Student Retention at a Higher Educational Institution: A Comparison Between Evaded and Nonevaded Groups Using Partial Least Square

Abstract

Students’ desertions can be a threat to the proper functioning and management of a Higher Educational Institution. Based on the theoretical model named the Relationship Quality-based Student Loyalty that combines relationship marketing to educational studies retention constructs, the main objective of this study is the evaluation of the Relationship Quality-based Student Loyalty constructs in a Brazilian Higher Educational Institution reality and the identification of significant differences when evaded and nonevaded students groups are compared. A survey was carried out with 314 undergraduate students along with their secondary data. Data analysis was performed using Partial Least Square technique to test the model validity, Partial Least Square–Multiple Group Analysis to compare groups, and Mann–Whitney U test to identify significant differences in the parameters. Six of the eight hypotheses were entirely or partially confirmed. Those results can help educational managers to adopt a different positioning about student retention factors, in attempt to improve commitment and perceived quality.

Keywords

student retention higher education evasion factors perceived quality commitment

To meet the demand for higher education in Brazil, new Higher Education Institutions (HEIs) have been established since 1997, as well as public and private sector courses. The number of organizations jumped from 900 (1997) to 2,537 (2018), an increase of 282% (Instituto Nacional de Estudos e Pesquisas Educacionais Anísio Teixeira [MEC/INEP], 2018). According to the Higher Education Census (INEP), the country reached a significant milestone of 8.4 million students enrolled in undergraduate courses, in 2018. The growth registered by the last census may be related to the expansion of counties offering undergraduate courses, as well as the expansion of the Federal Institutes of Education, Science and Technology (IFs) network and their internalization into the Brazilian municipalities, which registered a total of 197,000 enrolled students in 2018. At the regional level, the Goiano Federal Institute enrolled more than 25,000 students in 2018, from which more than 4,500 are at the undergraduate degree at 13 units located in the counties of the Goiás State.

Faced with growing competition in the educational sector, the IFs, despite the fact they are nonprofit organizations, should compete with the most diverse educational organizations and demonstrate accountability as the financial resource they receive from the government is based on the number of enrolled students. Adopting customer relationship management strategies in the educational sector, aiming to know the target market, is an essential way to achieve goals such as attracting, maintaining, and building relationships with society. As in the case of business organizations for goods and services, student retention is necessary such as customer retention, since the cost of capturing new customers or students and the loss of revenue from desertion, as well as idleness, can weaken the results of an institution.

Research Problem and Objective

The central problem in this research is the evasion of undergraduate students from the IF Goiano’s face-to-face courses, especially during the 1st year. This fact can drive public resources into a garbage can. In this sense, we adopted a theoretical model named the Relationship Quality-based Student Loyalty (RQSL) that combines relationship marketing factors arising from customer retention theory—Perceived Quality, Trust, Commitment, and Loyalty—with constructs emerged from students retention educational studies—Academic and Social Integration—thus seeking a combined approach between marketing and education. Therefore, the main objective of the study is the evaluation of RQSL constructs in a Brazilian HEI reality to understand the interaction between the constructs of the conceptual model and to obtain the validation metrics of those relations, identifying the best antecedents for student’s retention. In addition, we perform a comparison between evaded and nonevaded student groups to identify significant differences when using this approach.

Brazilian Higher Education Recent History

In 1996, through the new regulation named LDB-Lei das Diretrizes e Bases da Educação, the education market expansion was born. Since this moment, the implementation of new institutions in public and private spheres has been improved (Carneiro, 2012). From 2003 to 2018, the total number of enrollments in undergraduate courses more than doubled, from 3.9 million to 8.4 million (MEC/INEP, 2018). Despite this significant growth, dropout rates at public and private universities in Brazil continue to rise, thus, creating possible concerns for the HEIs.

There were high dropout rates in Australia, the United States, and South Africa, most of them in the early years (Weng et al., 2010). Ireland and England are the countries with the lowest rates of evasion in the European continent, and Japan is the country with the lowest level of evasion in the world (Furtado & Alves, 2012). In the case of an emergent country such as Turkey, voluntary and involuntary departures from college bring an immense cost to the efforts of expanding its higher education participation (Aypay et al., 2012).

Based on INEP data, in 2010, 11.4% of students dropped out of the course they were admitted to. In 2014, this number reached 49%. In 2017, a dropout from on-campus undergraduate courses in Brazil fell slightly from 27.2% in the previous year to 25.9%. It also decreased in On-Line mode (Educação à Distância ) from 36.1%, in 2016, to 34.3% in 2017 (Sindicato das Entidades Mantenedoras de Estabelecimentos de Ensino Superior no Estado de São Paulo [SEMESP], 2019). Despite this slight percentage reduction, the evasion rate is about one third of the total enrollments, revealing the reduced efficiency of the Brazilian educational system. There are also 107,000 vacancies remaining in federal institutions. Considering all the categories of public HEIs, the number rises to approximately 178,000 places (Portal Brasil, 2016).

Thus, desertions can be a threat to the proper functioning and management of a specific course or institution (Bôas, 2008), highlighting the importance and necessity of managing the relationship with students, from their initial contact to graduation date (Fontaine, 2014). Understanding the evasion factors to devise student retention strategies becomes essential to achieve satisfactory rates in HEIs, as in any business organization from other segments.

Theoretical Framework

The theoretical framework for this study is drawn from four related research themes: student evasion factors; customer and student retention; perceived quality, trust, and commitment; and academic and social integration.

Students Evasion Factors

Researchers from all over the world propose to study students evasion and seek to approach it from a variety of perspectives such as social (Aypay et al., 2012), behavioral (Arce et al., 2015; Schargel & Smink, 2002), economic (Cabrera et al., 1992; Weng et al., 2010), and expectations’ frustration up to the 1st year (Grebennikov & Shah, 2012). Some of them use computer systems (Tontini & Walter, 2011), or strategies related to marketing and consumer behavior (Ackerman & Schibrowsky, 2007; Fontaine, 2014; Kotler & Fox, 1994; Masserini et al., 2019; Schlesinger et al., 2017). Tinto (1975, 2001, 2006) developed an integrated model called Student Integration Model, that relates constructs, such as academic and social integration, to the student retention, signaling the level of commitment and possibly, driving evasion.

For Tinto, the students’ retention becomes high when there is a natural adaptation to the university environment and sufficient integration of the student. These intentions can set the level of commitment to the course completion goals. The higher the level of commitment, the greater the likelihood of completion (Tinto, 1975, 1993, as cited in Vasconcelos et al., 2009). Although the qualities that students bring upon joining the HEIs are important, retention depends mainly on what happens after admission (Ackerman & Schibrowsky, 2007).

Cabrera et al. (1992) identified in American students a direct effect of finances concerning the decisions of remaining, besides the influence in social integration. Schargel and Smink (2002) related abandonment to the accumulation of negative experiences. Ackerman and Schibrowsky (2007) identified graduation rates in the United States as close to 50%. They emphasized that the permanence is linked to the service received and the relationships experienced by the student, that is, building strong relationships leads to loyalty (Schlesinger et al., 2017). In addition to individual aspects, evasion can be related to academic management, curricular issues, and the prestige of the organization (Albuquerque, 2008; Aypay et al., 2012). For 1st-year students, the route to improving retention rates lies in increasing student learning skills and academic efficacy (Ryan & Glenn, 2002).

At the Brazilian national level, since 1995, the topic of evasion in public HEIs have entered into the agenda of the Federal Government. There was a study that identified possible evasion factors, considering individual, internal, and external aspects (Secretaria de Educação Superior [MEC/SESU], 1997). Due to the relevance of the matter, evasion has already been audited by the Federal Audit Court (TCU) with a focus on the Professional, Scientific, and Technological Education Federal Network (Tribunal de Contas da União [TCU], 2012). In response to the recommendations, in 2014, the Secretary of Education issued guidelines to strengthen educational action, encouraging actions that increase students’ chances of remaining at school and success, focusing on the quality of teaching, and attending to diversity (Secretaria de Educação Profissional e Tecnológica [MEC/SETEC],2014).

