Social Responsibility Attitudes and Behaviors’ Influence on University Students’ Satisfaction

: This study focused on university social responsibility (USR). Corporate social responsibility is currently an extremely common strategy implemented by organizations. Higher education institutions are also introducing this strategy to enhance their performance, seeking to ensure that every university action is socially responsible and oriented toward achieving advantages over competitors. This competitive advantage is the result of a social responsibility vision, which has an ethical core, that the University has implemented or is implementing among all its stakeholders. These institutions work in four areas: instruction, research, management, and projection to society. Universities must thus strive to meet the interests of different stakeholders’ interests. This research concentrated on university students as an important stakeholder. The main objective was to evaluate university students’ participation in USR activities, as well as assessing the impact of relevant university practices. In addition, the study sought to measure the existing causal relationship between students’ participation and their university’s practices in terms of student satisfaction. The fieldwork was conducted with an electronic survey distributed to a group of University of Extremadura students in Spain. A total of 362 valid questionnaires were collected, which were processed using structural equation modeling and partial least squares. The results have implications for university management in the area of social responsibility, with regard to the new USR trends are revealed. In terms of originality and value, this research emphasized a specific stakeholder in universities, namely students, and ways their satisfaction can be achieved through USR.

The European Commission (2001), in its Green Paper: To Promote a European Framework for Corporate Social Responsibility, considers CSR to be companies' voluntary integration of social and environmental concerns into their business operations and interactions with stakeholders. In 2011, this definition was strengthened by the European Comission (2011) when it pointed out that firms are responsible for their impact on society. This organization made an explicit reference to the need for collaboration with stakeholders to integrate social, environmental, and ethical concerns and respect for human rights and consumer concerns into company strategies and business operations. Other types of companies, such as cooperatives, foundations, and non-profit entities, are also working on these issues based on the triple bottom line perspective (Basterretxea and Martínez 2012;Castilla-Polo et al. 2017;Winston 2002). In addition, universities have begun to adopt the CSR approach in such a way that this strategy has already permeated numerous universities' functions (Adomssent et al. 2007;Aznar Minguet et al. 2014;Ferrer-Balas et al. 2008;Latif 2018).
The present study focused on university social responsibility (USR). Currently, universities define their mission, vision, values, principles, and corporate culture in ways that ensure socially responsible operations (Larrán Jorge et al. 2012a, 2012b. This context conditions students' ethical attitudes, which depend on many internal and external factors (Leonard and Cronan 2005), generating both individuals' holistic development and added value through their actions (Guédez 2006;Vallaeys 2007).
USR promotes higher education institutions' ability to adhere to a series of principles and values through strategies based on management, instruction, research, and extension into or projection onto society (Adomssent et al. 2007;Domínguez Pachón 2009;Ferrer-Balas et al. 2008). Baker-Shelley et al. (2017) and Rockström et al. (2009) point out that universities have a moral obligation to provide key social transformations through education and research that fully respects the planet's limitations. However, despite universities' obvious orientation toward meeting society's needs, these institutions have not yet achieved the degree of involvement in CSR that other organizations have. Universities' potential to integrate sustainability deeply into functions thus remains largely untapped (Baker-Shelley et al. 2017;Lozano 2011;Lozano et al. 2013;Moneva and Martín 2012). Notably, the concept of CSR did not reach Spanish universities until the end of the 1990s (Gaete Quezada 2011).
This situation translates into a significant opportunity for change in terms of universities' strategic orientation toward social responsibility values (Bok 2009). USR is, therefore, a concept that is being continuously developed, including moving social cohesion to the fore as a primary objective. These institutions' proposals revolve around four key parameters: instruction, research, social leadership, and commitment (Beltrán-Llevador et al. 2014). Vallaeys et al. (2009) also define USR as these institutions' impact on their environment in the organizational, educational, cognitive, and social spheres. Gaete (2012), in turn, suggests that USR is the incorporation of different interest groups into university administrations. Other researchers (Larrán-Jorge and Andrades-Peña 2015) have affirmed that USR policies should be based on stakeholder theory to bring together different groups' interests and expectations (e.g., students, professors, researchers, administration and service personnel, and managers).
The present study's objectives included three aspects. First, the research included evaluating university students' participation in USR activities. Second, an analysis was conducted of students' perceptions of USR practices in their university. Last, the study examined the causal relationship between USR and student satisfaction. The research applied the technique of structural equation modeling based on partial least squares (PLS). This paper is structured as follows. After the present introduction, a preliminary review is presented of the literature on CSR and USR. Then, the proposed conceptual model and research hypotheses are discussed, after which the study's methodology is explained. The results are given next, followed by the conclusions, limitations, and future lines of research.

