The aim of this study is to develop an explanatory model through empirical research and examine the impact of corporate reputation, social responsibility and student loyalty on student retention in private educational institutions. We adopted a non-experimental, cross-sectional, quantitative design because it allows the collection and analysis of data at a specific point and allows easy identification of relationships between variables without manipulating the research environment of the study (Hair et al., 2019).

Based on the literature previously mentioned, the following research questions were formulated:

Hypotheses were developed as shown in Figure 1.

The sample consisted of 409 students and graduates of Peruvian and foreign higher education institutions. Random convenience sampling was used to facilitate the data collection process, ensuring that participants were selected based on their immediate availability and accessibility. To mitigate possible biases, efforts were made to diversify the sample within logistical constraints, including data collection in different Peruvian and foreign university institutions and at different times of the year.

Of the participants, 44% (n = 180) were men and 56% (n = 229) were women; 37% (151) were students and 63% (258) were graduates. Regarding marital status, 39% (158) were single, 54% (213) were married, 26% (26) were cohabiting and 3% (12) were divorced. The mean age of the participants was 34.4 years, with ages ranging from 18 to 69 years. Regarding place of residence, 10% (42) were in Uruguay, 11% (43) in Ecuador, 8% (33) in Colombia and 71% (291) in Peru. As can be seen in Table 1.

The following measurement instruments were used.

Given that the loyalty scale was validated in English, it was back translated and adapted to the Peruvian context by means of two focus groups (Beaton et al., 2000). First, a certified translator oversaw the initial translation of the research instrument. Then, adjustments were made to ensure its adaptation to the local environment through focus groups conducted on two occasions, the first for the adaptation of the semantics and the second resulted in the validation of the modifications of the first one, leaving the questionnaire ready for application. The reliability of the scale, as well as its psychometric properties, was tested using the Partial Least Square PLS-SEMB (Hair et al., 2017).

Data collection was carried out using a Google Form, via email and social networks, such as WhatsApp. Once collected, the data were processed and subjected to statistical analysis using SPSS v.26 and AMOS v.24.

In the data analysis, Partial Least Square PLS-SEM was used to test the hypotheses. PLS- SEM is a comprehensive multivariate statistical analysis approach that includes measurement and structural components to simultaneously examine the relationships between each of the variables in a conceptual model, which has the characteristic of multivariate analysis, i.e., it involves a number of variables equal to or greater than three (Hair et al., 2013). In addition, PLS-SEM was employed in the present study because it facilitates theory building (Hair et al., 2017). WarpPLS version 8.0 was used to perform the PLS-SEM analysis. This software was used because according to Kock (2014), WarpPLS provides options to use different algorithms for external and internal models in the calculation of latent variable scores, such as the path coefficient and the parameters associated with the p-value, identifying and considering nonlinear relationships in the structural model.

To assess the quality of reflective constructs, the convergent validity and reliability of the construct, i.e., internal consistency, must be assessed (Chin, 2010; Hair et al., 2011; Kock, 2015), and the indicators presented must be met (see Table 2).

Table 3 shows that all the indicators evaluated are met. First, all loadings exceed the value of 0.7. Additionally, the construct shows good results overall, as both Cronbach’s Alpha and CR exceed the 0.7 threshold. Similarly, the AVE also meets the requirements, with all values being above 0.681. In addition, the values of the VIF Total Collinearity are below 3.632, which is within the acceptable range. Therefore, since all indicators are satisfactory, we proceed to the discriminant evaluation.

Discriminant validity provides an indication of the extent to which each construct is distinct from other constructs in the model (Chin, 2010). To meet discriminant validity the square root of the AVE for each construct must be greater than the highest correlation between the construct and other constructs in the model (Chin, 2010; Hair et al., 2011; Kock, 2014). Table 3 indicates that the square root of the AVE for all constructs is higher than the correlation with the other constructs, suggesting that the model exhibits adequate discriminant validity.