Silva Filho et al. (2007) reported a 22% national average rate for all HEIs in Brazil, 12% in public, and 26% in private schools. In a recent study, Ambiel et al. (2016) addressed evasion in Brazilian states by investigating students from public and private institutions and found the absence of family support and the financial question as to the main reasons for dropouts. Faced with evasion factors, to manage relationships, one must first understand who the students are, what they want, and how to serve them effectively and efficiently (Fontaine, 2014). Thus, to strengthen the intentions to remain, until the course conclusion, it will depend on the service provided and, on the relationships, experienced overtime (Ackerman & Schibrowsky, 2007).

To classify the factors and variables of the evasion into structural groups (MEC/SESU, 1997)—Individual, Internal, and External—Table 1 was elaborated in a nonexhaustive but summarized approach.

Table 1.
Evasion Factors Groups.

Subfactors Variables References

Individual factors Learning Educational backgroundPreparation for the courseSkillsLimitations failures Furtado & Alves (2012); Lobo (2012); Albuquerque (2008); Silva Filho et al. (2007); Schargel & Smink (2002); Ryan & Glenn (2002); Tinto (2001); MEC/SESU (1997); Tinto (1975)

Relationship MaturityCommitmentSocial and academic integration Arce et al. (2015); Lobo (2012); Aypay et al. (2012); Schargel & Smink (2002); Tinto (2001); MEC/SESU (1997); Tinto (1975)

Individual lifetime Personal lifeProfessional lifeAgenda and appointments Furtado & Alves (2012); Tontini & Walter (2011); MEC/SESU (1997)

Internal factors Infrastructure LocationClassroomsLibraryConservation Ambiel et al. (2016); Lobo (2012); Tontini & Walter(2011)

Education quality Subjects/course requirementsFinancial aidService and supportRecommendation Tontini & Walter(2011); MEC/SESU (1997)

External factors General lifetime Marriage/childrenAddress changeFamiliar referenceEngagements Ambiel et al. (2016); Tontini & Walter (2011); Albuquerque (2008); Schargel & Smink (2002); Tinto (1975)

Finance Personal incomeFamily supportEmployer incentives Ambiel et al. (2016); Arce et al. (2015); Furtado & Alves (2012); Lobo (2012); Tontini & Walter(2011); Silva Filho et al. (2007); Schargel & Smink (2002); MEC/SESU (1997); Cabrera et al. (1992); Tinto (1975)

Professional Expectations/uncertaintiesProfessional vocationPersonal and professional ascensionEmployability Tontini & Walter(2011); Weng et al. (2010); Albuquerque (2008); MEC/SESU (1997)

Source: Author.

In the public sector, the importance of understanding the students’ evasion mainly for federal educational institutions is that the phenomenon causes public money losses, by the time vacancies are not filled in (Palharini, 2010). In this sense, the competition in the market has forced HEIs to adopt a more targeted customer philosophy (Kotler & Fox, 1994; Masserini et al., 2019; Schargel & Smink, 2002). They should look for what is perceived as an advantage by the customer, such as studying in an institution that offers internships, practical classes, complete laboratories, status, and recognition by their peers (Bôas, 2008), besides a robust institutional image (Masserini et al., 2019; Schlesinger et al., 2017).

So, actions to acquire, satisfy, and retain students are the most significant challenge for the higher education sector (Abubakar & Mohd Mokhtar, 2015; Bowden, 2011). Kotler and Fox (1994) emphasize key activities to adopt soon at the student’s entrance: designate a commission to manage retention, encourage teachers and staff to take the attitude of serving students, and promote the feeling of belonging to the HEIs. Besides counseling and tutoring services, permanent grants, and students advising increase the perception of value and strengthen retention (Tinto, 2001).

Customer and Student Retention

Although attracting new customers continues to be relevant, the focus of the organizations has been changed to relationship marketing, which seeks to create, maintain, and improve customer relationships (Aurier & N’Goala, 2010; Rosa, 2001). It is also observed that customer evasion, the opposite movement to retention, is very expensive to corporations due to the new customer acquisition high costs (Zeithaml et al., 2014). Understanding and exploring the relationship between provider and client are seen as a successful mechanism to establish itself in the market and to extend existing relationships (Aurier & N’Goala, 2010; Lin & Wu, 2011). The loss of customers is estimated between 15% and 20% per year (Reichheld & Sasser, 1990). With the desertion of customers, there is a need to attract new customers, which implies a range of high costs including advertising, promotions, time, sales resources, and operational expenses, and that may reflect in the loss of all consumption that customers could have throughout their life (Reichheld & Sasser, 1990).

Relationship marketing can be easily adapted for use in HEIs (Masserini et al., 2019; Schlesinger et al., 2017; Subrahmanyam, 2017) to guide student retention initiatives (Ackerman & Schibrowsky, 2007; Bergamo et al., 2011) and assist management in decision making (Bowden, 2011). In educational organizations, retaining enrolled students is as important as attracting and enrolling them (Kotler & Fox, 1994). A relationship with students, following the same context of regular customers, emphasizes the importance of promoting an interactive channel between institutions and students (Bowden, 2011; Chen, 2017).

Consumers generally remain in the relationship when what they obtain (quality, satisfaction, and specific benefits) exceeds what they give (monetary and nonmonetary costs; Zeithaml et al., 2014), creating a high satisfaction that can lead to establish a long-term relationship (Reichheld et al., 2000; Zeithaml et al., 2014). Customer loyalty is motivated by the interaction between relational benefits and the quality of the relationship (Bergamo et al., 2011; Hennig-Thurau & Klee, 1997; Hennig-Thurau et al., 2001). Satisfied students can attract new students through positive word-of-mouth communication, which promotes a competitive advantage for educational institutions (Subrahmanyam, 2017).

Nowadays, this can also be explained by a new marketing field of study, named customer experience (CX) that reflects many of those constructs. The CX in their journey with the company can bring satisfaction and loyalty or disappointments that lead to a situation of desertion in the future. From several definitions of CX retrieved in the literature, the one that most explains the term’s complexity describes it as “comprised of cognitive, emotional, physical, sensory, spiritual and social elements” that point out the customer’s direct or indirect interactions with the relationship company (De Keyser et al., 2015, p. 1).The CX is the customer’s internal and subjective reaction to any direct or indirect contact with a company. In general, direct contact occurs during the purchase, use, and service and is usually initiated by the customer. Indirect contact often involves unforeseen encounters with products, services, or brands of the company and based on word of mouth, advertising, press reports, evaluations, and others (Meyer & Schwager, 2007).

In a distinct line of thought, it can be said that the CX has antecedent elements, such as satisfaction and perceived quality, as well as consequent elements, such as commitment and trust. In practical terms, the CX starts at the prepurchase moment, goes through the purchase up to the postpurchase, interactively and dynamically, incorporating past impressions and external factors (Lemon & Verhoef, 2016). The company that manages complete journeys reconciles the vision of individual transactions with the understanding of the root causes of interactions, improving them through recursively given feedbacks, which could contribute to unequivocal rewards such as increased satisfaction and revenues, reduced churn, and higher employee satisfaction (Rawson et al., 2013). For the repurchase or the future use of a service, it is crucial that the customer feels satisfied and that there is a sense of trust and perceived value about the provider and their services (Lin & Wu, 2011; Zeithaml et al., 2014).

Everything said in terms of a regular customer relationship with a company can be inferred to a student as a customer that relates to an HEI. When the student is analyzing the intention of leaving the course he is enrolled, he should consider the costs of change such as incompatibility of curricula between HEIs, time spent in studying complementary subjects, financial expenses, among others. Customer orientation has been considered something new in universities when compared with commercial organizations (Fontaine, 2014), but an important initiative to achieve a sustainable competitive advantage (Bowden, 2011). Thus, HEI has been working on relational marketing, with a focus on customer retention, maximizing the customer or student experience, as well as the importance of addressing aspects such as the perceived quality of services rendered, commitment and trust, among other constructs (Abubakar & Mohd Mokhtar, 2015; Bergamo et al., 2011; Bowden, 2011; Chen, 2017; Masserini et al., 2019; Schlesinger et al., 2017).

Perceived Quality, Trust, and Commitment

Perceived quality is defined as the judgment of the consumer on the excellence or superiority of a product/service, which means a perception based on impressions and previous experiences (Anjos Neto, 2003; Chenet et al., 2010). Service Quality for College students can be related to the evaluation of the quality regarding teaching staff, infrastructure (library, equipment, physical facilities of rooms, and laboratories), educational methodology, resources and facilities, learning evaluation and goals achievement (Anjos Neto, 2003; Hennig-Thurau et al., 2001), and lectures and courses offering (Masserini et al., 2019). As important as learning environments, students can consider other facilities as essential such as cafeteria, parking, sports gym, among others (Fontaine, 2014). In the university environment, student motivation can be a major factor in assessing quality (Subrahmanyam, 2017).

