Analyzing self-concept, emotional intelligence and violence according to sport modality in Higher Education

José Luis Ubago-Jiménez

Félix Zurita-Ortega

Eduardo Melguizo-Ibáñez

José Manuel Alonso-Vargas

Abstract

Violence in the university context presents a serious concern for higher education, due to the increasing number of violent episodes among young people. The sport modality played together with psychosocial factors such as self-concept or emotional intelligence effectively counteract these disruptive behaviours. The aim of this study is to test the relationship between the type of sport played, self-concept and emotional intelligence with violence in university students. The study was conducted with a sample of 1,057 Spanish university students among whom 43.8% (n = 463) were female and 56.2% (n = 594) were male. The TMMS-24 test was used to measure emotional intelligence and the SC-5 scale to measure self-concept. Those who do not play sport were found to have higher rates of direct and indirect violence. Likewise, young people involved in contact sports have the highest rates of direct violence. It is necessary to promote the practice of physical activity related to the development of emotional intelligence in order to reduce violent behaviour.

Introduction 

Aggressiveness is a construct that has been widely examined in the field of psychology, and is defined as the performance of an action ranging from threatening gestures to actual attacks against another (Giammanco et al., 2005). Its manifestation into aggressive behaviour depends on multiple factors internal and external to the person (Haller, 2020). In this line, violence arises as a consequence of the individual’s inability to control and manage his or her own impulses, turning him or her into a hostile and socially maladjusted subject in various environments. From a narrow perspective, violence implies the use of force with the intention of harming another person, but reducing violence to the physical act is a reductionist attempt to focus only on what can be measured (Martínez-Pacheco, 2016). 

On the contrary, it is appropriate to consider violence from a relational perspective, where it appears as a consequence of the mechanisms of power that are produced in the interplay of social relations, in any context. There are many triggers that can lead to violent behaviour, such as the need for social recognition, anger or the need to maintain a certain status regardless of the consequences, among others. To mitigate this situation, it is important that the individual learns a repertoire of coping strategies that help to regulate and alleviate such impulses (García-Martínez et al., 2021).

Psychosocial factors have traditionally been related to people’s overall well-being. Among them, emotional intelligence (EI), together with self-concept, have been identified as two reliable factors in determining the physical and psychological health of university students. In this respect, EI has been conceptualised using various models, most notably the ability model (Salovey & Mayer, 1990) and mixed models (Pérez-González et al., 2020). In this sense, EI can be defined as a person’s ability to identify, express and regulate his or her emotions and those of others, with a view to building stable bonds with the environment around him or her. 

On the other hand, self-concept is associated with the person’s perception of him/herself in different environments, ranging from emotional and physical aspects to performance in academic, social and family environments. The study of both constructs contributes to the development and construction of a profile of the individual that enables one to predict the individual’s behaviours to a certain extent. When examined from the field of physical activity and sport, it is observed that sport practice has also been identified as a factor that influences health, as a practice that is effective for channelling emotions and adverse feelings that a person may feel and maintaining balance. 

Accordingly, distinctions are made depending on the type of sport played, with a distinction being made between individual and group sports. Another common criterion usually applied is the existence or non-existence of contact between athletes, with a distinction being made between contact and non-contact modalities. From the above criteria, the four categories to be considered in this study are: individual non-contact, individual contact, group non-contact and group contact.

In view of the above, the aim of this study is to examine the levels of aggressiveness, self-concept and emotional intelligence of university students according to the type of sport they play by means of an explanatory structural equation model.

Materials and methods

Design and sample

With regard to the study design, a cross-sectional, descriptive and non-experimental analysis was performed. It was carried out with a sample of 1,057 university students from Andalusian public universities selected using convenience sampling. Regarding the gender distribution of the sample, 43.8% (n = 463) were female and 56.2% (n = 594) were male, aged between 18 and 23 years (20.78 ± 2.85). 

