1 INTRODUCTION

The search for the reasons behind attrition has been the object of national and international research, with special attention to higher education courses (Gomes, 1998; Vergidis & Panagiotakopoulos, 2002; Andriola et al., 2006; Adachi, 2009; Alves & Alves, 2012; Castro, 2012; Díaz et al., 2012; Barbosa, 2013; Sales, 2013; Silva, 2013; Slhessarenko et al., 2014; Camelo Neto, 2014; Pereira et al., 2014; Rafael et al., 2015; Silva, 2016; Gama, 2015; Ambiel, 2015; Barbosa et al., 2016; Cornélio et al., 2016; Feitosa, 2016; Bisinoto & Arenas, 2016; Ambiel et al., 2016).

There are studies that seek indicators that explain the phenomenon of student evasion (Canziani, 2015; Lima & Zago, 2016; Vitelli 2013); that assess evasion in the face of public education policies, such as the Unified Selection System (SISU) (Gilioli, 2016; Machado & Szerman, 2017); in different institutions (Braga et al., 2003; Palharini, 2004; Adachi, 2009); in hard sciences (Lima, 2013; Gomes, 2015; Silva, 2016; Freitas et al., 2017); focused on degrees (Vitelli, 2013; Rocha, 2015; Massi & Villani, 2015); and those who outline strategies to minimize the evasion phenomenon (Tontini & Walter, 2014).

As in other areas, in accounting, evasion has grown over the years. According to information made available in 2018 by the National Institute of Educational Studies and Research (INEP), we can see that, in 2014, 953 institutions offered the undergraduate program in Accounting: the number of enrollments was 353,597 and the number of students who dropped out was 131,338, representing more than 37% of freshmen. In 2016, there were already 995 institutions offering the course, with 355,425 enrollments and the number of evasions reaching 159,410. Comparing the years 2014 and 2016, there is an increase of 21.37% in the evasion rate of undergraduate programs in Accounting.

In the case of graduate programs in Accounting, the first class of Masters in Accounting, at the University of São Paulo (USP), graduated in 1974 and in 1977 for Ph.D. (Cunha, 2007). Over the years, there has been a significant growth in the number of Masters and Doctorates in Accounting recognized by the Coordination of Higher Education Personnel (Capes) in Brazil, which in itself justifies the study of evasion at this level of education.

Through the information provided by Capes in relation to the last quadrennium (2013-2016), the number of Higher Education institutions offering graduate programs increased from 15.73% to 17.32% in the period. In relation to the Professional Masters, the growth in the number of enrolled students was 56.77% in the same period. In 2017, there were 30 graduate programs offering the Masters in Accounting and, among these, 14 also offered the Doctorate (Capes, 2017). According to Miranda et al. (2012), this growth should significantly change the dynamics of the teaching-learning process in Accounting in Brazil.

Also between 2013 and 2016, the percentage of evasion was 22.53% in 2013, 20.36% in 2014, 18.35% in 2015 and 23.96% in 2016. In academic Master’s programs, in which the increase in enrolled students was 14.36%, the percentage of dismissals was 14.45% in 2013; 12.93% in 2014; 11.79% in 2015 and 10.63% in 2016. In Doctoral programs, the number of enrolled students also grew, reaching a 21.72% increase. The percentage of evasion was 16.17% in 2013; 13.75% in 2014; 12.07% in 2015; and 10.55% in 2016.

Despite the high percentage of evasion in graduate programs in Brazil, no studies were found that specifically investigated, the reasons that would contribute to such a scenario. In the international literature, there are also not many works addressing the subject at the graduate level (Golde, 1998; Zewotin et al., 2015).

In this context, and considering, mainly, the growth in the number of Master’s and Doctorate programs in Accounting in the country, it is essential to know the reasons that would contribute to the evasion of students in these types of education. Thus, the research question that guided this study was: What are the reasons that contribute to the evasion of students from graduate programs in Accounting in the country?

The objective of this study was to identify the perception of graduate students about the reasons that would contribute to the evasion of students from graduate programs in Accounting. Additionally, we sought to verify the reasons that led graduate students to enroll in graduate programs in Accounting.

The choice of this topic was due to the need to develop more specific studies in the Master’s and Doctorate programs in Accounting, in order to better understand the reasons that would lead students to drop out, especially to allow interventions that help keep them on the program. Thus, this study intends to stimulate further research in the area, also serving as a starting point for analysis in other areas of knowledge.

2 THEORETICAL FRAMEWORK

2.1 Attrition Theories

Initial studies on student evasion were developed by Spady (1970), Tinto (1993, 1997), Bean (1980), Pascarella and Terenzini (1980) and Astin (1985).

The theory of the evasion process developed by Spady (1970, 1971) emphasizes the process of integrating student attributes, values, interests, skills and attitudes into the school environment. Thus, having a harmonious relationship between the student and the school, the process of social and academic assimilation will be satisfactory and its possibilities of permanence will be greater, since, the more satisfied the student is, the greater will be their commitment to the institution, culminating in lower evasion.

The model pictured above was revised by Spady in 1971, as there were differences based on the students’ gender. The author observed that, for men, performance in relation to grades was the most important factor for dropping out of the course, while for women, the commitment to the institution was the determining factor for the evasion. In its adjusted model, there was also the inclusion of structural relationships in the support of friends construct. Structural relationships can be influenced by the relationship between teachers and students, by the relationship with other students of the opposite gender, for academic performance and for the support of friends.

