Social identity and engagement: An exploratory study at university

Abstract

Within the framework of research on students’ active performance in their study habits, the aim of this study is to analyze a model predicting the effect of social identity and personal initiative on engagement in university students. We conducted a cross-sectional study on 266 students from different Spanish universities. The resulting data were analyzed using SPSS Macro MEDIATE. Evidence was found for the proposed model. Only group-identity predicted personal initiative and engagement. Analysis revealed the mediating role of proactive behavior on engagement in university students. It is concluded that the university management may intervene, from an organizational-culture approach, promoting guidelines to reinforce students’ sense of belonging by enhancing initiative and autonomous problem solving in learning behaviors.

Resumen

Dentro del marco de investigación sobre desempeño activo de estudiantes, el objetivo de este trabajo es analizar un modelo predictor del efecto de identidad social e iniciativa personal en el compromiso de estudiantes universitarios. Se llevó a cabo un estudio transversal con 266 alumnos de diversas universidades españolas. Los datos se analizaron con la Macro MEDIATE de SPSS. Se encontraron pruebas para el modelo propuesto. Sólo la identidad grupal predecía la iniciativa personal y el engagement. El análisis mostró el papel mediador de la iniciativa personal en el engagement de estudiantes universitarios. Entre las conclusiones se destaca la posibilidad de promover desde la Universidad el sentido de pertenencia para generar iniciativa personal y la solución autónoma de problemas en los comportamientos de aprendizaje.

It is a common goal of higher education institutions to achieve successful academic performance of their students. Doing more than expected, suggesting new working objectives, and actively studying to achieve a degree are some examples of desirable behaviors in the new framework of higher education focusing on the achievement of professional skills and requiring a great deal of autonomous work from students. In recent years, attention has been focused on engagement, a significant predictor of academic performance. Engagement is defined as a positive, work-related state of mind characterized by vigor, dedication, and absorption at work (Schaufeli & Bakker, 2003). Apart from work setting, surveys on engagement have also been conducted to measure students’ academic performance. Results reveal that engagement in studying allows students a better management of difficulties encountered in the daily academic life (Salanova, Martínez, Bresó, Llorens, & Grau, 2005). Subsequent research has found that personal resources are among the aspects contributing to engagement (Xanthopoulou, Bakker, Demerouti, & Schaufeli, 2007). Personal resources are positive self-evaluations linked to resiliency and referred to individuals’ sense of their ability to control and impact upon their environment (Hobfoll, Johnson, Ennis, & Jackson, 2003). Proactive behavior has been identified as one of the resources contributing to an active performance (Lisbona & Freser, 2012; Salanova & Schaufeli, 2008). Furthermore, engagement is the positive pole of two underlying processes and is characterized by high energy levels at work and a strong identification with the organization (González-Romá, Schaufeli, Bakker, & Lloret, 2006).

The Social Identity Theory (SIT) suggested by Tajfel (1978) and its application to the organizational context proposed by Mael and Ashforth (1992) can explain this identification process. From the SIT approach, Social Identity is defined as the perceived sense of belonging to a group category, whereas in the business context Organizational Identification is considered as the perception of the individual as a member of an organization and the experience of the organization's success and failure as one's own (Mael & Ashforth, 1992). Therefore, organizational identification explains the relationship between an individual and the organization as a cognitive link binding the definition of the organization with the definition of self (Dutton, Dukerich, & Harquail, 1994). Accordingly, in broad sense we can speak of identification with the work group when an individual becomes a member of a profession or a group (Mael & Ashforth, 1992). Literature shows that both identification types are compatible (Foreman & Whetten, 2002). The individual may hold multiple identities or nested identities coexisting in harmony. The individual may also get on well with his or her colleagues and feel at ease with the values of the organization (Foreman & Whetten, 2002). Nevertheless, not all identities are equally valued; therefore, there is a hierarchical organization based on category salience with varying relative weights. For instance, it has been pointed that, for a given individual, the identification with the organization may decrease while the identification with the group rises (Cappelli, 2000). SIT establishes that individuals are self-categorized as members of a group category with a positive valence. However, a critical review of SIT (see Scandroglio, López, & San José, 2008) described that group processes should be studied starting with a viewpoint that places greater emphasis on the complex character of a dialogic social reality. In this sense, these authors note that recent research indicates that the processes of self-definition and hetero-definition are dynamic and changing, combining a number of formal and motivational elements and resulting from the interaction between characteristics of the environment and a set of a subject's resources, articulated in a multidimensional space that combines different criteria of inclusiveness and differentiation.