In light of the above on perceived quality, there are three hypotheses to be tested:
Hypothesis 1 (H1): Perceived Quality has a positive influence on student Trust in HEI.

Hypothesis 2 (H2): Perceived Quality has a positive influence on student Commitment.

Hypothesis 3 (H3): Perceived Quality has a positive influence on student Loyalty.
As fundamental as perceived quality, another key element is trust. Doing business with people we trust is more predictable, efficient, and profitable (Reichheld et al., 2000), narrowing the interactions and lowering transaction costs (Hennig-Thurau & Klee, 1997). Trust is when one relies on the exchange partner, what means the genuine trust that one partner can count on the other, implying the behavioral intention to trust (Morgan & Hunt, 1994), and it is an essential requirement for relational exchanges (Garbarino & Johnson, 1999; Lin & Wu, 2011; Morgan & Hunt, 1994). Trust, in the long run, favors the relationship, since it reduces the transaction costs associated with the relationship between customer and company (Chenet et al., 2010; Moreira & Silva, 2015).

Trust is a factor that maximizes the likelihood of college students remaining in the institution, especially when they realize that the organization is working for their interests (Ackerman & Schibrowsky, 2007). This construct is measured through the employee and organizational dimensions, which in the HEI translates into compliance with promises, policies, and rules agreed with the students (Anjos Neto, 2003). It is also noteworthy that students who do not have confidence are likely to demonstrate a low level of learning (Chen, 2017). It is created and strengthened by positive and satisfactory evaluations, influencing, and making more predictable future consumption in the established relationship (Aurier & N’Goala, 2010).

For Hennig-Thurau et al. (2001), student trust in the university is based on the integrity and reliability of actions and behaviors, based on the personal experience each student has with faculty members, for example, the punctuality of teachers, the content of the examinations, and excellent communication. Trust is a construct that can be assumed as a direct antecedent of student loyalty. The recommendation of the institution and the evaluation of its reputation regarding the image can also be added (Dziminska et al., 2018). Given those considerations, the following hypotheses are presented:
Hypothesis 4 (H4): Trust in HEI has a positive influence on Commitment.

Hypothesis 5 (H5): Trust in HEI has a positive influence on Loyalty.
The increase in the efficiency of a relationship is also related to a high level of commitment, which is characterized when there is both a rational connection (an increase of net benefits) and an affective bond (emotional ties) with the relationship (Bowden, 2011; Hennig-Thurau & Klee, 1997). The commitment is a central construct in the relations with the consumer, identified by the client’s psychological attachment, their identification with the organization, which signals their concern for the future well-being of the company (Aurier & N’Goala, 2010; Garbarino & Johnson, 1999), expressed by the continuous use of the service and the recommendation to others (Moreira & Silva, 2015). However, there will only be a commitment in the relationship when it is considered substantial (Ackerman & Schibrowsky, 2007; Morgan & Hunt, 1994). The commitment has several dimensions such as affective (preference for the company, emotional ties, and friendship); continuity or calculation (costs of change and benefits if the relationship is maintained; Bergamo et al., 2011; Bowden, 2011), normative (duty to remain); instrumental (dealing with stimuli, prizes, and punishments); and behavioral (perception of the supplier in assisting the exchange partner; Bergamo et al., 2011).

In higher education, students’ commitment to HEIs is a construct that is determinant of student loyalty (Bowden, 2011; Hennig-Thurau et al., 2001). Educational services have specific characteristics such as long-term nature, the need for active coproduction and learning, and several external challenges that may interfere with the student’s goals (Chen, 2017). Students who perceive a substantial and mutual commitment between themselves and the HEIs are more likely to remain and recommend it to friends (Ackerman & Schibrowsky, 2007). Given the aspects concerned with commitment, we have the hypothesis:
Hypothesis 6 (H6): The student’s Commitment to the institution has a positive influence on Loyalty to the institution.

Academic and Social Integration

In the last decades, studies related to customer loyalty have contributed to the literature in several aspects and markets including the loyalty/retention view of college students (Anjos Neto, 2003; Basso et al., 2015; Bergamo et al., 2011; Hennig-Thurau et al., 2001). Most of them use Hennig-Thurau et al.’s model, which integrates a quality model of the relationship with Tinto’s model (1975) and emphasizes aspects of the student including academic and social integration that can influence the commitment to HEIs (Tinto, 1975, 1993, as cited in Vasconcelos et al., 2009). In this sense, academic and social integration is approached as an influencing antecedent factor, since this construct is significant for the student’s emotional commitment. This approach was not found in the Federal Institutes of Education (FIs), which justifies its application. Thus, high academic and social integration can positively affect the student’s commitment to HEIs and, consequently, their loyalty. Therefore, we have the following hypothesis:
Hypothesis 7 (H7): Academic & Social Integration have a positive influence on the student’s Commitment to HEI.
Probably, the perceptions of evaded students and students in the course at the HEIs are not the same. One of the differentiating factors could be the level of interaction and relationship between students and members of the institution (teachers, technical administrative, and other course colleagues). Bearing in mind that there are several different behaviors of variables between groups, demonstrated in the nonparametric Mann–Whitney tests, it is suggested that the whole model should be tested between groups (current and dropout students), according to the following hypothesis:
Hypothesis 8 (H8): The proposed model has differences in path coefficients between the groups of evaded and non-evaded students.

Method

Sample

The unit investigated is the Federal Institute of Education, Science and Technology Goiano, located in a rural region, 18 kilometers from Morrinhos-GO. It is a school-farm and offers public and free education, from high school to postgraduate courses. The target population consisted of 596 undergraduate students in progress (Group 1), and 241 undergraduate students evaded between 2012 and 2016 (Group 2). The sample consisted of 209 students in progress, and of 105 evaded, totaling 314 participants. The identification of the evaded students occurred with the aid of the secretary of school records.

Primary data were collected through a survey applied to both groups of students. The first group answered the questionnaires face-to-face at the Morrinhos Campus. The second group (evaded students) was not at this school anymore. So, the survey was applied by electronic media. Secondary data about all the students were collected directly from the HEI records.

Instrument

This research is classified as a cross-sectional study, whose data collection was concluded in 2017. Descriptive quantitative research was carried out to reach the objectives, considering the conceptual and theoretical model described in the previous hypothesis. In the surveys, the same script was used, adjusting the question statements according to the investigated group.

The questionnaire was divided into two parts. The first one sought sociodemographic and academic information, and the second one was based on the Hennig-Thurau et al. (2001) variables, aiming to identify the student’s loyalty through the constructs Perceived Quality, Trust, Commitment, and Academic and Social Integration. Table 2 shows the proposed items and their respective constructs.

Table 2.
Proposed Questionnaire Items.

Questionnaire Items (variables)^a Constructs

Q12. Como você avalia, em média, a competência dos professores. Q13. Como você avalia a infraestrutura da instituição na qual estuda. Q14. Opinião sobre a orientação dada pelos professores. Q15. Critérios de avaliação adotados em seu curso. Q16. Oferta de serviços oferecidos da instituição. Q17. Oferta de lazer e descanso da instituição. Q18. Classificação dos resultados que você está obtendo com o curso. QUALIDADE PERCEBIDA (Perceived Quality)

Q19. Eu recomendaria o meu curso para outra pessoa. Q20. Eu recomendaria minha instituição para outra pessoa. Q21. Tenho muito interesse em manter contato com meus professores. Q22. Se eu tivesse que tomar a decisão novamente, escolheria o mesmo curso. Q23. Se eu tivesse que tomar a decisão novamente, escolheria a mesma instituição. Q24. Eu me tornaria membro de algum tipo de grupo de ex-alunos ligado a instituição ou a docentes. LEALDADE (Loyalty)

Q25. Integridade é uma palavra que eu usaria para descrever os funcionários da instituição. Q26. Estou certo de que os funcionários (técnicos-administrativos) sempre agem em prol dos meus interesses. Q27. Eu confio plenamente nos funcionários (técnicos-administrativos) da instituição. Q28. Os funcionários (técnicos-administrativos) da instituição sempre cumpriram suas promessas comigo. CONFIANÇA (Trust)

Q29. Eu me sinto muito ligado a minha instituição. Q30. Eu me sinto muito ligado aos meus professores. Q31. Eu sinto orgulho por poder estudar nesta instituição. Q32. Eu sinto orgulho de poder fazer este curso. Q33. Eu escolhi esta instituição por razões práticas. Q34. Quando estabeleço metas para mim eu sempre as alcanço. COMPROMETIMENTO (Commitment)

Q35. Eu regularmente participo de atividades de lazer ligadas à instituição, como esporte ou feiras. Q36. Tenho contato intenso com meus colegas. Q37. Eu regularmente convivo com meus colegas fora da instituição. Q38. Eu regularmente participo de atividades extra acadêmicas, como cursos e eventos. Q39. Eu regularmente me envolvo em grupos de estudo formados pelos próprios estudantes; INTEGRAÇÃO SOCIAL E ACADÊMICA (Social and Academic Integration)

Source: Adapted from Hennig-Thurau et al. (2001)

^aThe questionnaire was applied in Portuguese.