The Student Integration Theory, developed by Tinto (1975), aims to elucidate the particularities and processes that influence a student’s decision to withdraw from a program. In addition, it seeks to understand how these processes converge in the decision to evade or stay. For the author, the student’s choice to stay or not in the course is a consequence of the relationship between them and the institution itself, going through fundamental concepts, such as academic and social integration to the institution. Therefore, for Tinto (1975), students enter the institution with a series of characteristics such as race, gender, past skills acquired along their journey, previous experiences gained, academic and social performance and in a familiar context. Each of these characteristics directly or indirectly impacts the student’s performance at the institution and in the enrolled program.

In 1993, Tinto made changes that refined the previous model, now considering the importance of factors external to the institution, in addition to issues of a financial nature. Tinto (1993), in addition to observing the student’s integration into the context of the institution, believed it was important to add the personal characteristics that contribute to the student’s commitment to their goals and to the institution, emphasizing previous experiences and also the family context.

In addition, Tinto (1997) considered the learning process as an important aspect to determine the permanence or abandonment of the student in the program and in the institution. In this case, social and academic integration began to have a new reach, becoming a crucial item in the learning process, along with pedagogical techniques and technologies made available to students and teachers.

The student attrition theory proposed by Bean (1980) stemmed from the contributions of the model proposed by Tinto (1975) and the assumption that the student no longer has a traditional profile, characterized as that individual who, in addition to carrying out their academic tasks, also performs work outside the institution. This type of student has specific characteristics, such as socioeconomic status, being influenced, too, by the context in which they are inserted and by the various pressures of social groups that force them to join a progam without full identity, nor partially, which leads one simply to give up the course they started and start another one, or even abandon academic life altogether.

When it comes to professional life, changes in the workplace can be decisive when the student decides to stay or leave the program. For Bean (1980), just as the worker prioritizes the remuneration factor when deciding on his/her permanence in the job, the student also considers their academic performance in the assessments when deciding on their permanence in the program.

The model by Pascarella and Terenzini (1980) was also based on the student integration model proposed by Tinto (1975) and in the relationships between student, institution and environment. For the authors, the permanence or not of the student is related to their attendance and the quality of their contacts and information outside the classroom. The other individual characteristics influence the institution’s environment and the students’ social, academic and extracurricular experiences. Such experiences, in turn, can influence the amount of formal contact and all the factors that affect educational outcomes.

Also influenced by Tinto’s model (1975), Astin (1985) developed a model of student involvement focusing on the behavioral aspect, and on the way the individual acts as a determinant in understanding their involvement with the institution.

For Astin (1985), student involvement is related to the quantity and quality of physical and psychological energy invested during their journey. This involvement can take on many faces, such as the absorption of academic study, participation in extracurricular activities and interaction with faculty and other people in the institution. In this sense, the greater the student’s involvement with university life, the greater the chances of his/her remaining at the institution.

2.2 Empirical Studies on Attrition

Through the theories developed by Spady (1970), Tinto (1975, 1993, 1997), Bean (1980), Pascarella and Terenzini (1980) and Astin (1985), national and international studies on attrition were developed, mainly in the undergraduate context. Through these studies, several reasons were found that influence the student to evade or not from the program in which they are enrolled. In Chart 1, we highlight the main ones.

We can observe from the analysis of Table 1 that reasons include both aspects related to the individual’s adaptation process to the study environment (lack of relationship with other students and lecturers) as to the structure and location of the educational institution and the student (difficulties in the subjects, vocation), in addition to contingent ones (entry to another course) and involving lecturers (little motivation).

Additionally, we emphasize that only the studies by Canziani (2015), Golde (1998), Vergidis and Panagiotakopoulos (2002) and Zewotin et al. (2015) were carried out within graduate programs (Table 1). Canziani (2015) investigated the causes of evasion from certificate programs at the University of Southern Santa Catarina; Golde (1998) explored the evasion process of three doctoral students in Accounting at Oxford University; Vergidis and Panagiotakopoulos (2002) traced the causes of evasion in a graduate program in education offered by the Universidade Aberta Helénica; and Zewotin et al. (2015) evaluated the time it took students to successfully complete or drop out of a Master's program at the University of KwaZulu-Natal, on the east coast of South Africa. The other studies found on the subject were developed in the context of undergraduate programs.

Chart 1
Summary of reasons found in previous literature

Source: elaborated by the authors.