Resources 

Four tools were used to collect data for this research. The ad-hoc questionnaire collected the socio-demographic data of the sample, such as gender, age and type of sport played.

The Trait Meta-Mood Scale (TMMS-24) developed by Salovey et al. (1995) was used to measure EI, but in this study the version translated into Spanish by Fernández-Berrocal et al. (2004) was used. This questionnaire assesses emotional intelligence as a trifactorial construct using 24 items, where the dimensions emotional attention, emotional clarity and emotional repair are scored through a summation. In this study, a Cronbach’s alpha of α = .910 was obtained for the overall scale.

With regard to the measurement of self-concept, the Self-Concept Form 5- SC5 (García & Musitu, 2014) was used. This scale is composed of 30 items and is divided into five categories: academic, social, emotional, family and physical. The questionnaire is completed on a 5-choice Likert-type scale, where 1 is “never” and 5 is “always”. The internal consistency of the scale was α = .820 in this research.

The Scale of Violent Behaviour at School, originally developed by Little et al. (2003) and adapted to Spanish by Estévez (2005), was used to measure violent behaviour. This questionnaire is divided into two categories: direct aggression is generated in a personal encounter between the aggressor and the victim; and indirect aggression, which occurs when the aggressor remains anonymous. The Likert-type scale is completed by responding to 25 items ranging from 1 “never” to 4 “always”. In this research, an internal consistency of α = .880 was obtained. The scale has been used in studies on adolescents, such as that of San-Román-Mata et al. (2019) or Zurita-Ortega et al. (2018) and in primary school students (Rojas-Jiménez & Castro-Sánchez, 2020; Sánchez-Zafra et al., 2018).

Procedure  

The research process was divided into three phases. In the first, permission was requested from the University of Granada (Spain), and was granted by the corresponding ethics committee (2342/CEIH/2021). In the second phase, an informative letter was drafted explaining the objectives of the study and requesting informed consent from the participants. Following the agreed participation of 1,234 students, the questionnaire was sent to the students by email. 

During the third phase, the 1,085 student responses were reviewed and 28 questionnaires were discarded because they were not properly completed. The data analysis was carried out between November and December 2021, ensuring the confidentiality of the participants. Data analysis was carried out following the human research guidelines of the Ethics Committee of the University of Granada and the ethical principles established by the Declaration of Helsinki in 1975 and its revision in Brazil in 2013.

Data analysis

SEM analysis was used to test the fit of the proposed theoretical model with the data obtained due to its suitability for testing the mediation hypotheses, as opposed to other techniques such as linear regression analysis (Gunzler et al., 2013). Furthermore, this analysis also enables the estimation of measurement error (Garson, 2012). IBM Statistical Package for Social Sciences version 25.0 (IBM Corp., Armonk, NY, USA) was used for descriptive and exploratory analysis while IBM Statistical Package for Social Sciences Amos version 26.0 (IBM Corp., Armonk, NY, USA) was used for SEM analysis.

Little’s test (1988) was used and produced values for total and random absence (MCAR). Because such data were available in the database and did not exceed 3% of the total number of cases (28 participants had missing values), the Garson (2012) approach was used and those cases with missing values were removed. This procedure was carried out in order to avoid compromising the reliability of the data. Thus, the sample was slightly reduced to 1,057 participants. 

Following this process, the multivariate normality assumptions were checked. The Mardia coefficient test obtained 12.33, indicating that the data were not normal (Ullman, 2006). This requires the use of robust adjustment statistics, such as the Satorra-Bentler, which allows for the reduction of biases such as those caused by non-normal data distributions (Kline, 2015). Moreover, it is important to treat ordinal variables with 5 or more categories as continuous variables and to use the maximum likelihood method with a robust statistic (Rhemtulla et al., 2012).

Convergent validity and reliability or internal consistency were assessed according to the average variance extracted (AVE) and composite reliability (CR). AVE .50 and CR .70 were considered adequate (Fornell & Larcker, 1981).