After validation in the pretest, the questionnaires were applied to students in progress, at the HEIs, following a face-to-face strategy. For the evaded students, an electronic form (Google Forms) was employed. According to the original proposal, a 6-point scale was adopted for all questions. Profile data were also inserted at the final part of the instrument.

Design

The conceptual model used is an adaptation by Hennig-Thurau et al. (2001) of the RQSL model of Student Loyalty. In the RQSL model, the constructs Commitment to Work, Family, and Nonacademic activities are approached at the lower level as second-order constructs. Basso et al. (2015) have not used those for their study, probably as those factors were external to the academic environment. This exclusion may be related to the validity of the constructs (Basso et al., 2015) that are not suitable for application in Brazilian HEIs. The original and complete theoretical model of the research is presented in Figure 1.

Figure 1.
Theoretical Model. Source: Adapted from Hennig-Thurau et al. (2001).

Statistical software SPSS version 22 was used for descriptive data analysis and Kolmogorov–Smirnov normality tests, in addition to the Mann–Whitney test, to compare the behavior of the two representative subsamples of the population. To understand the interaction between the constructs of the conceptual model and to obtain the validation metrics of the constructs, the Partial Least Squares (PLS) technique was used, through the software SmartPLS, version 3.2.6. To perform a comparison between sample groups (evaded and nonevaded), the theoretical model was tested to check the differences between groups, against measures, using the Multiple Group Analysis technique (PLS-MGA). This technique allows determining whether there are significant differences in the parameters estimated in groups (Hair et al., 2009). For both cases, there is a minimum recommended number of cases per arrow (10:1) pointed to latent variables to work with the PLS-MGA technique. In this sense, it was necessary to ensure that this minimum sample requirement is observed (Hair et al., 2009).

Results

From the data collected, it can be seen that the evaded students’ profile is single majority, aged between 18 and 25 years (60%); family income of up to R$2,999 (31.4%); full-time work (55.7%); come from public schools (93.4%); and evaded mainly in the 1st year (53.8%). In terms of students in progress, there is a single majority (86.5%), aged between 18 and 25 years (84.7%); family income of up to R$1,999 (41.6%); not working (51.6%); come from public schools (91.7%); and most freshmen (40.1%). A Kolmogorov–Smirnov test was performed to check the normality of the variables including the indicators (Q12–Q39) of the model. The results of the Kolmogorov–Smirnov test showed no normality (p <.05) for all items, indicating the use of the Mann–Whitney nonparametric test (Field, 2009) for comparison between evaded and nonevaded groups.

In the matter of Perceived Quality, significant statistical differences were found in Q12, Q13, and Q18 (p <.05), indicating that there is a distinction between the groups in the evaluation of teachers’ competence, infrastructure, and results obtained in the course. In terms of Loyalty, significant statistical differences were found in Q19, Q21, and Q22 (p <.05), indicating that there is a different perception regarding the recommendation of the course as to the personal interest in maintaining contact with the teachers and to the course choice reiteration. In Loyalty, higher average values were observed in the items for current students, showing that students in daily contact with the institution, faculty, employees, and colleagues present a higher level of loyalty than those who lost this link. In respect of Trust, significant differences (p <.05) were found only in Q27, indicating different perceptions between the groups for trust in the employees. The values for the evaded students were higher, expressing a different behavior from those obtained in Perceived Quality and Loyalty. For Commitment, a statistical difference (p <.05) was achieved only in Q32, indicating that there is a distinction between the groups regarding the student’s feeling of pride in attending the course at HEI. Finally, for the Academic and Social Integration construct, a significant statistical difference was found only in Q38, indicating that there was a distinction between the groups regarding the student’s participation in extra-academic activities. All other variables had no significant values in the test.

Analyzing the original theoretical model (MTO), after the confirmatory factor analysis treatment, there was a high risk of overestimation of the proposed model (Figure 1), due to some constructs being at nonideal levels of explained variance, such Commitment (average variance extracted [AVE] = 0.468) and Perceived Quality (AVE = 0.481), which may indicate problems with the convergence of the constructs. When AVE > 0.50, it is assumed that the model converges to a satisfactory result, according to Fornell and Larcker (Hair et al., 2009). Also, a Discriminant Validity discrepancy was observed when the Commitment and Loyalty constructs were correlated, as the acceptable value should be less than 0.684 compared with the founded value of 0.784. To correct these questions, the variables with the worst factorial loads in the constructs were excluded, respectively, Q13 (0.673), Q16 (0.640), Q17 (0.633), Q18 (0.669), Q22 (0.619), Q24 (0.629), Q33 (0.322), and Q34 (0.362), whose presence impaired or contributed little to the explanation of variance.

After the elimination of the variables, a more synthetic model, called the adjusted theoretical model (MTA), was obtained. They preserved the interactions between constructs, given the theoretical structure of the model and the set of hypotheses that it was intended to test. When the suggested adjustment was made, the MTA presented an improvement in the validity of the Commitment (0.661) and Perceived Quality (0.702) constructs. These values show an AVE above the minimum reference value, as well as in the Discriminant Validity, with no additional discrepancies when compared with the reference that changed to 0.776.

Regarding the Cronbach’s alpha (AC) and composite reliability (CR), both models (MTO and MTA) were above the reference value (CR > 0.70; AC > 0.70), indicating that the sample is free of bias, as well as there is consistent internal reliability of the scale (Field, 2009; Hair et al., 2009). The results of AVE, CR, AC, and R² for the Adjusted Model are presented in Table 3. Concerning the coefficient R², it can be stated that the observed values are at a proper level for the social sciences, above 0.26 (Cohen, 1988). Commitment = 0.562, Trust = 0.290, and Loyalty = 0.657, for the MTO. After adjusting the model (Table 3), the values obtained were slightly smaller and remained close to those obtained initially (Commitment = 0.496, Trust = 0.205, and Loyalty = 0.649). Nevertheless, they have gained in terms of parsimony.

Table 3.
MTA Results.

Construct Average variance extracted Composed reliability Cronbach’s alpha R ²

Commitment 0.661 0.886 0.828 0.496

Trust 0.678 0.893 0.840 0.205

Academic and social integration 0.556 0.862 0.801

Loyalty 0.596 0.854 0.772 0.649

Perceived quality 0.702 0.876 0.788

Source: Research data.

In the analysis of the direction of the variables of the original model (path analysis), there was a direct, nonsignificant effect between Trust and Loyalty (0.026; p = .590), and for other interactions (p = .000). In the adjusted model, the path coefficients were similar to the original model and were not significant for the interaction Trust and Loyalty (0.067) with (p > .05).

Although the eliminated variables described the concepts of the constructs faithfully, the exclusion could make the model more harmonious, as Hair et al. (2009) suggest. Thus, of the 28 indicators from the original model, only 20 were kept after adjustment (Figure 2).

Figure 2.
MTA—Adjusted Theoretical Model—Factor Loads for Observable Variables and Path Coefficients. Source: Research Data.

Regarding Perceived Quality, the MTA explains, in a summarized way, criteria such as the evaluation of teachers’ competence, the orientation given by them and the evaluation process adopted in the course, as proposed by Hennig-Thurau et al. (2001) and Anjos Neto (2003), except for aspects of infrastructure and physical facilities (Fontaine, 2014), which did not remain as indicators of this construct. As for commitment, the MTA makes it possible to harmonically explain the construct and its variables, such as the secure connection between students and teachers, feelings of belonging to HEIs and pride in studying in the course/institution, evidenced in this research and defended by authors such as Bergamo et al. (2011) and Bowden (2011). However, the eliminated variables that deal with the choice of HEIs for practical reasons and the achievement of personal goals mentioned by Basso et al. (2015) and Hennig-Thurau et al. (2001) did little to explain the construct.