Reasons	References
Family pressure when choosing a program	Gomes (1998)
Difficulties in adapting to the new environment	Gomes (1998), Barbosa (2013), Rafael et al. (2015), Massi and Villani (2015)
Lack of choice when choosing a program	Gomes (1998), Sales (2013), Vitelli (2013), Rafael et al. (2015), Golde (1998)
Lack of program information	Gomes (1998), Braga et al. (2003), Alves and Alves (2012), Castro (2012), Amaral (2013), Slhessarenko et al. (2014), Camelo (2014), Canziani (2015), Gama (2015), Massi and Villani (2015), Ambiel et al. (2016)
Difficulties in reconciling the program with work	Gomes (1998), Biazus (2004), Sales (2013), Amaral (2013), Villar (2014), Canziani (2015), Rafael et al. (2015), Durso (2015), Gama (2015), Rocha (2015), Ambiel et al. (2016), Vergidis and Panagiotakopoulos (2002)
Socioeconomic variables	Gomes (1998), Braga, et al. (2003), Alves and Alves (2012), Amaral (2013), Vitelli (2013), Nagai (2017), Durso (2015), Gomes (2015), Rocha (2015), Gama (2015), Barbosa et al. (2016), Feitosa (2016), Bisinoto et al. (2016), Lima and Zago (2016), Silva (2016)
Low academic performance and failures	Braga et al. (2003), Sales (2013), Lima (2013), Vitelli (2013), Villar (2014), Gomes (2015), Rocha (2015), Cornélio et al. (2016), Silva (2016), Díaz et al. (2012), Gonzalez (2017)
Lack of teaching methodology by lecturers	Braga et al. (2003), Biazus (2004), Slhessarenko et al. (2014), Canziani (2015), Durso (2015), Bisinoto et al. (2016)
Evaluation process	Braga et al. (2003), Slhessarenko et al. (2014), Villar (2014), Camelo (2014), Canziani (2015), Durso (2015)
Little motivation by lecturers	Biazus (2004), Slhessarenko et al. (2014), Villar (2014), Cornélio et al. (2016)
Lack of integration between institutions	Biazus (2004), Villar (2014)
Structure of the institution	Palharini (2004), Andriola et al. (2006), Castro (2012), Barbosa (2013), Slhessarenko et al. (2014), Camelo (2014), Gama (2015), Massi and Villani (2015), Ambiel et al. (2016), Cornélio et al. (2016), Golde (1998), Díaz et al. (2012)
Lack of relationship with lecturers	Bardagi (2007), Castro (2012), Massi and Villani (2015)
Lack of relationship with other students	Bardagi (2007), Castro (2012), Massi and Villani (2015)
Distance between the Institution and the residence	Alves and Alves (2012), Sales (2013), Bisinoto et al. (2016)
Joining a new program	Amaral (2013), Slhessarenko et al., (2014), Silva (2016)
Dissatisfaction with the program	Amaral (2013), Vitelli (2013), Slhessarenko et al. (2014), Rafael et al. (2015), Silva (2016)
Schedule mismatch	Canziani (2015), Rafael, et al. (2015)
Difficulties with the subjects	Lima (2013), Vitelli (2013), Villar (2014), Gomes (2015), Rocha (2015), Barbosa et al. (2016), Cornélio et al. (2016)
Early choice of profession	Villar (2014), Gama (2015), Barbosa et al. (2016)
Difficulties in reconciling studies with family and friends	Villar (2014), Canziani (2015), Massi e Villani (2015), Vergidis e Panagiotakopoulos (2002)
Shift that the institution offers the program	Pereira et al. (2014)
Unsatisfactory past performance	Pereira et al. (2014), Nagai (2017), Gomes (2015), Lima e Zago (2016)
Vocation	Gama (2015), Barbosa et al. (2016)
Lack of time for dedication to studies	Canziani (2015), Rafael et al. (2015), Durso (2015), Gama (2015), Rocha (2015), Ambiel et al. (2016), Bisinoto et al. (2016), Vergidis e Panagiotakopoulos (2002)
Psychological factors	Golde (1998)

3 METHODOLOGICAL PROCEDURES

This descriptive and explanatory research, with a quantitative approach, is classified as a survey (Martins & Theóphilo, 2016).

The population of this study includes Master’s and Doctoral students regularly enrolled in graduate programs in Accounting in Brazil, registered at Capes, in the year 2018. According to the information made available by Capes through the Sucupira platform, and accessed on March 21, 2018, 1,164 students were enrolled in the Master’s (1006 in the Master’s of Science and 158 in the Professional Master’s), in addition to 369 in the Doctorate.

With the contacts of the students, obtained through the secretariats of the educational institutions, it was sent, on September 3, 2018, via e-mail, the data collection instrument accompanied by the Informed Consent Form (TCLE), approved by the Research Ethics Committee of the Federal University of Minas Gerais, under the number 85109518.4.0000.5149. On October 30, 2018, the data collection process ended. The sample of graduate students was composed of 619 students who accepted the invitation and answered the questionnaire: 447 in the Master’s program (390 in Master’s of Science and 57 in the Professional Master’s) and 172 in the doctoral program in Accounting.

The data collection instrument, applied to the population of this study, was built based on the study by Durso (2015) and adapted to graduate programs. The goal by Durso (2015) was to seek evidence that would allow characterizing the evasion process of Accounting majors of a Brazilian public Higher Education Institution (HEI). Adaptations were necessary due to the difference in the contexts analyzed: this research was carried out within graduate programs, while the former focused on undergraduate program and included a sample of subjects who had already dropped out of the program. Consequently, language adjustments were needed in the instrument used by Durso (2015), in order to make the assertions aimed at the researched modality and not at the university system in general. Thus, the reasons: “Difficulty in adapting to the university system”; “Lack of social integration with other undergraduate students”; “The course had little emphasis on vocational subjects”; and “UFMG did not provide the necessary support for them to adapt to the university pace”, listed in the instrument by Durso (2015), were replaced by, respectively: “Difficulties in adapting to the program”; “Lack of social integration with classmates”; “The program has little emphasis on subjects that interest me” and “The institution does not provide the necessary support for me to adapt to the pace of the program”. Additionally, two reasons that were not investigated in the studies by Durso (2015), namely “Health problems” and “Loneliness during the program”.