Goodness-of-fit for both the measurement model and SEM was assessed by: a) S-B x², degrees of freedom (df) and p-values; b) comparative fit index (CFI) as incremental fit index; c) normalised fit index (NFI) analysis; d) incremental fit index (IFI); e) Tucker-Lewis index (TLI); and f) root mean square error of approximation (RMSEA) with 90% confidence interval (CI). Taking into account the sample size and the number of indicator variables, an adequate model fit was defined as S-B x² value of p ≥ .05, CFI .92 and RMSEA .07 (Hair et al., 2014).

Finally, Harman’s one-factor test was used to check the variance problem of the common method, as this could compromise the validity of the results (Podsakoff et al., 2003). 

Results

First, the model developed for students who play contact group sport. The model analysis shows a good fit (S-B x² = 210.416; df = 32; p < .000; NFI = .943; IFI = .958; TLI = .986; CFI = .971; RMSEA = .062). 

Table 1 and figure 1 show the regression weights of the theoretical model, with statistically significant differences at p < .001. In terms of self-concept, positive relationships are observed with the social (SO) (β = .719), family (FA) (β = .619), physical (PH) (β = .641), emotional (EM) (β = .337) areas. Similarly, EI is found to be positively related to emotional clarity (EC) (β = .950) and emotional attention (EA) (β = .277). Regarding direct violence, a negative relationship with self-concept (β = -.150) and a positive relationship with EI (β = -.211) is observed. Finally, for indirect violence, negative relationships with self-concept (β = .044) and negative relationships with EI (β = -.108) are observed.

Table 1

SEM for contact team sports players.

See Table

Figure 1

See Full Size

Examining the model developed for students who play non-contact group sport, it can be seen that the model fit is good (S-B x² = 210.853; df = 32; p < .018; NFI = .924; IFI = .976; TLI = .909; CFI = .965; RMSEA = .057). 

Table 2 and figure 2 show the regression weights of the theoretical model, with statistically significant differences at p < .001. Direct violence is negatively correlated with EI (β = –.071) while indirect violence is positively associated with EI (β = .301).

Table 2

SEM for non-contact team sports players.

See Table

Figure 2

See Full Size

The proposed SEM for students who play individual non-contact sport shows a good fit (S-B x² = 326.986; df = 32; p < .000; NFI = .955; IFI = .981; TLI = .953; CFI = .969; RMSEA = .053). 

Table 3 and figure 3 show the regression weights of the theoretical model, with statistically significant differences at p < .001. The self-concept values reveal positive relationships with the categories AC (β = .580); SO (β = .704); FA (β = .578); and PH (β = .498). On the other hand, EI is positively related to ER (β = .521); EC (β = .789); and EA (β = .310). Observing both times of violence (direct and indirect), they reveal negative associations with EI (β = –.022; β = –.004) and with self-concept (β = –.310; β = –.156).

Table 3

SEM for individual non-contact sports players.

See Table

Figure 3

See Full Size

Finally, the SEM associated with university students who play individual contact sport shows a good fit (S-B x² = 124.229; df = 32; p < .000; NFI = .925; IFI = .992; TLI = .947; CFI = .978; RMSEA = .049). 

Table 4

SEM for individual contact sports players.

See Table

Figure 4

See Full Size

As can be seen in the table, the regression weights of the theoretical model are shown, with statistically significant differences at p < .001. The self-concept values reveal positive relationships with the categories AC (β = .724); SO (.813); EM (β = .347) FA (β = .849); and PH (β = .551). On the other hand, EI is positively related to ER (β = .350); EC (β = .356); and EA (β = .327). Focusing on direct violence, negative associations are found with EI (β = –.873) and with self-concept (β = –.003). Finally, indirect violence is negatively associated with EI (β = –.797) and with self-concept (β = –.053) (figure 4).