Stakeholder Theory
CSR involves an integrated approach to business management, production, marketing, and relationships with third parties (e.g., workers, customers, suppliers, distributors, or regional communities) (Lafuente et al. 2003). From this perspective, a company is considered socially responsible when it meets the expectations of different stakeholders affected by its operations (Clarkson 1995;Frooman 1999;Mitchell et al. 1997;van Marrewijk 2003). Stakeholder theory covers the ways companies and organizations can respond to the pressures exerted by various groups (e.g., non-governmental organizations [NGOs], administrations, associations, customers, and media) that benefit from firms' responsible behavior (Garriga and Melé 2004).
Stakeholder theory suggests that, currently, all interest groups involved in each organization have to be satisfied, so these organizations need to work to achieve the best outcomes for their stakeholders. Amaral and Magalhães (2002) argue that, in university contexts, the concept of stakeholder refers to any singular or collective entity with a legitimate interest in higher education, and thus the right to intervene in these institutions. This definition includes numerous interest groups such as students, teachers, administrations, services personnel, local government, unions, companies, employers, families, and society at large (Burrows 1999;Jongbloed et al. 2008). The present study focused on a specific, extremely important stakeholder: students. The research's aim was to extract information about their attitudes toward and behaviors in the University of Extremadura, Spain, regarding social responsibility, as well as its repercussions for these students' satisfaction.
The factors that enhance student satisfaction have become an increasingly important issue for academic institutions because the parameters of satisfaction are linked to universities' future. These limiting factors provide the information needed to prepare academic success strategies that generate competitive advantages, implement plans for improvement, or analyze these institutions' evolution over time (Secchi 2006;Zu and Song 2009). The current study's approach was to examine links between student satisfaction and the development of USR initiatives. In this context, universities that implement socially responsible practices have to strengthen these institutions' teaching-learning processes through USR activities, which constitute a key aspect of their students' future. Among the most significant challenges universities face is how to enhance students' motivation (European Commission 2001) and satisfaction (Lamberton 2005;Sánchez-Hernández and Mainardes 2016;Saleem et al. 2017;Va 'zquez et al. 2016) in a socially responsible environment.
Universities' current reality includes the fact that the traditional functions of instruction, research, and facilitation of their students' social and personal expansion have had to adapt to deal with new trends and generations (Molina et al. 2012). These institutions have thus incorporated another series of values based on their stakeholders' social responsibility, with a special role given to students (Latif 2018;Moneva and Martín 2012). Universities now have opportunities to take advantage of the socioeconomic development generated by USR values, thereby transcending traditional academic limits in the context of a globalized environment (Beltrán-Llevador et al. 2014). Socially responsible universities have incorporated into their own best interests efforts to help meet the needs of the different groups with which these institutions interact (e.g., students), with the ultimate goal of achieving both academic and social objectives (Gaete Quezada 2012).
In this regard, researchers have highlighted cases such as Bilgi University in Istanbul, Turkey, which has adopted an action strategy oriented toward long-term social responsibility practices. These kinds of institutions seek to develop a better reputation and differential value compared to other universities, and various relevant groups' satisfaction (Jongbloed et al. 2008;Saleem et al. 2017;Sanje and Senol 2012). The present research focused primarily on characterizing the attitudes of the University of Extremadura's students toward USR and, simultaneously, their behavior as a core group of higher education stakeholders (Jongbloed et al. 2008). The current study thus examined a set of USR initiatives in order to understand these stakeholders' commitment to their university, with the final goal of defining how closely this loyalty is related to their satisfaction.