In terms of Loyalty, the MTA facilitated the explanation of aspects such as the HEI’s recommendation and the students’ intention to maintain contact with teachers. For Bowden (2011) and Hennig-Thurau et al. (2001), a strong commitment of students to HEI signals their loyalty, which in Ackerman and Schibrowsky’s (2007) vision signalizes students retention in the HEIs, which is motivated by the interaction between relational benefits and quality of the relationship (Bergamo et al., 2011). In this work, differently, variables that express a postperiod relationship with the institution did not remain as indicators of the model. After evaluating the consistency measures, Table 4 shows the t test for the MTA. We also analyze the path coefficients and validate the hypotheses.

Table 4.
t test for MTA.

Structural relation Original sample (O) Bootstrap sample Standard deviation t statistics p

Commitment → Loyalty 0.602 0.604 0.050 12.151 .000

Trust → Commitment 0.251 0.252 0.053 4.706 .000

Trust → Loyalty 0.067 0.065 0.046 1.467 .143

Integration → Commitment 0.311 0.314 0.045 6.883 .000

Perceived Quality → Commitment 0.380 0.377 0.056 6.809 .000

Perceived Quality → Trust 0.452 0.451 0.052 8.632 .000

Perceived Quality → Loyalty 0.242 0.263 0.059 4.468 .000

Source: Research data.

The path coefficients between Perceived Quality and Trust, Perceived Quality and Commitment and between Perceived Quality and Loyalty, whose significant factor loads are, respectively, 0.452, 0.380, and 0.242, could be accepted (p <.05) and also have a minimal accepted strength, validating the hypothesis number H1, H2, and H3. The path coefficient between Trust and Loyalty presented a value of 0.067, below the acceptable minimum to be understood as significant. This result shows a low connection between those constructs, which may be related to the maturity of the group of interviewed students, whose relationships with HEIs can be considered in the initial phase (up to 2 years in the vast majority), which could contribute for not establishing Loyalty based on Trust. The path coefficient between Academic and Social Integration and Commitment also exhibits a significant connection (p <.05) and a factorial load (0.311), which strength is also minimally accepted. Such a situation evidences that the Integration is an antecedent of Commitment for this study.

Finally, it is verified that the path coefficient between Commitment and Loyalty was the relation with the highest factorial load (0.602). This result highlights a significant connection between the constructs and signals a strong influence in the formation of Loyalty based on Commitment. Considering the results of the Structural Equation Modeling-Partial Least Square (SEM-PLS) application on the proposed conceptual model, we summarize the hypotheses validation (Table 5).

Table 5.
Hypotheses Validation.

Hypotheses t test (p < .05) Reject H₀?

H1 Perceived Quality has a positive influence on the student trust in HEI 8.632 Yes

H2 Perceived Quality has a positive influence on the student commitment 6.809 Yes

H3 Perceived Quality has a positive influence on the student loyalty 4.747 Yes

H4 Trust in HEI has a positive influence on commitment 4.706 Yes

H5 Trust in HEI has a positive influence on loyalty 1.467 No

H6 The student’s commitment to the institution has a positive influence on loyalty to the institution 12.151 Yes

H7 Academic and Social Integration have positive influence on the student’s commitment to HEI 6.883 Yes

Note. HEI = higher education institutions.

Source: Author.

To validate H8, an additional comparative analysis between groups was required, as indicated in the methodology. The application of the nonparametric PLS-MGA test was adopted to assess the differences between groups (evaded and nonevaded), which results are shown in Table 6. Analyzing Table 6, we can observe that there are only two paths (Commitment to Loyalty and Perceived Quality to Commitment) at a significant level (p < .05) for the PLS-MGA test. This result shows the paths that behave differently between groups. In this sense, we can conclude that both the evaded and the nonevaded students’ groups have similar behavior, except for the relationships Perceived Quality to Commitment and Commitment to Loyalty. Thus, the H8 can be only partially validated as several paths are nonsignificant (Table 7).

Table 6.
PLS-MGA Test: Differences Between Evaded and Nonevaded Groups.

Constructs Path coefficients p

Commitment → Loyalty 0.167 .047

Trust → Commitment 0.155 .913

Trust → Loyalty 0.023 .601

Social & Academic Integration → Commitment 0.108 .101

Perceived Quality → Commitment 0.239 .017

Perceived Quality → Trust 0.151 .087

Perceived Quality → Loyalty 0.079 .773

Source: Research data.

Table 7.
Hypothesis No. 8 Validation.

Hypothesis t test (p < .05) Reject H₀?

H8 The proposed model has differences in path coefficients between the groups of evaded and non-evaded students Table 6 p values Partial

Source: Author.

For explaining this unexpected result, it is necessary to evaluate the Mann–Whitney U Test and the descriptive for each significant question of the MTA for evaded and nonevaded groups (Table 8). It can be seen that nonevaded students evaluate in a better grade than evaded students the following aspects such as teachers’ competence, results obtained with the course, recommendation, intention to keep contact, choice preference, proud to be an HEI student, and extracurricular activities. This result refers to elements in four of five constructs, revealing a better perception of Quality, Commitment, Social and Academic Integration, and Loyalty.

Table 8
Mann–Whitney U Test: Variables With Significant Differences Between Evaded and Nonevaded Groups.

Construct Variables Mann–Whitney U Z p value Nonevaded Evaded

Mean SD Mean SD

Perceived Quality Q12. Como você avalia, em média, a competência dos professores 8.991 –2.78 0.005 5.08 0.88 4.73 1.02

Q13. Como você avalia a infraestrutura da instituição na qual estuda 9.413 –2.13 0.033 4.08 1.13 4.35 1.24

Q18. Classifique os resultados que você está obtendo com seu curso 7.569 –4.65 0.000 4.73 1.01 3.95 1.44

Loyalty Q19. Eu recomendaria meu curso para outra pessoa 8.348 –3.74 0.000 5.34 0.94 4.64 1.56

Q21. Eu tenho muito interesse em manter-me em contato com meus professors 8.888 –2.84 0.004 4.83 1.18 4.19 1.7

Q22. Se eu tivesse que tomar a decisão novamente, eu escolheria o mesmo curso 7.921 –4.24 0.000 4.91 1.59 3.82 2.11

Trust Q27. Eu confio plenamente nos funcionários (técnico-administrativos) do IF Goiano 9.157 –2.38 0.017 4.11 1.59 4.53 1.29

Commitment Q32. Eu sinto orgulho por poder fazer este curso 8.623 –3.27 0.001 5.2 1.18 4.52 1.73

Social and Academic Integration Q38. Eu regularmente participo de atividades extra-acadêmicas, tais como cursos e eventos. 9.173 –2.41 0.016 4.17 1.6 3.6 1.88

Source: Research data.

In terms of confidence in employees, that is a Trust variable, the evaded students have a better evaluation grade, probably because they left the school at the beginning of the course and did not have time to interact deeper. Thus, it seems to be aligned with the PLS-MGA results, as there is a significant difference in the path Perceived Quality → Commitment → Loyalty.

Conclusion

This study validated, to a large extent, the proposed theoretical model, demonstrating that Perceived Quality, Trust, and Academic and Social Integration have a significant influence on Commitment that impacts students Loyalty. The results showed that Trust had a nonsignificant impact on Loyalty but only a mediating effect by Commitment, not acting as a direct antecedent of Loyalty in this investigation, different from that presented by Garbarino and Johnson, 1999, Hennig-Thurau et al., 2001, among others. This low connection between Trust and Loyalty can be related to the lack of contact of the interviewees with certain services or the low maturity of the investigated, most of whom are in the initial phase of the course. Otherwise, Perceived Quality also has a direct effect on Loyalty and Trust, besides the impact on Commitment. Thus, Perceived Quality seems to be the first antecedent of the whole model.