Discussion

This study examined the levels of emotional intelligence, self-concept and violence according to the sport modality practised, while establishing relationships between the variables considered. Thus, EI and self-concept have been found to have a positive relationship, in line with previous studies (Martínez-Monteagudo et al., 2021; Sánchez-Zafra et al., 2022).

A negative relationship between EI and self-concept has also been identified with both types of violence. In this regard, the study by (Díaz-López et al., 2019) with adolescents found a negative relationship between EI and aggressive behaviour. In turn, previous studies have also identified a negative relationship between self-concept and violent behaviour (Castro-Sánchez et al., 2019; Sánchez-Zafra et al., 2019). In this sense, EI has been shown to improve emotional expression (Castillo-Viera et al., 2021).

Focusing on the forms of violence, self-concept is found to be negatively associated with direct and indirect violence, in line with previous literature (Blakely-McClure & Ostrov, 2016; Jenkins & Demaray, 2012; Malhi et al., 2014).

Similarly, EI was also found to be strongly and negatively related to both types of violence. This relationship has also been observed in the literature, and there is research in which EI programmes have been implemented to prevent violence in adolescents, an approach that supports the importance of fostering emotional skills among students with a view to reducing this type of behaviour (Garaigordobil & Peña-Sarrionandia, 2015). In addition, EI is also related to the practice of physical activity, with a direct relationship between the two factors as reported in the systematic review by Puertas-Molero et al. (2017).

EI is negatively associated with direct violence and positively associated with indirect violence. In this regard, Porche (2016) states that EI should be understood as a cognitive-behavioural strategy for overcoming violence, as it enables the determination of the person’s potential to perform violent actions, self-awareness about the emotions they feel and that motivate them to react violently or the role that past and present experiences of violence have on their behaviour.

Regarding the relationship between sport and violence, this study found that people who do not play sport tend to develop both types of violence. Along these lines, the study by Medina-Cascales and Reverte-Prieto (2019) concluded that high level of sport participation in less competitive environments is associated with lower rates of violence. In contrast, the research by Martínez-Martínez et al. (2017) did not report significant relationships between the sport modality played and violent behaviour, despite the benefits that physical activity brings to the management and control of aggression according to other studies (Magnan et al., 2013; Sánchez-Alcaraz et al., 2020).

At the same time, it is important to highlight the limitations and prospects of this study. Firstly, the cross-sectional methodological design does not allow for the identification of the evolution of these variables, as well as causal relationships between them. In this line, longitudinal research, with pre-post methodological designs, will contribute to a better understanding of the behaviour of these variables, which will allow for the examination of the effectiveness of certain psychological constructs on student behaviour, as well as for testing educational programmes aimed at providing students with the strategies they need to cope with situations that challenge their psychosocial well-being. 

Conclusions

From the findings obtained, it can be concluded that there is a pressing need to develop educational programmes based on physical activity and the promotion of psychosocial factors among university students. In particular, EI-based programmes will lead to a reduction in violent behaviour. This initiative could equip students with the necessary strategies to cope adequately in different contexts in diverse areas of their lives.

Likewise, from the findings it has been observed that promoting group contact sports requires this type of programme to a greater extent in order to mitigate the predisposition towards direct violence. To a certain extent, it is a matter of promoting educational programmes aimed at replacing violent behaviour with more adaptive behaviour from an integral education perspective. In contrast, the finding of higher emotional control and lower propensity towards violence among university students who play individual contact sports (wrestling, karate, etc.) calls for reflection on the type of sports promoted among students. Although it is true that the results obtained should be interpreted with caution, the findings obtained encourage reflection on the sports that need to be promoted in educational programmes in higher education and, considering that this study involves future teachers, it is possible to transfer these good practices to lower levels of education. One area to consider is the development of physical activity programmes or active breaks within the different subjects taught in university degrees in order to release stress and anxiety produced by academic performance.