Theoretical Approaches to CSR
We would like to note that throughout this paper, the concepts of CSR and social responsibility are used interchangeably. After this precision, in recent years, more and more organizations worldwide have emphasized CSR activities because of their general implications for, and importance to, society (Baldarelli and Gigli 2014;Bradford and Fraser 2008;Deakin and Hobbs 2007;Gallardo-Vázquez et al. 2013;Sánchez-Hernández 2014a, 2014b;González-Loureiro and Pita-Castelo 2012;Herrera-Madueño et al. 2016;López-Cózar-Navarro and Benito-Hernández 2017;Perrini et al. 2007;Turker 2009). CSR refers to both obligations and ethical commitments that go beyond strictly legal issues and that are derived from the impacts that companies and organizations' activities have on different areas, such as social, environmental, or human resources (Deakin and Hobbs 2007;Doh and Guay 2006;McWilliams et al. 2006;Yu et al. 2017). Bowen (1953) defined companies' social responsibility policies as these firms' obligation to follow accepted norms for assuming responsibility when making decisions or setting objectives, thereby showing respect for society's values. Carroll (1979) subsequently referred to corporate responsibility as organizations going a step beyond economic profitability and assuming social tasks. The cited author proposed a model with four interdependent CSR dimensions. The first is economic aspects, which aims to achieve profitability by marketing products or services. The second is the legal dimension of compliance with current legal frameworks. The third is ethics, which refers to business activities that comply with ethical standards, and the last is the philanthropic dimension, which comprises companies' voluntary participation in socially responsible activities.
CSR has been gaining increasing prominence within firms' strategies, actions, and objectives (Aguinis and Glavas 2019). This trend has increased both companies' investment in social works and communication about these efforts (Marín Rives and Ruiz de Maya 2007) since this can positively influence the potential final consumers' behavior (Sureshchandar et al. 2001) or form a competitive difference in terms of other companies (Gallardo-Vázquez and Sánchez-Hernández 2012). CSR, therefore, implies a change in firms' management philosophy with regard not only to their business activities' economic performance but also a genuine corporate commitment to creating internal and external value for society (Obeidat 2016; Scherer and Palazzo 2011).

USR Strategy
Ineffective governance and a lack of vision regarding social responsibility can restrict any higher education institution's ability to contribute to societies' sustainable development (Pollock et al. 2009). Thus, approaches to integrating USR have been proposed by numerous authors, with a broader or narrower scope. Some researchers have focused on integrating the analysis, implementation, and evaluation of CSR policies into strategy (Leitao and Silva 2007). Others have concentrated on USR's significant effect on society (Ahmed 2012) and, more specifically, on the environmental dimension (Jabbour 2010;Nejati et al. 2010). Baker-Shelley et al. (2017) point out that universities play an important role in an era in which societies must undergo a profound transformation due to climate change and other circumstances. This process gives universities the impetus to pursue changes that contribute to sustainability (Adomssent et al. 2007;Aznar Minguet et al. 2014;Baker-Shelley et al. 2017;Beynaghi et al. 2016;Latif 2018;Trencher et al. 2014).
Enríquez et al.'s (2018) work provides a description and explanation of university administrations' contributions to the management of education based on higher education institutions' social commitment in Ecuador. These universities play a fundamental role in implementing strategies that align these institutions with the needs of a globalized contemporary world in a highly competitive environment (Hemsley-Brown and Oplatka 2006;Veldman 2018). In addition, different researchers have focused on the need to develop a methodological model of USR with the relevant indicators, which could help universities obtain the information needed to analyze their level of compliance and influence (Noguera et al. 2014).
The Baldassarre Report (European Parliament 2013a), entitled Corporate Social Responsibility: Responsible and Transparent Behavior of Companies and Sustainable Growth, makes some references to training in USR. This report specifically highlights "the importance of offering specific training in matters of corporate social responsibility in universities" (p. 20). The report also underlines the "importance of the link between companies and schools that provide upper secondary education, particularly at the level of [h]igher [e]ducation" (p. 43). The Baldassarre Report further emphasizes "the role of corporate social responsibility in the acquisition of practical experience by students during their studies" (p. 43).
In addition, the Howitt Report (European Parliament 2013b), with the title Corporate Social Responsibility: Promoting the Interests of Society and a Path towards Sustainable and Inclusive Recovery, makes numerous references to CSR in education: [The report first] asks the Member States, in cooperation with the Commission and, if appropriate, with the universities, to examine how to integrate [C]SR, responsible management and responsible citizenship in education and training [to ensure the] professional … management of all future business leaders (p. 24).
This report also "urges the … [European Union] to play an active role in raising awareness of the contribution that companies can make to society in the field of culture, education, sports and youth through [C]SR" (European Parliament 2013b) (p. 25). Finally, the report recognizes "the importance of increasing [C]SR practices and policies in large companies, therefore, asking the Commission to emphasize the importance of including [C]SR in education and training, especially in business and administrative schools" (p. 32).