In terms of the Academic and Social Integration factor and its connection with the Commitment, it was verified that it influences the Commitment, which flags the retention, especially in the 1st year, when the ties are experimental, corroborating with Tinto (1975, 2001). Another relevant connection was between Commitment and Loyalty. The commitment is the identification of the consumer and their affiliation with the company, which generates the continuous use of the service and the recommendation to others (Aurier & N’Goala, 2010; Moreira & Silva, 2015). Students who perceive a mutual commitment between themselves and HEIs are more likely to stay and recommend the institution, as observed by other researchers (Ackerman & Schibrowsky, 2007).

Finally, about 65% of student Loyalty (R² = 0.649) can be explained by the influence of their relationship with the constructs Perceived Quality, Trust, Commitment, and Academic and Social Integration in the MTA. There is a significant difference in the path of Perceived Quality—Commitment—Loyalty when we compare evaded and nonevaded student groups, revealing that the second group has a better evaluation of attributes from those constructs. Since the student has evaded, their assessment becomes lower and more critical to some aspects, mainly related to those constructs.

Regarding the practical contributions, even though the studied HEI is free of charge, federal public Institution, it suffers from student evasion. Therefore, working relationship marketing, especially on the elements contained in the constructs validated in this study, could improve student retention. In this sense, the institution must seek to know its customers all the time. Also, the intention of considering the student as a customer that has needs, preferences, and opinions can help to promote a better channel of interaction in the relationship between institution and student (Bowden, 2011).

There are three main limitations for this study: (a) the cross-section sample that makes the analysis valid only for that time and not along the time; (b) there was no differentiation among types of courses, as the sample elements are from any course; (c) the research was carried out only at the Morrinhos Campus and given the nonprobabilistic characteristics of the sample and unit investigated, the results cannot be inferred for other campuses, nor the institution as a whole. Otherwise, the proposed model for this research could be replicated in any HEIs.

It is suggested for future researches: (a) to apply the model used in a probabilistic sample, covering an individualized analysis of all HEI courses; (b) to expand the research to all IF Goiano campuses, making it possible to make inferences for the whole institution, including On-Line Programs; (c) given the student dropout at the beginning of the course, it is suggested to carry out a qualitative research to highlight the causes that lead students to escape in the first semester of the course; (d) deepen the study of Trust in the context of public HEIs, for a better understanding of its little influence on Loyalty, which reflects student retention.

Considering that Trust in the technical administrative was not significant for this study, it is also suggested to evaluate the causes of this lack of confidence within the students. Finally, it is recommended to replicate this research for students of technical courses, who represent a large percentage of the target audience of the Federal Institutes of Education, Science, and Technology in Brazil.

	Subfactors	Variables	References
Individual factors	Learning	Educational backgroundPreparation for the courseSkillsLimitations failures	Furtado & Alves (2012); Lobo (2012); Albuquerque (2008); Silva Filho et al. (2007); Schargel & Smink (2002); Ryan & Glenn (2002); Tinto (2001); MEC/SESU (1997); Tinto (1975)
Relationship	MaturityCommitmentSocial and academic integration	Arce et al. (2015); Lobo (2012); Aypay et al. (2012); Schargel & Smink (2002); Tinto (2001); MEC/SESU (1997); Tinto (1975)
Individual lifetime	Personal lifeProfessional lifeAgenda and appointments	Furtado & Alves (2012); Tontini & Walter (2011); MEC/SESU (1997)
Internal factors	Infrastructure	LocationClassroomsLibraryConservation	Ambiel et al. (2016); Lobo (2012); Tontini & Walter(2011)
Education quality	Subjects/course requirementsFinancial aidService and supportRecommendation	Tontini & Walter(2011); MEC/SESU (1997)
External factors	General lifetime	Marriage/childrenAddress changeFamiliar referenceEngagements	Ambiel et al. (2016); Tontini & Walter (2011); Albuquerque (2008); Schargel & Smink (2002); Tinto (1975)
Finance	Personal incomeFamily supportEmployer incentives	Ambiel et al. (2016); Arce et al. (2015); Furtado & Alves (2012); Lobo (2012); Tontini & Walter(2011); Silva Filho et al. (2007); Schargel & Smink (2002); MEC/SESU (1997); Cabrera et al. (1992); Tinto (1975)
Professional	Expectations/uncertaintiesProfessional vocationPersonal and professional ascensionEmployability	Tontini & Walter(2011); Weng et al. (2010); Albuquerque (2008); MEC/SESU (1997)

Questionnaire Items (variables)^a	Constructs
Q12. Como você avalia, em média, a competência dos professores. Q13. Como você avalia a infraestrutura da instituição na qual estuda. Q14. Opinião sobre a orientação dada pelos professores. Q15. Critérios de avaliação adotados em seu curso. Q16. Oferta de serviços oferecidos da instituição. Q17. Oferta de lazer e descanso da instituição. Q18. Classificação dos resultados que você está obtendo com o curso.	QUALIDADE PERCEBIDA (Perceived Quality)
Q19. Eu recomendaria o meu curso para outra pessoa. Q20. Eu recomendaria minha instituição para outra pessoa. Q21. Tenho muito interesse em manter contato com meus professores. Q22. Se eu tivesse que tomar a decisão novamente, escolheria o mesmo curso. Q23. Se eu tivesse que tomar a decisão novamente, escolheria a mesma instituição. Q24. Eu me tornaria membro de algum tipo de grupo de ex-alunos ligado a instituição ou a docentes.	LEALDADE (Loyalty)
Q25. Integridade é uma palavra que eu usaria para descrever os funcionários da instituição. Q26. Estou certo de que os funcionários (técnicos-administrativos) sempre agem em prol dos meus interesses. Q27. Eu confio plenamente nos funcionários (técnicos-administrativos) da instituição. Q28. Os funcionários (técnicos-administrativos) da instituição sempre cumpriram suas promessas comigo.	CONFIANÇA (Trust)
Q29. Eu me sinto muito ligado a minha instituição. Q30. Eu me sinto muito ligado aos meus professores. Q31. Eu sinto orgulho por poder estudar nesta instituição. Q32. Eu sinto orgulho de poder fazer este curso. Q33. Eu escolhi esta instituição por razões práticas. Q34. Quando estabeleço metas para mim eu sempre as alcanço.	COMPROMETIMENTO (Commitment)
Q35. Eu regularmente participo de atividades de lazer ligadas à instituição, como esporte ou feiras. Q36. Tenho contato intenso com meus colegas. Q37. Eu regularmente convivo com meus colegas fora da instituição. Q38. Eu regularmente participo de atividades extra acadêmicas, como cursos e eventos. Q39. Eu regularmente me envolvo em grupos de estudo formados pelos próprios estudantes;	INTEGRAÇÃO SOCIAL E ACADÊMICA (Social and Academic Integration)

Construct	Average variance extracted	Composed reliability	Cronbach’s alpha	R ²
Commitment	0.661	0.886	0.828	0.496
Trust	0.678	0.893	0.840	0.205
Academic and social integration	0.556	0.862	0.801
Loyalty	0.596	0.854	0.772	0.649
Perceived quality	0.702	0.876	0.788

Structural relation	Original sample (O)	Bootstrap sample	Standard deviation	t statistics	p
Commitment → Loyalty	0.602	0.604	0.050	12.151	.000
Trust → Commitment	0.251	0.252	0.053	4.706	.000
Trust → Loyalty	0.067	0.065	0.046	1.467	.143
Integration → Commitment	0.311	0.314	0.045	6.883	.000
Perceived Quality → Commitment	0.380	0.377	0.056	6.809	.000
Perceived Quality → Trust	0.452	0.451	0.052	8.632	.000
Perceived Quality → Loyalty	0.242	0.263	0.059	4.468	.000

Hypotheses	t test (p < .05)	Reject H₀?
H1	Perceived Quality has a positive influence on the student trust in HEI	8.632	Yes
H2	Perceived Quality has a positive influence on the student commitment	6.809	Yes
H3	Perceived Quality has a positive influence on the student loyalty	4.747	Yes
H4	Trust in HEI has a positive influence on commitment	4.706	Yes
H5	Trust in HEI has a positive influence on loyalty	1.467	No
H6	The student’s commitment to the institution has a positive influence on loyalty to the institution	12.151	Yes
H7	Academic and Social Integration have positive influence on the student’s commitment to HEI	6.883	Yes

Constructs	Path coefficients	p
Commitment → Loyalty	0.167	.047
Trust → Commitment	0.155	.913
Trust → Loyalty	0.023	.601
Social & Academic Integration → Commitment	0.108	.101
Perceived Quality → Commitment	0.239	.017
Perceived Quality → Trust	0.151	.087
Perceived Quality → Loyalty	0.079	.773