To put it differently, the classification within an organization or group entails feelings of self-esteem and recognition that have an impact on the behaviors of individuals, and interaction with the context and the motivational aspects will also have a role in this self-categorization. In this line, it has been observed in an organizational context that identification with the group predicts its members’ attitudes and behaviors (Van Dick & Wagner, 2002). It has also been pointed out that identification with the group explains a higher variance with respect to job satisfaction, participation in group activities, commitment, and personal-initiative in organizational contexts (Lisbona, Morales, & Palací, 2006; Snape, Redman, & Chan, 2000; Van Dick & Wagner, 2002; Van Dick, Wagner, Stellmacher, & Christ, 2005; van Knippenberg & van Schie, 2000; Veenstra & Haslam, 2000). In contrast, identification with the organization is positively related to trust, motivation, performance, and citizenship behaviors (Abrams, Ando, & Hikle, 1998; Bhattacharya, Rao, & Glynn, 1995; Haslam, 2001). Hence, evidence suggests that identification with the organization and with the group converge in extra-role or citizenship behaviors at the workplace that go beyond the requirements of the job. These behaviors have a strong proactive component, understood as a stable trend to effect changes in order to preserve the life of the group (Crant, 2000).

Within this approach to proactive behavior, Personal Initiative is defined as a self-starting response and proactive behavior in the pursuit of an active participation in performance (Frese & Fay, 2001). This self-starting behavior has demonstrated to be a predictor of engagement in different organizational settings (Lisbona & Frese, 2012). Accordingly, employees with strong personal-initiative tend to seek solutions to problems, to act regardless of what others do, etc. The model of personal-initiative proposed by Frese and Fay (2001) identifies two types of causes determining personal-initiative: distal and proximal causes. Distal causes, on the one hand, (i.e., environmental support, knowledge and skills, and personality) predict personal-initiative indirectly through proximal causes; social identity may be included in this group. On the other hand, proximal causes are orientations promoting personal-initiative by allowing people to believe that they can show personal-initiative in a particular context (i.e., self-efficacy, change orientation) (Frese & Fay, 2001). Hence, according to this classification and starting from the assumptions of the SIT, we would expect that students with high identification with the organization and with the group will develop self-starting behaviors in their approach to study work, and this will predict engagement as it provides them with a positive value of identification with the group.

Therefore, the objective of this study is to analyze a predictive model of social identity on engagement with the mediation of personal-initiative in university students.

The hypotheses of the study are the following:Hypothesis 1

Organizational- and group-identification predict students’ personal-initiative.

Hypotheses 2

Personal-initiative mediates the relationship between social identity and engagement in students.

Method

Participants

The present study was conducted with a sample of 266 participants from three universities: Universidad Miguel Hernández of Elche (49.2 percent), Universidad Nacional de Educación a Distancia (UNED) (39.1 percent), and Universidad de Alicante (9.4 percent). Most participants were studying Psychology, Occupational Therapy, and Advertising and Public Relations (84.8 percent). The average age of participants was 28.42 years (SD = 9.84); 71.8 percent were females; and 63.3 percent were first- and second-year students. UNED students’ profile is different from traditional university students, since they are students who combine studies with a full-time job, thereby needing more years to complete their studies; furthermore, a lot of UNED's students are workers that decided to start studying in the adulthood.

Variables and Instruments

Social identity. It was measured with a social identification scale developed ad hoc for this study, considering the review of the major social and organizational identification scales conducted by Haslam (2001) and the recommendations of van Knippenberg and van Schie (2000) and Grice, Jones, and Paulsen (2002). The organizational-identification scale and the group-identification scale included, respectively, six items; an example of item is “I feel personally insulted when someone criticize my university” and “I present myself as a student of a degree”. We used a 1-to-5-point Likert-type scale for the responses (from I totally disagree to I totally agree). The KMO test shows satisfactory results (KMO = .77, χ² = 1282.3, df = 91, p = .00); two factors explain 39% of variance. Cronbach's alpha is .82 and .73 for organizational-identification and group-identification respectively.

Personal initiative. It was measured with the self-report scale by Frese, Fay, Hilburger, Leng, and Tag (1997) adapted to the Spanish population by Lisbona and Frese (2012). It includes six items in a 1-to-5-point Likert-type scale (from I totally disagree to I totally agree). An example of item is “When I have the opportunity to get involved in something, I do it” The KMO test shows satisfactory results (KMO = .78, χ² = 366.3, df = 21, p = .00); this factor explained 39% of variance. Cronbach's alpha is .70.

Engagement. It was measured with the Spanish version for students of the vigor, dedication and absorption - Utrecht Work Engagement Scale (Schaufeli & Bakker, 2003). An example of vigor (six items) is “I can keep on studying during long periods”, of dedication (five items) “I am proud of studying this degree”, and of absorption (six items) “I am absorbed in my studies”. The responses were measured on a 0-to-6 Likert-type scale (from never to always or every day). The average of scale is used as an indicator. The KMO test shows satisfactory results (KMO = .91, χ² = 2618.87, df = 120, p = .00), the three factors explaining 60% of variance. Cronbach's alpha is .88 for the global scale.