Conceptual Model of Student Satisfaction in Universities and Hypotheses Developed
Based on the above review of the literature, a conceptual model was proposed (see Figure 1), and working hypotheses were formulated. The present study sought to contribute to a better understanding of the process of strengthening university student satisfaction through socially responsible attitudes and behaviors, and to analyze USR practices' impacts on this aspect of universities. These organizations' different interest groups have been studied by increasing numbers of scholars from the perspective of how important these groups' participation is (Bingham et al. 2005). In the context of USR practices, students stand out as a key group because of their level of involvement and participation in USR initiatives (Alves and Raposo 2006;Jongbloed et al. 2008; Larrán-Jorge and Andrades-Peña 2015; Saleem et al. 2017;Sánchez-Hernández and Mainardes 2016;Va 'zquez et al. 2016). Universities have the ability to promote students' interactions with socially responsible practices, which is essential to making these individuals become involved in spreading positive change in society (Molina et al. 2012). Various authors (Gaete Quezada 2012) have gone even further, considering different stakeholders' participation levels to be a crucial indicator of USR. Based on these findings, the present study's first hypothesis was formulated as follows:

Hypothesis 1 (H1). University students' participation in USR activities directly and positively influences their universities' practices.
Socially responsible practices need to complement universities' academic instruction. In this way, social responsibility can be perceived by university students as going beyond educational impacts to produce more relevant social and motivational impacts on their lives (Vallaeys et al. 2009). Cereceda's (2005) study of the Catholic University of Chile students' perceptions of USR practices verified these initiatives' positive influence on the students' satisfaction, especially in terms of such aspects as excellence and integrity.
Similar to private companies with employees and clients, universities must meet the challenge of increasing student satisfaction through the philosophy of social responsibility, which is a key tool used to achieve this objective (Vallaeys et al. 2009). Along these lines, Larrán-Jorge and Andrades-Peña (2015) confirmed that meeting the expectations of higher education institutions' interest groups is a quite important result of applying a model of socially responsible management to universities. After reviewing previous researchers' contributions regarding students' perceptions of USR practices, the present study's second working hypothesis was developed:

Hypothesis 2 (H2). Universities' practices directly and positively influence their students' satisfaction.
Students who participate in USR initiatives and practices or in volunteer opportunities offered by their universities report higher satisfaction rates regarding the academic institutions in which these individuals carry out these different activities. In addition, students perceive their universities' image more favorably, as well as the other stakeholders involved in their institution (Gil et al. 2012). Therefore, providing students with an agenda and action plan that encourages their participation in and development of USR activities can transform social responsibility into an innovative, distinctive feature that builds students' emotional links with their university (Gallardo-Vázquez and Sánchez-Hernández 2012). Alnawas's (2015) research, in turn, identified student satisfaction as the cited study's final construct, basing this conclusion on other traditional aspects of university contexts such as evaluations, in addition to other, more novel, ways of improving and adapting teaching processes. These innovative methods include fostering the performance of other types of activities, as is the case with USR. In accordance with the approaches that researchers currently advocate (Cereceda 2005), the present study's third and last hypothesis was formulated as follows: Hypothesis 3 (H3). University students' participation in USR activities directly and positively influences their satisfaction with their university.