The structure of the questionnaire encompassed demographic and socioeconomic issues of respondents; a question asking the respondent to mark the reason that led him/her to join the program in which he/she was enrolled and 22 reasons that, according to theory, would contribute to the evasion of students. These sentences represent issues related to institutional problems (adaptation to the institution, lack of information on the part of the program), with lecturers (didactics, lack of integration), external (reconciling the program with work, family pressures, social pressures) and the student themselves (lack of motivation, difficulty following the program, health problems).

To achieve the objective of this study, three steps were carried out: (i) sample characterization, (ii) reasons that led graduate students to enroll in graduate programs in Accounting and (iii) perception of the reasons that would contribute to the evasion of students from graduate programs in Accounting.

In the first stage, we made a description of the characteristics of the graduate students who participated in the research, and the frequency of responses was analyzed, in relative and absolute terms. In the second stage, we identified the main reasons that led respondents to join the program under review. Through descriptive statistics, a ranking of these reasons was constructed. In the third stage, we asked respondents to assign scores from 1 to 10, listing the reasons that would contribute the most (10) or the least (1) for the evasion of students from graduate programs in Accounting. A hierarchy was created to categorize motivations as low, medium and high propensity to evade. We considered the reason as a low degree of evasion propensity when the mean of the score attributed to it was between 1 (one) and 3 (three); as of medium degree of propensity to evasion when the mean of the scores attributed to it was above 3 (three) and below 6 (six); and, as a high degree of evasion propensity when the mean of the scores assigned to it was above 6 (six). The division of scores awarded into low, medium and high scores was performed at the researchers’ discretion.

Additionally, we performed t-tests of combined variance to identify the existence of statistically significant differences between the means of scores assigned by students with different profiles.

The sample of this study comprised students enrolled in all graduate programs in Accounting in Brazil (see Appendix A).

Mostly, the 390 graduate students enrolled in the Academic Master’s are white, single, female, aged between 20 and 30 years old, they live in the same city as the program, they are not scholarship holders, they work (mostly in the private sector), have a family income between two to five minimum wages and participate in family economic life. A profile of the 57 graduate students enrolled in professional Master’s programs who participated in the survey points to the male gender, age group between 31 and 35 years old, residents in the city where the course is located, non-scholarship holders, working in the private sector, with family income above ten minimum wages and participants in family economic life.

Regarding the 172 doctoral students who participated in the research, predominantly, are female, aged between 26 and 30 years old, live in the same city as the course, do not have scholarships, work (especially in the public initiative), have a family income between five and ten minimum wages and participate in family economic life.

4 RESULTS

4.1 Reasons for joining graduate programs in Accounting

In addition to information about the profile of students, graduate students were also asked about the main reason that led them to enroll in graduate programs in Accounting. For Lapini (2012), the objective of the graduate program in the academic modality is to enable the student to exercise the teaching career. According to the ranking in Graph 1, for students in Master’s of Science programs, the aptitude for the academic area was the reason that led 38% of them to enroll. Followed by the increase in compensation, marked by 11% of students. In addition, 10% joined to deepen in the area and 7% to achieve greater employability. We also realized that the lack of option, the prestige of the program and the higher productivity were reasons for admission listed by only 1% of the students, each of them. No student identified the program being tuition-free as a reason, regardless of whether they were enrolled in public or private institutions.

Graph 1
Reasons that led students to enroll in the Master’s of Science program
Source: elaborated by the authors.

According to Capes (2014), the Professional Master’s adds a higher level of competitiveness and productivity to public or private companies, contributing to the national productive sector. This modality of Master’s is regulated by the ordinances n° 389 by MEC and nº 131 by Capes, 2017. In it, 25% of students claimed to have entered the program to acquire professional differentiation, which was expected, since this is the main objective of courses in this modality. On the other hand, 19% of students entered due to their aptitude in the academic area, 19% for deepening the area, 14% for greater employability and 9% for greater professional mobility. No respondent listed the tuition-free program, greater productivity, development of cognitive skills, prestige or lack of option as a reason. The other reasons can be seen in the ranking presented in Graph 2.

Graph 2
Reasons that led students to enroll in the Professional Master’s program
Source: elaborated by the authors.

Still according to Capes (2014), the training of lecturers, the encouragement and development of academic research through the adequate preparation of researchers are the two practical objectives of the Master’s of Science and Doctorate. According to the ranking presented in Graph 3, the aptitude for the academic area was the reason that led 44% of students enrolled in Doctoral courses to enroll in them, even if the aptitude for the academic area, according to the objectives of Capes (2014), is already enabled for students with a Master’s degree. On the other hand, greater productivity and deepening in the area, which are also fundamental objectives of Doctoral programs, were reasons for admission for 2% and 9% of students, respectively. The increase in compensation was the reason for admission indicated by 12% of students.

Graph 3
Reasons that led students to enroll in the Doctorate program
Source: elaborated by the authors.