Procedure

The tests were administered after an informative meeting with some university lecturers of the different universities. To select the sample, we used a non-probability method. We asked for volunteers among students from different years of different degrees. At every moment, the participants were aware of the goals of the research and the voluntary nature of their participation in the survey and the study.

Data Analysis

Data were analyzed using SPSS statistical software. To analyze the mediation effect MEDIATE (Hayes & Preacher, 2014) was used. Regarding measuring instruments, confirmatory factor analysis and Cronbach alpha analysis showing the construct validity and internal consistency of the measures were used. To prevent random capitalization in the study of the models, sample was randomly divided into two groups. The study model of mediation with different dependent variables is analyzed, with descriptive statistics and correlation for both groups, and additionally Baron and Kenny criteria for the analysis of mediation models were taken into account. Age was taken as a control variable. The significant test value (p < .05) in the OMINIBUS test is indicative of mediation effect (Hayes & Preacher, 2014).

Results

To analyze the two models that were proposed, the sample was divided randomly. Sample 1 (n₁ = 136) was composed by 66 percent women, 60 percent were in the first three years and the mean age was 28.2 years (SD = 9.5). As for sample 2 (n₂ = 129), 77 percent were women; the mean age was 28.6 years (SD = 10.1). Fifty-nine percent of the sample was in the first three years. Table 1 shows the descriptive results for both groups. Specifically, in the group 1, group-identification has a significant relationship with personal-initiative and with engagement, while organizational-identification has only a statistically significant relationship with engagement (Table 1). Obviously, group and organizational-identification are related. In group 2, all indicators show a statistically significant relationship to each other.

Table 1.

Descriptive Results.

	M	SD	1	2	3	4
Sample 1 (n = 136)
1. Engagement	4.00	1.02	-	.58^**	.35^**	.59^**
2. Group Identity	4.00	0.73		-	.42^**	.31^**
3. Organizational Identity	2.96	0.80			-	.19
4. Personal-initiative	3.49	0.57				-
Sample 2 (n = 129)			1	2	3	4
1. Engagement	4.04	0.92	-	.60^**	.36^**	.55^**
2. Group Identity	3.88	0.66		-	.41^**	.25^*
3. Organizational Identity	2.99	0.84			-	.34^**
4. Personal-initiative	3.42	0.60				-

*

p < .05, ^**p < .01

For the mediation analysis (see Table 2), the results regarding engagement as a dependent variable show that group-identification has a direct effect on engagement (β = .76, SE = 0.14, t = 5.26, p = .00), but the same effect is not observed between organizational-identification and engagement (β = .07, SE = 0.14, t = .53, p = .59). Controlling for personal-initiative, an indirect effect of group-identification on engagement is observed (β = .59, SE = 0.13, t = 4.60, p = .00), but not with organizational identity (β = .04, SE = 0.12, t = 0.35, p = .72). Thus, group identification and personal-initiative explains engagement responses (R² = .75, F = 21.95, p = .00; see Figure 1). Regarding the age as control variable, an effect in the proposed model for engagement is not observed (β = .01, SE = 0.01, t = 1.76, p = .08).

Table 2.

Mediate Analyses.

	b	SE	t	p-value	R^2
Engagement
Step 1
Organizational identity	.07	0.14	0.53	.59	.65^**
Group identity	.76	0.14	5.26	.00
Step 2
Organizational identity	.04	0.11	0.44	.65	.34^**
Group identity	.25	0.11	2.21	.03
Step 3 and 4
Personal initiative	.65	0.13	4.87	.00	.75^**
Organizational identity	.04	0.12	0.35	.72
Group identity	.59	0.13	4.60	.00
Personal initiative
Step 1
Organizational identity	.19	0.10	1.87	.06	.41^**
Group identity	.21	0.13	1.53	.13
Step 2
Organizational identity	.06	0.10	0.65	.51	.74^**
Group identity	.93	0.13	6.86	.00
Step 3 and 4
Engagement	.37	0.12	2.99	.00	.53^**
Organizational identity	.16	0.09	1.73	.08
Group identity	-.13	0.17	-0.80	.42

**

p < .01.

Figure 1.

Proposed Model. ** p < .01.

Regarding the model with personal-initiative as dependent variable (see Table 2), the results show no direct effect of organizational-identification on personal-initiative (β = .19, SE = 0.10, t = 1.87, p = .06), even for group-identification (β = .21, SE = 0.13, t = 1.53, p = .13).