Structural Equation Models
To test the proposed conceptual model of causal relationships, structural equation models were applied based on PLS, with the help of Smart PLS version 3 software ). These models are multivariate techniques that combine aspects of multiple regression and factor analysis to simultaneously estimate a series of interrelated dependency relationships (Hair et al. 2011. At the same time, the models foster an econometric perspective focused on prediction, as well as a psychometric approach that makes room for latent or unobserved variables inferred from indicators (i.e., directly observed or manifest variables) (Chin 1998b).
The relevant literature (Fornell 1982;Chin 1998a;Hair et al. 2011Hair et al. , 2013 reports that these models, which are considered a second-generation multivariate method, facilitate, first, the incorporation of abstract constructs that are not directly observable. Second, the models allow researchers to determine the degree to which the variables that can be measured describe the latent variables. Third, this method offers a model of relationships between multiple predictor variables and criteria and, last, combines and compares hypotheses arising from prior theoretical knowledge with the empirical data obtained.

Population and Sample Selection
The population is formed of the university students enrolled at the University of Extremadura, Spain, who were enrolled during the 2017/2018 academic year, in order to provide homogeneity to data collection and subsequent analysis. It is made up of both Spanish and foreign students (including mobility students of the Erasmus Program), amounting to 18,985 students.
The present research sample 1 was composed of 362 students. The students in the sample were not selected by the authors. It is a representative sample, collecting students from all the degrees and Masters that were taught at the University during that academic year. This way, there is representation of students of all the studies offered by the University. In order to obtain the relevant data, the questionnaire was distributed via email to the entire population and students voluntarily answered it, together with a message from the vice-chancellor of the university's Infrastructure and University Services, which includes the University Social Responsibility Office. In this email, the students were invited to complete the questionnaire as a way to share their opinions of the USRrelated aspects under study. For a 95% confidence level, the sample error stood at 5.1%. Technical data are shown in Table 1.

Students' Participation in University Activities
To measure the dimension of student participation, the current research took items from scales developed by Larrán et al. (2012Larrán et al. ( , 2012aLarrán et al. ( , 2012b, Pérez et al. (2013), Vallaeys et al. (2009), andWeber et al. (2004). The first six items selected assessed the importance that university students attach to environmentally respectful initiatives and the degree of intensity with which these individuals incorporate these initiatives into their own practices. The rest of the items refer to students' volunteer 1 Regarding the sample's size, the effect size has to be specified for each regression analysis using Cohen's (1988) power tables or Green's (1991) approximation. According to Roldán and Sánchez-Franco (2012), if the researcher assumes an average effect size as defined by Cohen (1988) and aims to obtain a power of 0.80 and an alpha level of 0.05, a minimum sample of 53 cases is required if the number of predictors is 1. Based on this rule, the present study had the minimum number of cases needed to estimate the proposed model. activities that seek to improve the relevant social groups' living conditions. In total, the scale had 14 items, as shown in Table 2.

PART3
I turn off the lights in classrooms and hallways when they are unnecessary.

PART4
I share the use of private vehicles with other colleagues.

PART5
I use bicycles because of their benefits.

PART6 a
I am concerned about activities that enhance the maintenance of my student center and university's facilities.

PART7 a
I contribute to dialogues and participate in ways that ensure students' interests are incorporated in the university's decisions. I contribute during blood donation campaigns, in particular, and, in general, to other social causes.

PART13
I participate in sports activities that promote solidarity with the community.

Universities' Practices
As with the previous dimension, university practices were measured by using items from the scales developed by Larrán et al. (2012Larrán et al. ( , 2012aLarrán et al. ( , 2012b, Pérez et al. (2013), Vallaeys et al. (2009), andWeber et al. (2004). This second dimension covers the importance that students give to their university's initiatives that promote integration in the workplace, multicultural, or support contexts and overcome physical and psychopedagogical barriers. Overall, these university activities are implemented with social integration in mind, but they also include the component of students' future employment prospects. More specifically, 11 items were selected, with responses given on a 7-point Likert scale, as shown in Table 3. The university offers support and integration programs to students over 25 years of age.