In summary, as expected, the aptitude for the academic area was the main reason listed by Master’s and Doctoral students in the academic modality, considering that the objective of this modality is precisely to enable the student to exercise the teaching career, being essential for this aptitude for the area. In the professional modality, the highlight was the item acquiring professional differentiation, which was also an expected result for this modality. An interesting result was observed in the item aptitude for the academic area, listed by 19% of students, even though they have chosen the modality of professional education to the detriment of the academic one, which was perhaps the most appropriate in these cases.

4.2 Reasons that would contribute to attrition in graduate programs in accounting

Through the mean of the scores assigned by students enrolled in graduate programs in Accounting in the third part of the research instrument, it was possible to identify the reasons that would most contribute to attrition. The reason was considered as having a low degree of evasion propensity when the mean of the scores assigned was between 1 (one) and 3 (three); as of medium degree of evasion propensity when the mean of the scores assigned was above 3 (three) and below 6 (six); and, as a high degree of evasion propensity when the mean of the scores given was above 6 (six).

For students enrolled in the Master’s in Accounting programs (scientific and professional modalities), the difficulties to follow the course due to the time needed for the studies was indicated as the reason that would most lead to evasion (mean = 6.266), representing a high degree of propensity. The difficulty in reconciling academic activities with professional activities (mean = 5.884) and the difficulties to follow the course due to the lack of financial resources (mean = 5.091) represented medium-scoring reasons. On the other hand, admission to the program by passing the selection process even though it is not the student’s choice (mean = 2.327) and the loss of prestige of the accounting career (mean = 2.564) were the reasons with the lowest means, representing low-scoring reasons, as shown in Table 1.

Table 1
Scores given by Master's students

Source: elaborated by the authors.

N°	Reasons	Mean	Degree of propensity
2	Difficulties in following the program due to the time needed for studies	6.2662192	High
1	Difficulties in following the content taught in the program	4.57718121	Medium
3	Difficulties in following the program due to lack of financial resources	5.0917226	Medium
4	Unsatisfactory academic performance	4.0917226	Medium
5	Lecturers’ didactic deficiency	4.20581655	Medium
6	Lack of motivation to continue the program	4.88590604	Medium
7	Difficulties in adapting to the program	4.62863535	Medium
10	The program has little emphasis on subjects that interest me	4.08277405	Medium
11	Deficiency in the institution’s infrastructure	3.08501119	Medium
13	The program was not what I imagined	4.27069351	Medium
15	Lack of receptivity from program lecturers	3.77628635	Medium
16	The institution does not provide the necessary support for me to adapt to the pace of the program	4.05369128	Medium
17	The profession does not allow me to achieve the financial return I want	3.32885906	Medium
19	Difficulties in reconciling academic activities with professional activities	5.8836689	Medium
20	No necessary information was made available for me to better understand the course I took	3.4049217	Medium
21	Health problems	4.27293065	Medium
22	Loneliness during the program	4.15212528	Medium
8	Lack of social integration with program students	2.95749441	Low
9	The program curriculum is out of date	2.98657718	Low
12	Loss of prestige in the accounting career	2.56375839	Low
14	I realized that I do not feel good about the profession	3.00223714	Low
18	I enrolled in the program because I passed the selection process, but actually this was not my choice	2.32662192	Low

Students enrolled in the Doctoral program in Accounting did not indicate reasons that would represent a high degree of contribution to the evasion propensity (see Table 2). As for the reasons for high school, as verified in relation to students of Master’s programs, they presented higher scores to the difficulties to reconcile academic activities with professional activities (mean = 5.383) and the difficulties to follow the program due to the time needed for studies (mean = 5.198). The loss of prestige of the accounting career (mean = 2.419) and the perception of not feeling well in the profession (mean = 2.593) were reasons evaluated by Doctoral students as low score.

Table 2
Scores given by Doctoral students

Source: elaborated by the authors.

N°	Reasons	Mean	Grau de propensão
1	Difficulties in following the content taught in the program	3.8313953	Medium
2	Difficulties in following the program due to the time needed for studies	5.1976744	Medium
3	Difficulties in following the program due to lack of financial resources	4.3604651	Medium
4	Unsatisfactory academic performance	3.4767442	Medium
5	Lecturers’ didactic deficiency	4.2151163	Medium
6	Lack of motivation to continue the program	4.5755814	Medium
7	Difficulties in adapting to the program	4.0290698	Medium
8	Lack of social integration with program students	3.0116279	Medium
9	The program curriculum is out of date	3.0290698	Medium
10	The program has little emphasis on subjects that interest me	4.0988372	Medium
11	Deficiency in the institution’s infrastructure	3.2325581	Medium
13	The program was not what I imagined	3.627907	Medium
15	Lack of receptivity from program lecturers	3.7790698	Medium
16	The institution does not provide the necessary support for me to adapt to the pace of the program	3.8081395	Medium
19	Difficulties in reconciling academic activities with professional activities	5.3837209	Medium
21	Health problems	4.6511628	Medium
22	Loneliness during the program	4.0465116	Medium
14	I realized that I do not feel good about the profession	2.5930233	Low
12	Loss of prestige in the accounting career	2.4186047	Low
17	The profession does not allow me to achieve the financial return I want	2.8953488	Low
18	I enrolled in the program because I passed the selection process, but actually this was not my choice	2.6662791	Low
20	No necessary information was made available for me to better understand the course I took	2.9011628	Low

In short, it was possible to observe that most Master’s and Doctoral students considered that the difficulties to follow the content taught in the program (reason 1), the difficulties to follow the program due to the lack of financial resources (reason 3), the difficulties of adapting to the program (reason 7) and the lack of necessary information (reason 20) are reasons that would contribute to the evasion of students from graduate programs in Accounting at a low scores (mean between 2.326 and 5.092). The average difference tests revealed that the means of the scores given by the Master’s and Doctoral students were statistically different (p-value < 0.05), despite having the same degree.

