few studies have specifically studied the relationship between emotional intelligence and virtues and/or character strenghts, there is an acknowledged role of emotions in the individuals' life and well-being, and also in the necessity of benefiting of positive moods (for example, ^{Lyubomirsky et al., 2005}). In this sense, it is suggested that emotional intelligence, with particular emphasis on emotion regulation and emotion repair, plays an important part in the development and maintenance of virtues and its character strenghts (^{Mirowska, 2011}). These results offer insights into how individual character strengths are related to emotional intelligence and especially to emotion regulation as well as emotion repair.

Differently, positive emotions in this field have been more explored, showing that positive emotions are related to several beneficial outcomes. ^{Fredrickson (2001}), for instance, developed a broaden-and-build theory which predicts that positive states enhance self-development, exploration and prosocial behaviours. According to his results, positive emotions also affect mindsets and broaden the range of cognitions and behavioural repertoires an individual engages in (Fredrickson & Losada, 2005). Additionally, it has been proven that positive emotional states convey certain information to the individual, including feedback regarding their overall state in life and progress towards goals (^{Lyubomirsky et al., 2005}).

In this direction, research so far has shown contradicting results, although several studies show that the endorsement of character strengths, such as hope, zest or love, is significantly related to a higher degree of well-being and Positive Affect (PA) (e.g., ^{Buschor et al., 2013}; ^{Martínez-Martí & Ruch, 2014}). At the same time, some authors have found that hope, zest, love, social intelligence, perseverance, and those strenghts related to the use and acquisition of knowledge seem to be the most related character strengths to PA while other authors find that all strengths result significantly and directly related to PA (^{Azañedo, 2014}; Martínez-Martí & Ruch, 2014). Contrary to that, modesty, appreciation of beauty and religiousness seem to be the least associated with PA (Martínez-Martí & Ruch, 2014).

Thus, given the current state of knowledge and the limited number of studies in the field, in the present work we aimed to study the association between Emotional Intelligence, Positive Affect and VIA's virtues and character strengths. Also, we had the objective of exploring if Emotional Intelligence, measured with the scales of Emotional Attention, Emotional Clarity and Emotional Repair, predicted the virtue's levels of the subjects. Finally, the mediating role of PA in the aforementioned relationship was also studied.

Method

Results

Table 1 shows the means, standard deviations and the range of 24 VIA strengths and 6 VIA virtues, the 3 scales of TMMS-24 and the PA scale of PANAS for the total sample.

Table 1: Descriptive statistics of the study variables. 

Pearson correlations among measures are shown in Table 2. As expected, our results yield a significant positive correlation between all the VIA strengths and virtues and TMMS-24 scales Emotional Clarity and Emotion Repair. Differently, Emotional Attention showed more selective correlations, specifically, positive and significant correlations emerged for all the virtues except Temperance. Character strenghts also showed some significant associations, being Humanity the virtue with more significative character strenghts (see Table 2).

Table 2: Correlations among measures. 

When further analysing Positive Affect, results showed that all the character strenghts and virtues of VIA correlated significantly, with values ranging from .200 (Prudence) to .400 (Zest). Regarding TMMS-24, Emotional Clarity and Emotional Repair also showed a significant positive correlation, with .233 and .268 values respectively.

Additionally, in order to study the role of PA in the relationship between emotional intelligence and virtues, we carried out a hierarchical regression model for each one of the virtues and we examied the potential mediating contribution that PA may have in the model. The first step of the model introduces TMMS-24 scales as possible predictors and one of the virtues. The second step introduces, in addition to the TMMS-24 scales, the PA scale. Those predictors which change their significance within the model when PA is added, may suggest a mediation effect of the latest variable.

All the models that were tested showed that TMMS-24 variables contribute to the model when predicting for the virtues, showing high significance and a relevant difference among R-squared when introducing PA. The increase among R-squared indexes of the analysed models ranges from .202 to .258 for Wisdom, from .189 to .266 for Courage, from .161 to .213 for Humanity, from .127 to .177 for Justice, from .074 to .131 for Temperance and from .185 to .248 for Transcendence.

Although the overall total variation of the model was not always high, it is important to note that all the entered variables resulted as highly significant contributors for the model. Moreover, in an attempt to go a step further and so as to explore the mediating role of PA, the changes in significance for all the independent variables were studied in further detail. Results showed that when predicting Temperance and Humanity, Emotional Repair shows a possible mediating effect of PA (see Tables 3 and 4). That is, when introducing PA in the model, Emotion Repair becomes non-significant, while the explained variability of the model increases up to a 21.3% for Humanity and a 13.1% for Temperance.

Table 3: Hierarchical multiple regression analysis and mediation for the scales of TMMS-24, Positive Affect scale of PANAS predicting Temperance virtue of VIA. 

Discussion and Conclusions

The present study aimed at analysing the relationship between Emotional Intelligence, Positive Affect and VIA's virtues and character strengths. We also intended to explore if Emotional Intelligence predicted the virtue's levels of the participants, as well as the possible mediating role of Positive Affect.

Despite the scarcity of studies regarding Emotional Intelligence and character, our results are in line with those previous ones showing that character is associated with emotional abilities and Positive Affect, becoming evident that some strengths and virtues consistently seem to yield higher correlations than others (^{Lyubomirsky et al., 2005}; ^{Martínez-Martí & Ruch, 2014}). Specifically, our results showed a significant positive correlation between all the VIA strengths and virtues and TMMS-24 scales Emotional Clarity and Emotion Repair. However, Emotional Attention showed more selective correlations, emerging positive and significant correlations for all the virtues except Temperance. This finding appeared as consistent with previous literature that exhibited more association with emotion regulation than other emotional skills. This can be explained by the fact that a less functional regulation of the emotions has a greater and a more stable impact in character and affect than other skills, more related to the awareness and identification of the emotion itself.

When exploring if Emotional Intelligence predicted the virtue's levels of the participants and the possible mediating role of Positive Affect, our results showed that all TMMS-24 scales predicted virtue's levels in a discreet but significant way. Additionally, when predicting the virtues of Temperance and Humanity, it appeared a possible mediating role of Positive Affect. In the first hierarchical step of the regression, Emotional Repair scale of TMMS-24 was significant, but when introducing Positive Affect in the second step of the regression, this scale ceased its significance. This can be explained because Emotional Repair only leads to high levels of Humanity and Temperance when Positive Affect emerges. That is, an individual exhibits high levels of Humanity and Temperance by means of having a positive affection. This is consistent with previous evidence, which shows that regulating and repairing emotions and, ultimately, being emotionaly efficient, leads to a higher satisfaction, Positive Affect and a healthier mindset, which at the same time, brings individuals to a better psychological adjustment and character features (^{Fernández-Berrocal, P., Ramos, N., & Extremera, N.; 2001}; ^{Nolen-Hoeksema, S., & Aldao, A., 2011}).

It is important to note that until now, most studies have focused on samples whose representativeness becomes limited. Studied participants are normally individuals from convenience samples that typically result in biased samples. Also, there is evidence showing that there may be differences among the relationship existing among character and affect and well-being (Martí & Ruch, 2014). Our results were obtained form a sample of undergraduate students, which show a certain amount of homogeneity regarding their characteristics and age. Therefore, we believe that future studies shoud be carried out with samples which amend this pitfalls.