PRAC6 a
The university has implemented measures that provide support to and encourage the integration of immigrant students. Larrán et al. (2012) PRAC7 a Activities have been designed so that students can acquire the skills that facilitate access to the job market.
The university has a psychopedagogical intervention unit set up for students.

PRAC9 a
Clear, well-defined strategies have been adopted to promote the elimination of physical and social barriers. Institutional support is given to activities promoting multiculturalism.
a The indicators that appear with a superscript letter are those in the model that were validated for this scale.

Student Satisfaction
To evaluate student satisfaction, items from Capelleras and Veciana Vergés's (2001) scale were used, as well as items adapted from Beerli and Díaz Meneses's (2003) scale. The present study's objective was to measure the respondents' perceptions of teaching staff's attitudes toward students, as well as the students' own degree of involvement and motivation, based on a 7-point Likert scale. A total of nine items were defined, as shown in Table 4.

SAT8 a Continuous communication occurs between teachers and students.
SAT9 a The staff clearly seek to stimulate the development of students' abilities.
a The indicators that appear with a superscript letter are those in the model that were validated for this scale.

Evaluation of Measurement Model
The model needed to be evaluated in terms of internal consistency and the scales' reliability and validity (Hair et al. 2011) (i.e., the inner model). First, each item's individual reliability was checked by examining their loadings (λ). The strictest criterion for accepting an indicator as part of a construct is that the indicator must have a loading greater than 0.707 (λ > 0.7) (Carmines and Zeller 1991;Chin and Dibbern 2010;Roberts et al. 2006). However, some authors believe that this rule is too strict and that loadings of 0.5 or 0.6 are acceptable in the initial stages of a scale's development (Chin 1998a).
In the present study, values greater than 0.6 were accepted. Based on this criterion, only the indicators PART3, PART4, PART5, PART12, PART13, and PART15 were eliminated, as the remaining indicators had appropriate loadings to remain in the model. The adjusted model is shown in Figure  2, in which all indicator loadings exceed the criterion of being greater than 0.6 and no cases present loadings less than 0.625 (see Table 5). Thus, a total of 29 indicators out of the 35 initial items were kept, and these indicators were confirmed as accurate reflections of the defined constructs, thereby confirming the reliability of the measurement scale developed.  Second, the constructs' reliability was checked. To carry out this evaluation, the constructs' composite reliability (ρc) was measured, as well as the more traditional Cronbach's alpha. Nunnally (1978), Nunnally and Bernstein (1994), and Vandenberg and Lance's (2000) recommendations were followed. According to the cited authors, values above 0.7 obtained for the composite reliability of a construct's indicator could be considered adequate because the present research was exploratory. For studies in more advanced phases, a value equal to or greater than 0.8 would be required. As Table 6 shows, all the model's constructs present values higher than 0.8, confirming the constructs' internal consistency. Third, the constructs' convergent validity was evaluated (Henseler et al. 2009) using AVE (Fornell and Larcker 1981;Hair et al. 2011). Fornell and Larcker (1981) assert that AVE values should be greater than 0.50. In Table 5, AVE values are listed for each of the model's constructs, revealing that the minimum recommended conditions were generally fulfilled. However, the value for the student participation construct is slightly less than 0.5. The results thus largely confirmed that the model constructs' convergent validity is satisfactory.
Fourth, the proposed model's discriminant validity was evaluated. Fornell and Larcker (1981) suggest using AVE, but its value needs to be greater than the square of the correlations between each construct and the other constructs in the model. The present analysis revealed that the square root of each construct's AVE (i.e., the value shown in the diagonal in bold) is higher than the correlation between each construct and the rest of the model's constructs (0.696 > 0.360 and 0.256; 0.757 > 0.557 and 0.360; 0.848 > 0.557 and 0.256).
In addition, the heterotrait-monotrait ratio (HTMT) was analyzed by comparing it with a predefined threshold of 0.85 (Henseler et al. 2015). Based on the HTMT.85 criterion, all the model's variables show discriminant validity. Therefore, the results support the conclusion that all the constructs under study fulfill the established criteria for discriminant validity (Nunnally and Bernstein 1994) based on Fornell and Larcker's (1981) suggestions and the HTMT criteria (see Table  7).