As for the didactic deficiency of lecturers (reason 5), the lack of motivation to continue the program (reason 6), the lack of social integration with program students (reason 8), to outdated program curriculum (reason 9), the little emphasis on subjects of interest to the student (reason 10), to the deficiency in the institution’s infrastructure (reason 11), to the loss of prestige of the accounting career (reason 12), not feeling well in the profession (reason 14), the lack of receptivity of the program lecturers (reason 15), the lack of necessary support from the institution so that the student can adapt to the pace of the program (reason 16), to the profession not allowing the student to achieve the desired financial return (reason 17), to join the program because one passed the selection process even though it was not his/her choice (reason 18), to health problems (reason 21), to loneliness during the program (reason 22) and the program is not what the student imagined (reason 13), were considered by most Master’s and Doctoral students as low-scoring reasons (mean between 2.418 and 4.885) in the contribution to the evasion of students from graduate programs in Accounting. However, the means of the scores assigned by Master’s and Doctoral students for these reasons were considered statistically equal (p-value > 0.05).

Difficulties in following the course due to the time needed for studies (reason 2) and unsatisfactory academic performance (reason 4) were considered high-scoring reasons (mean between 3.476 and 6.267) that would contribute to the evasion of students from graduate programs in Accounting, for most Master’s and Doctoral students. In this case, the means of the scores assigned by the Master’s and Doctoral students were considered statistically different (p-value = 0.0001).

Finally, the difficulties in reconciling academic activities with professional activities (reason 19) was a reason perceived as high (mean between 5.383 and 5.883), that would contribute to the evasion of students from graduate programs in Accounting, both by Master’s and Doctoral students, with means of the scores considered statistically equal (p-value = 0.0935).

The reasons that would most contribute to the evasion of students were also analyzed, taking into account other characteristics of the students’ profile, in addition to the course modality. Thus, the following characteristics were considered for the analysis: student’s gender, age, whether they are a scholarship holder or not, whether they work or not, income and whether or not they reside in the city where the program is located (see Appendices B and C).

For female students, the difficulties to follow the course due to the time needed for studies was indicated as the reason that would most contribute to evasion (reason 2, mean = 6.1351), representing a high degree of propensity. In relation to male students, there was no indication of reasons that represented a high level contribution to the propensity to drop out. Among the reasons listed as low-scoring, women indicated the loss of prestige of the accounting career (mean = 2.5616) and the possibility of enrolling in the course because it passed the selection process, but actually this was not the choice (mean = 2.4024). Men indicated these same reasons as low-scoring, in addition to listing three others: the lack of social integration with program students (reason 8, mean = 2.7624), the fact that they do not feel good in the profession (reason 14, mean = 2.4929) and the fact that the profession does not allow the student to achieve the desired financial return (reason 17, mean = 2.9965). The means of the scores given by female and male students were considered statistically different only in relation to reason 14 (p-value = 0.0008).

In relation to age, for older students (above the median age of 31 years old), the difficulties to follow the program due to the time needed for studies was indicated as the reason that would most contribute to evasion (reason 2, mean = 6.2491). Younger students (under the age of 31) did not indicate reasons that represented a high degree of contribution to the evasion propensity. As for the reasons listed as low in the evasion propensity, the younger students indicated the following: the lack of social integration with program students (reason 8, mean = 2.9942), the fact that the program curriculum is outdated (reason 9, mean = 2.8830), the loss of prestige of the accounting career (reason 12, mean = 2.5058) and the possibility of enrolling in the program because they passed the selection process, but actually this was not their choice (reason 18, mean = 2.3567).

Older students, on the other hand, indicated as low-scoring reasons for the propensity to drop out: reasons 8 (mean = 2.9458), 12 (mean = 2.5451) and 18 (mean = 2.3141), in line with the perceptions of younger students, in addition to indicating the fact that they do not feel good in the profession (reason 14, mean = 2.6715) and that the profession does not allow the student to achieve the desired financial return (reason 17, mean = 2.9711). The means of the scores given by older and younger students were considered statistically different only in relation to reasons 2 and 17 (p-values = 0.0361 and 0.0413, respectively).

Scholarship students did not indicate reasons that represented a high degree contribution to the evasion propensity. For non-scholarship holders, the difficulties in following the program due to the time needed for studies (reason 2, mean = 6.4237) and the difficulties to reconcile academic activities with professional activities (reason 19, mean = 6.4289) were indicated as the reasons that would most contribute to evasion, representing a high degree of propensity. As for the reasons listed as low in the evasion propensity, the scholarship students indicated: the fact that the program curriculum is outdated (reason 9, mean = 3,000), the loss of prestige of the accounting career (reason 12, mean = 2.4937) and the possibility of enrolling in the program because one passed the selection process, but actually this was not their choice (reason 18, mean = 2.5732).