To evaluate the structural model, it is essential to verify and report two preliminary criteria: the significance of the path coefficients and the value of the R² coefficient for the endogenous constructs. Each hypothesis is linked to a causal relationship in the structural model, representing the interactions between pairs of constructs. Path coefficients have been calculated for each relationship in the model, along with their respective p-values. While the path coefficients should be significant, the value of the R² coefficient is highly dependent on the area of investigation. Chin (1998) suggests values of 0.67, 0.33 and 0.19 as, respectively, substantial, moderate and weak measures of R. In behavioral studies, a value of 0.2 for R² is generally considered acceptable (Hair et al., 2014; Kock, 2013).

The findings indicate a significant positive effect of SR on LY (β = 0.2; p < 0.01), a significant positive effect of SR on RT (β = 0.12; p < 0.01) and a significant positive effect of SR on IR (β = 0.53; p < 0.01). This shows the importance of the social responsibility of higher education institutions in student retention and loyalty, as well as to improve their reputation in the community.

The findings indicate a significant positive effect of LY on RT (β = 0.77; p < 0.01). In the global model, this is the relationship that presented the largest effect size indicated by the standardized β coefficient with a high statistical significance, which indicates that a higher student retention will be obtained in those university institutions whose students present a high loyalty.

The results indicate that institutional reputation exerts a small but statistically significant effect on student retention among university students (β = 0.02; p < 0.38). This finding suggests that a university’s reputation influences students’ decisions to continue their enrollment at the institution. Although the effect size is modest, its statistical significance positions it as one of the contributing factors to student retention.

In this study, the R² for the IR, LY and RT coefficients were 0.28, 0.48 and 0.70, respectively. Therefore, all R² values had relatively high and acceptable values. The values in this study suggest that the variables (SR, IR, and LY) account for a high percentage of the variance in RT. Table 4 and Figure 2 present the results of hypothesis testing and evaluation of path coefficients. For the overall model fit index, the six goodness-of-fit indices (Kock, 2014) were considered, with a confidence level of 95%, the efficiency indices are as follows:

In the case of the present study the six fit indices suggested that the model fit was more than acceptable: average path coefficient (APC) = 0.368, p < 0.001; average R2 (ARS) = 0.488, p < 0.001; average adjusted average R2 -squared (AARS) = 0.486, p < 0.001; average block variance inflation factor (AVIF) = 1.645 (acceptable if ≤5, ideally ≤3.3); average full collinearity variance inflation factor (AFVIF) = 2.541 (acceptable if ≤5, ideally≤3.3); and Tenenhaus GoF (GoF) = 0.629 (small ≥0.1, medium ≥0.25, large ≥0.36). The predictive validity of a construct can be confirmed when the value of the associated R² coefficient is greater than zero. In this study, all the values of the endogenous variables in the model meet this criterion, suggesting acceptable predictive validity for the entire model.

According to the results of the first hypothesis (H1), social responsibility influences the loyalty of students in private educational institutions, these results are in agreement with those of Latif et al. (2021) who showed that university social responsibility (USR) positively predicts student loyalty, with perceived service quality, student satisfaction and trust in the university as key mediators. Similarly, Mostafa and Hamieh (2021) found that SR activities have a significant impact on students’ attitudinal and behavioral loyalty in the Lebanese education sector.

Also, Suranta and Rahmawati (2024) highlighted that higher education image and service quality measured the effect of SR on student loyalty in Indonesia. In the same vein, Wong et al. (2023) added that an institution’s brand reputation acts as a mediator between SR and student loyalty, suggesting that a positive perception of SR initiatives enhances loyalty through the institution’s reputation. So also, (Hossain et al., 2019). Research on the impact of CP on the health sector has shown that CP practices increase patient satisfaction and loyalty. These results are consistent with the notion that SR practices can increase loyalty among various stakeholders, including students at private institutions. The results of this study show that social responsibility plays an important role in motivating students in private educational institutions.