Evaluation of Structural Model
After the measurement model was accepted as satisfactory, that is, the constructs' measures were found to be valid and reliable, the internal or structural model needed to be interpreted correctly. This was necessary to verify whether the model reflects the relationships between the latent variables suggested by the relevant theories (Wright et al. 2012) and formulated as hypotheses. The structural model evaluates the weight and magnitude of the relationships between the model's different variables. To this end, the model's predictive power had to be confirmed.

Model's Predictive Power: Analysis of R-Squared (R 2 ) for Each Dependent Construct
This type of model's fundamental objective is prediction. According to Chin (2010), structural models' predictive power can be evaluated by obtaining R 2 values for the dependent constructs. A model's goodness of fit is thus determined through each structural path's strength and analyzed based on the latent dependent variables' R 2 values (i.e., the variance explained). The criterion for which level of R 2 is appropriate depends on the authors consulted. For example, Falk and Miller (1992) consider the appropriate explained variance values to be those equal to or greater than 0.1. In the present study, the R 2 values calculated for the dependent constructs included in the structural model were PRAC R 2 = 0.130 and SAT R 2 = 0.314. The result thus confirm that the model presents adequate predictive power.

Predictor Variables' Contribution to Endogenous Variables' Explained Variance
Regarding path coefficients or standardized regression weights, these are generally interpreted as indicators of statistical relationships' relative strength. According to Chin (1998b), standardized path coefficients should have values that exceed 0.2-ideally, greater than 0.3. However, Falk and Miller (1992) are less demanding, proposing instead that, when one variable explains at least 1.5% of another variable's variance (0.15), this can be considered confirmation that the first variable has a predictive effect on the endogenous variable.
To calculate the variance in an endogenous construct explained by another latent variable, the path coefficient or beta (β) is multiplied by the corresponding correlation coefficient between both variables. In the present proposed model, the predictive constructs' exact contribution to the explained variance R 2 of each of the model's endogenous constructs was given as the absolute value, obtained by multiplying the path coefficient between two constructs by the value of the existing correlation between constructs (Falk and Miller 1992) (see Table 8).

Predictive Relevance of Dependent Constructs (Stone-Geisser test)
To measure the dependent constructs' predictive relevance, PLS uses the criterion of Stone-Geisser Q 2 , which is based on the redundancies that arise from the product of communities (λ 2 ) with the AVE indicator and that are cross-validated. The Stone-Geisser Q 2 test postulates that each reflective measurement model must be able to adequately predict the indicators of every endogenous latent construct in the model [93]. According to Hair et al. (2011), the Q 2 value can be obtained through the blindfolding process.
The test's results should be interpreted as follows (Chin 2010;Hair et al. 2011). If Q 2 > 0, the model has predictive capability. If Q 2 < 0, the model has no predictive capability. The present model's results include the following values: PRAC Q 2 = 0.068 and SAT Q 2 = 0.204. Based on Chin's (1998a) guidelines, this means these two constructs have predictive relevance, because positive Q 2 values were obtained.

Analysis of Paths' Statistical Significance (β) and Comparison of Hypotheses' Results
An analysis of the paths' significance was conducted to verify if the hypotheses formulated for the present study had empirical support. If all the β were found to be statistically significant, the research hypotheses were supported. A nonparametric resampling technique was used to carry out the analysis, namely bootstrapping, which produces the parameters' standard error and Student's tstatistic. To this end, a bootstrap test was conducted with 5000 subsamples and a Student's tdistribution with n − 1 degrees of freedom, in which n is the number of subsamples (Hair et al. 2011). Table 9 and Figure 3 present the results obtained, which reveal that two of the structural paths in the proposed model are statistically significant. Thus, two hypotheses, H1 and H3, received support based on positive effects, significant at the 0.001 level. The Student's t-statistic values show that the structural paths proposed are statistically significant except for the relationship between university practices and student satisfaction. H1 correctly postulates a strong relationship between students' participation and university practices, according to the β value of 0.360. According to H3, students' participation has a positive and significant influence on student satisfaction, although this relationship is less intense according to the β value of 0.064.