Hypothesis	t test (p < .05)	Reject H₀?
H8	The proposed model has differences in path coefficients between the groups of evaded and non-evaded students	Table 6 p values	Partial

Construct	Variables	Mann–Whitney U	Z	p value	Nonevaded	Evaded
					Mean	SD	Mean	SD
Perceived Quality	Q12. Como você avalia, em média, a competência dos professores	8.991	–2.78	0.005	5.08	0.88	4.73	1.02
	Q13. Como você avalia a infraestrutura da instituição na qual estuda	9.413	–2.13	0.033	4.08	1.13	4.35	1.24
	Q18. Classifique os resultados que você está obtendo com seu curso	7.569	–4.65	0.000	4.73	1.01	3.95	1.44
Loyalty	Q19. Eu recomendaria meu curso para outra pessoa	8.348	–3.74	0.000	5.34	0.94	4.64	1.56
	Q21. Eu tenho muito interesse em manter-me em contato com meus professors	8.888	–2.84	0.004	4.83	1.18	4.19	1.7
	Q22. Se eu tivesse que tomar a decisão novamente, eu escolheria o mesmo curso	7.921	–4.24	0.000	4.91	1.59	3.82	2.11
Trust	Q27. Eu confio plenamente nos funcionários (técnico-administrativos) do IF Goiano	9.157	–2.38	0.017	4.11	1.59	4.53	1.29
Commitment	Q32. Eu sinto orgulho por poder fazer este curso	8.623	–3.27	0.001	5.2	1.18	4.52	1.73
Social and Academic Integration	Q38. Eu regularmente participo de atividades extra-acadêmicas, tais como cursos e eventos.	9.173	–2.41	0.016	4.17	1.6	3.6	1.88

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Fernando de Rosa

Author Biographies

Fernando de Rosa is a visiting professor of Business Administration at FGV Management in Brasilia, DF, Brazil. He earned his Ph.D. from the University of São Paulo. His current research interests include customer retention, digital channels, and customer relationship management, as well as those concepts applied to Higher Educational Institutions (HEI). He also researches quantitative models to assess Brazilian MBA students' internationalization of career and life.

Thiago Luiz de Oliveira is an administrative technician in education at the Federal Institute of Education, Science and Technology of Goiás, in Morrinhos, GO, Brazil. He obtained his master's degree in Business Administration at UniAlfa. He is currently a law student at the Goiatuba University Center. His current research interests include customer retention, customer relationship management, and institutional effectiveness, applied to Higher Education Institutions (HEIs).

References

Abubakar

M. M.

Mohd Mokhtar

S. S.

(2015). Relationship marketing, long term orientation and customer loyalty in higher education. Mediterranean Journal of Social Sciences, 6(4), 466–474. https://doi.org/10.5901/mjss.2015.v6n4p466

Ackerman

Schibrowsky

(2007). A business marketing strategy applied to student retention: A higher education initiative. Journal of College Student Retention: Research, Theory & Practice, 9(3), 307–336. https://doi.org/10.2190/CS.9.3.d

Albuquerque

(2008). Do abandono à permanência num curso de ensino superior [From evasion to retention in a higher education course]. Revista de Ciências da Educação, 7, 19–28. http://sisifo.ie.ulisboa.pt/index.php/sisifo/article

Ambiel

R. A. M.

Carvalho

L. F.

Moreira

T. C.

Bacan

A. R.

(2016). Funcionamento diferencial dos itens na escala de motivos para evasão do ensino superior (M-ES) [Differential item functioning in the Higher Education Dropout Reasons Scale (M-ES)]. Psico, 47(1), 68–76. https://doi.org/10.15448/1980-8623.2016.1.21765

Anjos Neto

M. R.

(2003). Construção e teste de um modelo teórico de marketing de relacionamento para o setor de educação [Construction and testing of a theoretical model of relationship marketing for the educational sector]. [Dissertação de Mestrado, Universidade Federal de Pernambuco]. http://www.liber.ufpe.br/teses/arquivo/20050224150704.pdf

Arce

M. E.

Crespo

Míguez-Álvarez

(2015). Higher education drop-out in Spain: Particular case of universities in Galicia. International Education Studies, 8(5), 247–264. https://doi.org/10.5539/ies.v8n5p247

Aurier

N’Goala

(2010). The differing and mediating roles of trust and relationship commitment in service relationship maintenance and development. Journal of the Academy of Marketing Science, 38(3), 303–325. https://doi.org/10.1007/s11747-009-0163-z

Aypay

Çekiç

Boyaci

(2012). Student retention in higher education in Turkey: A qualitative study. Journal of College Student Retention: Research, Theory & Practice, 14(1), 91–116. https://doi.org/10.2190/CS.14.1.e

Basso

Schwab

E. A.

Isbert

L. T. M.

Slongo

L. A.

(2015). Relações de interdependência entre confiança, comprometimento, qualidade percebida e lealdade em estudantes do ensino superior [Interdependent relationships between trust, commitment, perceived quality and loyalty in higher education students]. Revista Economia & Gestão, 15(41), 5–32. https://doi.org/10.5752/P.1984-6606.2015v15n41p5

10.

Bergamo

F. V. M.

Giuliani

A. C.

Galli

L. C. L. A.

(2011). Modelo de lealdade e retenção de alunos para instituições do ensino superior: Um estudo teórico com base no marketing de relacionamento [Student loyalty and retention model for higher education institutions: A theoretical study based on relationship marketing]. BBR-Brazilian Business Review, 8(2), 43–66. https://doi.org/10.15728/bbr.2011.8.2.3

11.

Bôas

R. V.

(2008). The campus experience: Marketing para instituições de ensino [The campus experience: Marketing for educational institutions]. Summus.

12.

Bowden

J. L.

(2011). Engaging the student as a customer: A relationship marketing approach. Marketing Education Review, 21(3), 211–228. https://doi.org/10.2753/MER1052-8008210302

13.

Cabrera

A. F.

Nora

Castañeda

M. B.

(1992). The role of finances in the persistence process: A structural model. Research in Higher Education, 33(5), 571–593.

14.

Carneiro

M. A.

(2012). LDB Fácil: Leitura crítico-compreensiva artigo a artigo [Easy LDB: Critical-comprehensive reading article by article]. Vozes.

15.

Chen

(2017). The relationships between brand association, trust, commitment, and satisfaction of higher education institutions. International Journal of Educational Management, 31(7), 973–985. https://doi.org/10.1108/IJEM-10-2016-0212

16.

Chenet

Dagger

T. S.

O’Sullivan

(2010). Service quality, trust, commitment and service differentiation in business relationships. Journal of Services Marketing, 24(5), 336–346. https://doi.org/10.1108/08876041011060440

17.

Cohen

(1988). Statistical power analysis for the behavioral sciences. Psychology Press.

18.

De Keyser

Lemon

K. N.

Klaus

Keiningham

T. L.

(2015). A framework for understanding and managing the customer experience (MSI Working Paper No. 15-121). https://www.msi.org/reports

19.

Dziminska

Fijalkowska

Sulkowski

(2018). Trust-based quality culture conceptual model for higher education institutions. Sustainability, 10, 1–22. https://doi.org/10.3390/su10082599

20.

Field

(2009). Descobrindo a estatística usando o SPSS [Discovering Statistics using SPSS]. Artmed.

21.

Fontaine

(2014). Student relationship management (SRM) in higher education: Addressing the expectations of an ever evolving demographic and its impact on retention. Journal of Education and Human Development, 3(2), 105–119. http://jehdnet.com/journals/jehd/Vol_3_No_2_June_2014/7.pdf

22.

Furtado

V. V. A.

Alves

T. W.

(2012). Fatores determinantes da evasão universitária: Uma análise com alunos da Unisinos [Determining factors of university evasion: An analysis with Unisinos students]. Contextus - Revista Contemporânea de Economia e Gestão, 10(2), 115–129. https://doi.org/10.19094/contextus.v10i2.32153

23.

Garbarino

Johnson

M. S.

(1999). The different roles of satisfaction, trust, and commitment in customer relationships. Journal of Marketing, 63(2), 70–87. https://doi.org/10.2307/1251946

24.