Social responsibility influences student retention in private institutions. There are several studies that have examined factors associated with student retention in higher education, such as (Thai and Tho, 2024) study that examined the impact of university service quality, student satisfaction, and change resistance on student privacy. Higher education institutions. Emphasizes service quality and student satisfaction in student retention. Similarly, Xuerong and Kanjanapathy (2024) indicated that the financial burden on students in higher education institutions indicated that student tuition fees and fees are the main cost of these institutions. Similarly, Al Hassani and Wilkins (2022) discussed the impact of organizational identification and institutional reputation on student satisfaction and behavior, noting that these factors play an important role in student retention. In terms of social responsibility, Nirmalasari et al. (2024) emphasized the social and emotional power of students to increase their concentration, attendance and learning ability. This suggests that creating a socially responsible environment in educational institutions can have a positive impact on student retention.

On the other hand, the results indicate states that social responsibility influences the institutional reputation of private educational institutions. This is consistent with the study by Rasoolimanesh et al. (2023) who found that demonstrating social responsibility practices not only contributes to the welfare of society, but also enhances the image and reputation of higher education institutions. In the same vein, Azeem et al. (2019) identified a positive relationship between the social responsibility of institutions and their reputation, indicating a strong link between these two factors. Besides, Chen et al. (2020) argued that maintaining social responsibility requires building and maintaining external social capital, which improves corporate reputation. In addition, Taamneh Mohammad et al. (2022) found various corporate social responsibility (CSR) practices, including ethical, philanthropic, and legal aspects, significantly improve the reputation of universities, especially in developing countries such as Jordan. Furthermore, other authors believe that the implementation of social responsibility in universities and involvement in student misconduct and humanitarian activities helps to improve the institution’s reputation (Berei, 2020; Jie and Huam, 2019). Furthermore, Prodanova et al. (2021) emphasized that perceived organizational support has a significant impact on student satisfaction, which affects the reputation of the institution.

The research results support that student loyalty affects student retention (Bakrie et al., 2019; Lee and Seong, 2020; Ong et al., 2023; Sugiharto et al., 2021; Wang et al., 2024; Xuerong and Kanjanapathy, 2024). These authors analyzed this relationship in detail and concluded that the relationship between the factors was associated with student loyalty and subsequent student retention. Likewise, in higher education, factors such as the quality of student instructors, institutional involvement, and reputation play an important role in increasing student loyalty (Bakrie et al., 2019; Snijders et al., 2020). Likewise, the commitment-trust theory positively supports student loyalty, promoting the development of strong relationships and a positive institutional image (Bakrie et al., 2019; Snijders et al., 2020). Furthermore, an institution’s reputation, along with service quality and student satisfaction, has a significant impact on student loyalty and retention (Bakrie et al., 2019; Sugiharto et al., 2021). The financial burden on students in the country is high, as tuition fees are the primary source of revenue for private institutions (Xuerong and Kanjanapathy, 2024). Thus, strategies to increase student loyalty by improving service quality, satisfaction, and engagement are critical to the sustainability and success of these institutions (Jie and Huam, 2019; Wang et al., 2024). Furthermore, the recognition of training and development programs for teachers and staff as a tool to enhance satisfaction, loyalty, and retention underscores the importance of investing in human resources to enhance the overall student experience (Mampuru et al., 2024).

This result is supported by Bakrie et al. (2019) who found in their study that there is a statistically significant relationship between university reputation and student loyalty, emphasizing the role of the reputation of institutions in fostering student loyalty. Besides, Wong et al. (2023) argue in their findings that institutional reputation is an influence on student loyalty, as well as satisfaction and service quality. Undoubtedly, Mulyono et al. (2020) found that reputation is an important predictor of loyalty in higher education, emphasizing its mediating role in shaping student loyalty. These results support the idea that an institution’s reputation plays an important role in shaping student loyalty. Indeed, Hasan and Hosen (2022) emphasized the relationship between university reputation and service quality, satisfaction, and student loyalty.