Discussion

The aim of the present study was to analyze a model predicting the effect of social identity on engagement mediated by personal-initiative in university students. Results confirm partially hypothesis 1, because group-identification predicts personal-initiative and engagement. Students may be identified as members of a degree rather than as members of a university, because the defining characteristics as a degree member are most salient for the student than the defining characteristics as a university member, regardless of students’ age. One explanation for this may be that students in the sample have a strong vocational component. In accordance with Scandroglio et al. (2008), this could be a personal resource that interacts with the environment (i.e., degree tasks or compared classmate) and creates dynamic process for self-definition.

According to these results, group-identification could be considered as a proximal cause for initiative behavior in Frese and Fay's (2001) model. Thus, if a certain level of prediction is observed for being a member of a group of fellow students, this would in turn predict proactive behaviors. Nevertheless, this effect is weak and therefore must be interpreted with caution, while other existing variables with greater weight, as self-efficacy or control aspirations, explain orientation towards personal-initiative (Lisbona et al., 2006). Along these lines, results show that only group-identification score similarly with personal-initiative. These results are in line with earlier studies in which organizational and group identification led to proactive behaviors when they are examined separately but not simultaneously (Abrams et al., 1998; Bhattacharya et al., 1995; Haslam, 2001; Lisbona et al., 2006; Snape et al., 2000; Van Dick & Wagner, 2002; Van Dick et al., 2005; van Knippenberg & van Schie, 2000; Veenstra & Haslam, 2000). Thus, according to SIT, students may show behaviors that are beyond their requirements so as to preserve a positive valence for one category. In this sense, only one identity predicts positive outcomes.

In relation to hypothesis 2, we expected personal-initiative to mediate the relationship between students’ both social identification and engagement. The results of our study confirm this hypothesis, but only in the case of group-identification. Although group-identification and personal-initiative explain active performance of students, personal-initiative mediates the action sequence. A strong identification with fellow students enhances engagement with the mediation of personal-initiative. In accordance with previous studies on proactivity (see Salanova & Schaufeli, 2008), our results indicate that personal-initiative is a proactive behavior that affects engagement.

The tested model, where personal-initiative is a mediator variable, shows betters results in the mediation analysis than when engagement is a mediator. However, the causal relationship between personal-initiative and engagement is unclear. Although in Frese and Fay's (2001) model there is a motivational element, it may respond to items of intrinsic motivation unlike the engagement which is generated by environmental elements (i.e., Job Resources; Demerouti, Bakker, Nachreiner, & Schaufeli, 2001). Thus, Fay and Frese (2001) mention these aspects in their proposal of the personal-initiative theory. Specifically, they said that “Self-starting behavior in terms of voluntary action has been described in the field of intrinsic motivation (…). The rewards of intrinsically motivated activities are the experience of effectance and autonomy and the experience of positive emotions such as enjoyment and excitement” (Fay & Frese, 2001). Notwithstanding, a better explanation for the results obtained is defining personal-initiative as a personal resource over agreement proposed by Xanthopoulou et al. (2007), that allows people to generate engagement responses, as occurs with self-efficacy.

Our findings suggest that group identification drive students to present and categorize themselves as members of grade group and to exercise personal-initiative behaviors in those aspects they consider to be important for the group, generating a better performance in their studies. However, it is worth noting that the mediation produced is partial and has a low effect. There are possibly other personal resources that may justify higher variances of the active performance, as it occurs with self-esteem (Salanova et al., 2005), which consequently could play a double role, as a proximal cause of personal-initiative and a mediator with engagement.

Regarding the limitations of the study, we would mention its cross-sectional design, which limits our ability to infer causality. To sum up, several conclusions may be drawn from the results presented here. First, only one type of identification may exist. Second, despite its weak effect, group social identity may be a proximal cause for personal-initiative. In this line, these empirical results support the need to introduce a new antecedent in the model of the antecedents and consequences of personal-initiative. Third, group-identification has an indirect effect on students’ engagement, with the mediation of personal-initiative. Nevertheless, due to the unclear causal relationship between personal-initiative and engagement showed in literature, there is a need to study these variables in a longitudinal design.

Practical implications of our study offer two possibilities for the analysis and application. In terms of organizational-culture, faculties may promote guidelines for the university and the studies such as to seek opportunities, do more than is required, seek solutions to problems, and to act regardless of what others do. All of this may contribute to the students’ personal-initiative. These guidelines would be distal causes for initiative and would impact as a proximal cause on social identity. Lastly, in pedagogical terms, conclusions suggest to propose study tasks that facilitate opportunity seeking behaviors, solving unforeseen events, and implementing ideas.

Conflict of Interest

The author of this article declares no conflict of interest.