The non-scholarship students listed, in addition to the reasons already highlighted by the scholarship holders (9, 12 and 18), two other additional reasons such as low-scoring evasion propensity: the lack of social integration with program students (reason 8, mean = 2.8684) and the fact that they do not feel good in the profession (reason 14, mean = 2.6605). The means of the scores given by scholarship holders and non-scholarship holders were considered statistically different only in relation to reasons 2, 14 and 19 (p-values = 0.0000, 0.0156 and 0.0000, respectively).

Students who work, in a manner equivalent to that indicated by non-scholarship students, listed the difficulties to follow the program due to the time needed for studies (reason 2, mean = 6.3366) and the difficulties to reconcile academic activities with professional activities (reason 19, mean = 6.4307) as the reasons that would most contribute to evasion, representing a high degree of propensity. Students who do not work did not indicate reasons that represented a high degree of contribution to the evasion propensity.

Among the reasons listed as low in the evasion propensity, working students indicated: the lack of social integration with program students (reason 8, mean = 2.7847), the fact that the program curriculum is outdated (reason 9, mean = 2.9455), the deficiency in the institution’s infrastructure (reason 11, mean = 2.9307), the loss of prestige of the accounting career (reason 12, mean = 2.4381), the fact that they do not feel good in the profession (reason 14, mean = 2.6361) and the possibility of enroll in the program because one passed the selection process, but actually this was not their choice (reason 18, mean = 2.5256). Students who do not work listed reasons 12 and 18 as low in the evasion propensity. The means of the scores given by students who work and by students who do not work were considered statistically different only in relation to the reasons 2, 8, 11, 14 and 19 (p-values = 0.0467, 0.0451, 0.0307, 0.0107 and 0.0000, respectively).

Students with higher income (higher than five minimum wages), equivalent to working students and non-scholarship students, perhaps because it is a single group of students, indicated the difficulties to follow the program due to the time needed for studies (reason 2, mean = 6.0455) and the difficulties to reconcile academic activities with professional activities (reason 19, mean = 6.0126) as those that would most contribute to evasion, representing a high degree of propensity. Students with lower income (up to five minimum wages) listed the lack of financial resources (reason 3, mean = 6.0807) as the reason that would most contribute to evasion. Regarding the reasons listed as low in the evasion propensity, students with higher income indicated the following: the lack of social integration with program students (reason 8, mean = 2.6970), the fact that the program curriculum is outdated (reason 9, mean= 2.9520), the loss of prestige of the accounting career (reason 12, mean = 2.3813), the fact that they do not feel good in the profession (reason 14, mean = 2.5682) and the possibility of entering the program because one got through the selection process, but actually this was not their choice (reason 18, mean = 2.2096). Students with lower income listed reasons 12 and 18 as low in the propensity for evasion. The means of the scores given by students with higher income and by students with lower income were considered statistically different only in relation to reasons 3, 8 and 14 (p-values = 0.0000, 0.0082 and 0.0017, respectively).

Finally, the last characteristic analyzed was the fact that the student resides in the city where the program is located or outside it. Students who live outside the city where the course is located, indicated the difficulties in following the program due to the time needed for studies (reason 2, mean = 6.1604) as the reason that would most contribute to evasion, representing a high degree of propensity. Students who reside in the city where the program is located did not indicate reasons that would represent a high degree of contribution to the propensity for evasion.

As for the reasons listed as low-scoring in the propensity for evasion, students who reside outside the city of the program indicated: the fact that the program curriculum is outdated (reason 9, mean = 2.8585), the loss of prestige of the accounting career (reason 12, mean = 2.6745) and the possibility of enrolling in the program because one got through the selection process, but actually this was not their choice (reason 18, mean = 2.4575). Students who reside in the city where the program is located, in addition to motivations 12 and 18, already listed by the group of students who reside outside it, they added the following reasons as of low degree in the propensity for evasion: the lack of social integration with program students (reason 8, mean = 2.9017) and the fact of not feeling good in the profession (reason 14, mean = 2.8133). The means of the scores given by students who live in the city where the program is located and by students who live outside it were considered statistically different only in relation to reason 2 (p-value = 0.0388).

In summary, evidence was found that female graduate students, older students, non-scholarship students, working students, with higher income and those who reside outside the city where the program is located, see the difficulties to follow the program due to the time needed for studies as a reason that would strongly contribute to evasion. These groups of individuals seem to have to reconcile other relevant activities with graduate school, in addition to spending a period with commuting, so they would consider dropping out of the program if time became too short to adequately dedicate themselves to the Master’s or Doctorate.

5 FINAL CONSIDERATIONS

The objective of this study was to identify the perception of graduate students about the reasons that would contribute to the evasion of students from graduate programs in Accounting. Additionally, we intended to verify the reasons that led them to enroll in the programs in question. We evaluated the perceptions of 447 students enrolled in the Master’s in Accounting programs (390 of the scientific modality and 57 of the professional) and of 172 enrolled in the Doctorate in Accounting programs, by completing a questionnaire.

Regarding the reasons that would contribute to the evasion of students from graduate programs in Accounting, those that stood out the most, culminating in a high degree of contribution to the dropout from the program, were the difficulties to follow the program due to the time needed for studies and the difficulties to reconcile academic activities with professional activities.