Finally, the findings shown that institutional reputation influences student retention in private educational institutions. Various studies carried out support this rejection since the important factors for student retention are service quality, student satisfaction and loyalty, and not only institutional reputation (Al Hassani and Wilkins, 2022; Forid et al., 2022; Moslehpour et al., 2020; Tan et al., 2022). Although institutional reputation can influence student loyalty (Raja, 2023), it does not determine student retention, suggesting the involvement of other intervening variables in this relationship. Furthermore, higher organizational reputation could instill high expectations in students, which they might not meet, resulting in higher dissatisfaction and lower organizational retention (Sears et al., 2017).

Private higher education institutions must retain recruited students in order to maintain their institutional budgets (Le et al., 2021). In addition, it has been highlighted that low student retention can have a negative impact on the reputation of the institution (Adeola Atobatele et al., 2024). However, studies have shown that student satisfaction, service quality, and barriers to change mediate the relationship between service quality and student behavioral intentions in private universities (Tan et al., 2022). So also, Forid et al. (2022) suggest that institutions that fail to ensure student retention jeopardize their reputation, financial gains, and possibly their survival. On the other hand, Moslehpour et al. (2020) indicate that student satisfaction mediates the influence of service quality on institutional reputation. Likewise, Varol and Catma (2021) and (Wade, 2019) noted that institutional selectivity and governance structures are also significant predictors of student retention.

The research expands the comprehension of quality management theory (Grönroos, 1984), incorporating variables pertaining to educational quality, such as social responsibility, institutional reputation, and student loyalty as significant factors in the retention of university students. This approach offers a more comprehensive perspective on the dynamics that influence student retention. The findings substantiate the theoretical significance of social responsibility, not only as an ethical practice but also as a strategic factor that directly impacts student loyalty and retention. This suggests the necessity to integrate social responsibility more prominently into theoretical models of university management. Furthermore, the study elucidates a more complex relationship between institutional reputation and retention than previously postulated, challenging certain prior theoretical assumptions and indicating the need to re-evaluate existing models of institutional reputation within the context of higher education.

The mediating role of student loyalty between various institutional factors and retention emerges as a significant theoretical finding. This contributes to a more nuanced understanding of how student retention is developed and maintained, underscoring the importance of considering loyalty as a central construct in retention models. These theoretical insights provide a solid basis for future research and the development of more accurate models in the field of higher education management.

Educational institutions should prioritize and integrate USR practices into their overall strategies, not only as an ethical obligation, but as an effective tool to improve student loyalty and retention. It is crucial to implement specific programs to cultivate student loyalty, such as student experience improvements, mentoring programs, and opportunities for participation in the university community. Effective communication about social responsibility initiatives and other factors that influence reputation and loyalty is critical, making sure students are aware of and value these efforts. Take a more holistic approach to their retention strategies, considering the interplay between MSW, reputation, loyalty, and other institutional factors. In addition, it is crucial to train academic and administrative staff on the importance of SR and its role in fostering student loyalty and retention. In addition, institutions can develop more personalized services and programs that address the specific needs of their student population, based on an understanding of the factors that influence loyalty and retention.

It is known that selecting a convenience sample comes with limitations, including a lack of representativeness and potential bias in participant selection. To minimize these effects, we implemented the following steps: We ensured quality control in both data collection and analysis by providing a variety of data collection centers by institution and country. Moreover, the cross-sectional nature of data collection can be a limiting factor in establishing causal relationships between variables (Finkel, 1995). Future studies should conduct longitudinal investigations of these variables. Because the variables are self-reported measures, the results may be biased (Podsakoff et al., 2003). This comprehensive study lays the groundwork for future research on the influence of various factors. Future research would benefit from expanding the analysis to include more institutional and cultural contexts, as well as using longitudinal methods to better understand how the influence of these factors changes over time. We also recommend examining potential mediators or moderators of the relationship, including quality of education, institutional support, and socioeconomic factors.