Model's Goodness of Fit
Henseler et al. (2014) use as their reference point the goodness of fit criteria based on covariance analysis. The cited authors developed and empirically validated a global goodness of fit measure applicable in research using PLS: standardized root mean square residual (SRMR). This measure represents the Euclidean distance between the empirical correlation matrix and the implied correlation matrix (Hu and Bentler 1999). Henseler et al. (2014) suggest that this measure's value should not exceed the maximum limit of 0.08, after which the model's general goodness of fit should be considered unsatisfactory. For the present study's model, the SRMR is 0.063, which is lower than the maximum recommended value, so the model's overall goodness of fit is satisfactory.

Conclusions, Limitations, and Future Lines of Research
USR is currently considered crucial to maintaining sustainability, so many universities have adopted this strategy as part of their initiatives. In this context, researchers have become aware of the need to develop measures that can help to improve many aspects of universities. Based on the present research's findings, evaluating students' perceptions can provide an opportunity to determine the ideal level of USR initiatives in the coming years.
The study's results corroborate that university students' participation in USR activities has an important influence on university practices, which is in alignment with relevant theories and much of the existing research (Alves and Raposo 2006;Bingham et al. 2005;Jongbloed et al. 2008;Larrán-Jorge and Andrades-Peña 2015). In addition, the present results confirm that university students' participation in USR activities has an impact on student satisfaction, a finding that is in line with the results reported by Gil et al. (2012), Gallardo-Vázquez and Sánchez-Hernández (2012) The current survey's findings thus reveal that students' perceptions are a significant factor in USR projects' development. Therefore, universities must continue working on strengthening their USR practices, especially in terms of student participation, which this research demonstrated has a positive influence on both university practices and student satisfaction. However, university practices do not positively influence student satisfaction, perhaps because of these institutions' lack of effective communication about their USR projects. This issue is especially important in public universities with a commitment to serving society at large (Moneva and Martín 2012).
In this context, higher education institutions' dedication to USR should include all training processes in their different modalities. Different types of training need to be taken into account in terms of objectives and content, as well as the various theoretical approaches that shape training programs' design and implementation. The results obtained also indicate that students know how to collaborate in USR activities, which mechanisms to participate in, and who benefits from these practices. Universities' initiatives affect both students and society by contributing to economic, political, and social development. In short, USR should be a primary objective for universities, in such a way that USR becomes the main axis of university policies, thereby complementing students' academic education with socially responsible practices (Molina et al. 2012).
With regard to the present study's limitations, the results obtained should not be generalized to other universities without due caution because the research context and cultural elements were almost certainly a decisive factor in the results. The initiatives implemented by the University of Extremadura may not be appropriate for other universities to emulate, and the surrounding culture may cause some aspects to be more influential than others in some contexts.
Nonetheless, this kind of study represents a first step toward extending universities' ability to incorporate significant actions that are part of their USR strategy (teaching, research, management and projection to society dimensions). Other universities would undoubtedly benefit from starting any evaluation of their USR practices by adapting the present questionnaire and taking this study's results into consideration. These can serve as good reference points, and the methods used can be enriched by introducing the particular variables that will allow other universities to use the proposed model to meet their specific needs.
A future line of research, to be conducted relatively soon, will involve the study of other university interest groups' perceptions of USR, specifically the administration, research and teaching faculty, and administrative and service staff. These three interest groups' perceptions should already provide enough information to help the University of Extremadura become fully aware of its current USR practices' effectiveness and the directions these should take in the immediate future. After analyzing the aforementioned groups' perspectives, research will also be carried out on stakeholders belonging to society at large. This future study will seek to verify whether society is aware of what the university is doing in terms of USR, as well as whether these stakeholders perceive any need for improvement and whether the University of Extremadura is meeting the present demands of society as a whole.
At the same time, other future work can be considered, such as expanding the sample with more Spanish universities and even foreign universities. Such a study would show how the social and cultural context can be decisive in the perception of the USR by students. Such a study would allow for greater generalization of the results and would contribute to the definition of commonalities in the recommended USRs for different countries and regions.