Grebennikov

Shah

(2012). Investigating attrition trends in order to improve student retention. Quality Assurance in Education, 20(3), 223–236https://doi.org/10.1108/09684881211240295

25.

Hair

J. F.

Jr. Black

W. C.

Babin

B. J.

Anderson

R. E.

Tatham

R. L.

(2009). Análise multivariada de dados [Multivariate Data Analysis]. Bookman.

26.

Hennig‐Thurau

Klee

(1997). The impact of customer satisfaction and relationship quality on customer retention: A critical reassessment and model development. Psychology & Marketing, 14(8), 737–764. https://doi.org/10.1002/(SICI)1520-6793(199712)

27.

Hennig ‐ Thurau

Langer

M. F.

Hansen

(2001). Modeling and managing student loyalty an approach based on the concept of relationship quality. Journal of Service Research, 3(4), 331–344. https://doi.org/10.1177/109467050134006

28.

Instituto Nacional de Estudos e Pesquisas Educacionais Anísio Teixeira. (2018). Sinopse Estatística da Educação Superior 2018 [Statistical Report of Higher Education 2018]. Ministério da Educação. INEP. http://portal.inep.gov.br/basica-censo-escolar-sinopse-sinopse

29.

Kotler

Fox

K. F. A.

(1994). Marketing estratégico para instituições educacionais [Strategic marketing for educational institutions]. Atlas.

30.

Lemon

K. N.

Verhoef

P. C.

(2016). Understanding customer experience throughout the customer journey. Journal of Marketing, 80(11), 69–96https://doi.org/10.1509/jm.15.0420

31.

Lin

(2011). The role of expected future use in relationship-based service retention. Managing Service Quality: An International Journal, 21(5), 535–551. https://doi.org/10.1108/09604521111159816

32.

Lobo

M. B. C. M.

(2012, Setembro). Panorama da evasão no ensino superior brasileiro: Aspectos gerais das causas e soluções [Overview of evasion in Brazilian higher education: causes and solutions aspects]. ABMES Cadernos, 25.

33.

Masserini

Bini

Pratesi

(2019). Do quality of services and institutional image impact students’ satisfaction and loyalty in higher education? Social Indicators Research, 146, 91–115. https://doi.org/10.1007/s11205-018-1927-y

34.

Meyer

Schwager

(2007). Understanding customer experience. Harvard Business Review, 85(2), 117–126. https://store.hbr.org/Article Reprint No. R0702G

35.

Moreira

A. C.

Silva

P. M.

(2015). The trust-commitment challenge in service quality-loyalty relationships. International Journal of Health Care Quality Assurance, 28(3), 253–266.

36.

Morgan

R. M.

Hunt

S. D.

(1994). The commitment-trust theory of relationship marketing. Journal of Marketing, 58(3), 20–38. https://doi.org/10.1177/002224299405800302

37.

Palharini

F. A.

(2010). Evasão, exclusão e gestão acadêmica na UFF: Passado, presente e futuro [Evasion, exclusion and academic management at UFF: Past, present and future]. ICHF.

38.

Portal Brasil. (2016). MEC defende reformas para reduzir a evasão em faculdades [MEC advocates reforms to reduce college evasion]. Ministério da Educação. www.brasil.gov.br/educacao

39.

Rawson

Duncan

Jones

(2013, September). The truth about customer experience. Harvard Business Review. https://store.hbr.org/product/the-truth-about-customer-experience/R1309G

40.

Reichheld

F. F.

Markey

R. G. J.

Hopton

(2000). The loyalty effect—The relationship between loyalty and profits. European Business Journal, 12(3), 134–139. https://pt.scribd.com/doc/95138476

41.

Reichheld

F. F.

Sasser

W. E.

(1990). Zero defections: Quality comes to service. Harvard Business Review, 68(5), 105–111. https://hbr.org/1990/09/zero-defections-quality-comes-to-services

42.

Rosa

(2001). Canais de atendimento eletrônico e satisfação, retenção e rentabilidade de clientes em bancos [Electronic service channels and customer satisfaction, retention and profitability in banks]. Tese de Doutorado, Universidade de São Paulo. https://teses.usp.br/teses/disponiveis/12/12139/tde-25032004-221736/pt-br.php

43.

Ryan

M. P.

Glenn

P. A.

(2002). Increasing one-year retention rates by focusing on academic competence: An empirical odyssey. Journal of College Student Retention: Research, Theory & Practice, 4(3), 297–324. https://doi.org/10.2190/KUNN-A2WW-RFQT-PY3H

44.

Secretaria de Educação Profissional e Tecnológica. (2014). Documento orientador para a superação da evasão e retenção na Rede Federal de Educação Profissional, Científica e Tecnológica. Ministério da Educação [Guide to overcome evasion and retention in the Federal Network of Professional, Scientific and Technological Education]. http://www.ifto.edu.br/portal/docs/proen/doc_orientador_evasao_retencao_setec.pdf

45.

Secretaria de Educação Superior. (1997). Comissão especial de estudos sobre a evasão nas universidades públicas brasileiras [Special committee on evasion studies in Brazilian public universities]. Ministério da Educação. http://www.udesc.br/arquivos/id_submenu/102/diplomacao.pdf

46.

Schargel

F. P.

Smink

(2002). Estratégias para auxiliar o problema de evasão escolar [Strategies to support the school evasion problem]. Dunya.

47.

Schlesinger

Cervera

Pérez-Cabañero

(2017). Sticking with your university: The importance of satisfaction, trust, image, and shared values. Studies in Higher Education, 42 (12), 2178–2194. https://doi.org/10.1080/03075079.2015.1136613

48.

Sindicato das Entidades Mantenedoras de Estabelecimentos de Ensino Superior no Estado de São Paulo. (2019). Mapa do Ensino Superior no Brasil 2019 [Map of Higher Education in Brazil 2019]. https://www.semesp.org.br/pesquisas/mapa-do-ensino-superior

49.

Silva Filho

R. L. L.

Motejunas

P. R.

Hipólito

Lobo

M. B. C. M.

(2007). A evasão no ensino superior brasileiro [Evasion in Brazilian higher education]. Cadernos de Pesquisa, São Paulo, 37(132), 641–659.

50.

Subrahmanyam

(2017). Relationship between service quality, satisfaction, motivation and loyalty. Quality Assurance in Education, 25(2), 171–188. https://eric.ed.gov/?id=EJ1136952

51.

Tribunal de Contas da União. (2012). Relatório de auditoria operacional em ações da rede federal de educação profissional, científica e tecnológica [Operational audit report on actions of the federal professional, scientific and technological education network]. TCU. Seprog.

52.

Tinto

(1975). Dropout from higher education: A theoretical synthesis of recent research. Review of Education Research, 45(1), 89–125.

53.

Tinto

(2001). Rethinking the first year of college. Higher Education Monograph Series. Syracuse University.

54.

Tinto

(2006). Research and practice of student retention: What next? Journal of College Student Retention: Research, Theory and Practice, 8(1), 1–19. https://doi.org/10.2190/4YNU-4TMB-22DJ-AN4W

55.

Tontini

Walter

S. A.

(2011, Dezembro). Podemos identificar propensão e reduzir a evasão de alunos? Ações estratégicas e resultados táticos para instituições de ensino superior [Can we identify student evasion propensity and reduce it? Strategic actions and tactical results for higher education institutions]. Anais do XI Colóquio Internacional de Gestão Universitária na América do Sul. Florianópolis, 7–9. http://repositorio.ufsc.br/xmlui/handle/123456789/25905

56.

Vasconcelos

R. M.

Almeida

L. S.

Monteiro

(2009). O insucesso e abandono acadêmico na universidade: Uma análise sobre os cursos de engenharia [Academic failure and evasion at the university: An analysis of engineering courses]. In Proceedings of the VI International Conference on Engineering and Computer Education, ICECE 2009 (pp. 457–461). https://copec.eu/congresses/icece2009/proc/ICECE.pdf

57.

Weng

Cheong

(2010). Modelling IS student retention in Taiwan: Extending Tinto and Bean’s model with self-efficacy. Innovation in Teaching and Learning in Information and Computer Sciences, 9(2), 1–12. https://doi.org/10.11120/ital.2010.09020012

58.

Zeithaml

V. A.

Bitner

M. J.

Gremler

D. D.

(2014). Marketing de Serviços: A empresa com foco no cliente [Service marketing: The customer-focused company] (6 ed.). AMGH.