This result confirms the theories by Spady (1970), Tinto (1975) and Bean (1980) and match the findings by Gomes (1998), Biazus (2004), Sales (2013), Amaral (2013), Villar (2014), Canziani (2015), Rafael et al. (2015), Durso (2015), Gama (2015), Rocha (2015), Ambiel et al. (2016), Vergidis and Panagiotakopoulos (2002) and Gonzalez (2017), evidencing that these are reasons that can be decisive for student evasion.

Analyzing the different profiles of students separately, it is noted that the reason - the difficulties to follow the program due to the time needed for studies - indicated as a strong contributor to evasion, was reinforced, especially by students enrolled in the Master’s in Accounting programs, by women, by older students, by non-scholarship students, by working students, by students with higher incomes and by those who reside outside the city where the program is located.

Empirical evidence has shown that, in relation to the reasons that led students to enroll in the graduate program in Accounting, the aptitude for the academic area predominated among the students of the Master’s of Science programs. This result was expected, considering that the main objective of this type of program is precisely to enable the student to exercise the teaching career. Regarding Professional Master’s Degree students, the most mentioned reason was to acquire professional differentiation, which was also expected. An interesting finding is that, even by enrolling in a Professional Master’s, 19% of students reported that the main reason for enrolling in the program was their aptitude for the academic area, which raises doubts regarding the chosen modality.

Perhaps, for these individuals, an Academic Master's Degree would be more appropriate. Regarding students enrolled in Doctoral Degree courses, most of them indicated their aptitude for the academic area as a reason for enrolling in the program, similarly to what was listed by the respondents of the Master’s of Science Degree. It can be seen, therefore, that such aptitude favors the interest in continuing studies, even though teacher training is already possible only with the Master’s program. Perhaps the interest in such continuity is also due to the vacancies offered by Higher Education Institutions (HEI) to fill their teaching staff, which almost always require the Doctoral degree or give higher scores to those who hold such degree.

These findings reinforce the need for discussions on the topic within the HEI, in order to support students throughout the performance of their academic activities. By recognizing possible difficulties reported by students, members of educational institutions have the opportunity to seek alternatives to help them deal with uncomfortable situations at the university, including the discussion of techniques that allow better planning of time dedicated to academic activities and avoid, for example, the procrastination process. Students, with the help of lecturers, educational institutions and their own colleagues, will be able to plan their time more appropriately, organizing the days so that the carrying out of activities is more efficient. Such planning, in addition to the support of the various actors involved in the teaching-learning process, can make the student feel assisted and improve their academic performance, culminating, ultimately, in the possibility of minimizing evasion.

As limitations of this study, it should be noted that the results found do not extrapolate the researched sample. In addition, the limitations of the data collection instrument itself are highlighted, as it contains closed questions, it prevents different reasons from being listed by respondents. For future studies, we suggest conducting a survey specifically with dropout students, so that it is possible to investigate the reasons that effectively contributed to the evasion process. We also suggest including in the studies faculty members and graduate program administrators to verify their perspectives on this phenomenon.

REFERENCES

Adachi, A. A. C. T. (2009). Evasão e evadidos nos cursos de graduação da Universidade Federal de Minas Gerais. (Dissertação de Mestrado). Belo Horizonte: UFMG / FAE.

Alves, V. V. F., & Alves, T. W. (2012). Fatores determinantes da evasão universitária: uma análise com alunos da Unisinos. Contextus, 10(2).

Amaral, J. B. D. (2013). Evasão discente no ensino superior: estudo de caso no Instituto Federal de Educação, Ciência e Tecnologia do Ceará (Campus Sobral). 2013. 48 f (Tese de Doutorado) do programa de Pós-Graduação em Superintendência de Recursos Humanos, Universidade Federal do Ceará, Fortaleza).

Ambiel, R. A. (2015). Construção da Escala de Motivos para Evasão do Ensino Superior. Avaliação Psicológica, 14(1), 41-52.

Ambiel, R. A., Carvalho, L. D. F., Moreira, T. D. C., & Bacan, A. R. (2016). Funcionamento diferencial dos itens na Escala de Motivos para Evasão do Ensino Superior (M-ES). Psico, 47(1), 68-76. DOI: 10.15448/1980-8623.2016.1.21765

Andriola, W. B., Andriola, C. G., & Moura, C. P. (2006). Opiniões de docentes e de coordenadores acerca do fenômeno da evasão discente dos cursos de graduação da Universidade Federal do Ceará (UFC). Ensaio: aval. pol.públ. Educ, 14(52). DOI: https://doi.org/10.1590/S0104-40362006000300006

Astin, A. W. (1985). Student involvement: A developmental theory for higher education. Journal of college student personnel, 25(4), 297-308.

Barbosa, C. L. D. (2013). Preditores de evasão em diferentes ambientes acadêmicos. (Dissertação de Mestrado). Programa de Pós-graduação em Psicologia. Universidade Federal da Bahia.

Barbosa, E. T., Nascimento Filho, R., Azevedo Filho, A. C., & Biavatti, V. T. (2016) Fatores determinantes da evasão no curso de Ciências Contábeis de uma Instituição Pública de Ensino Superior. In: Anais do XIII Congresso USP de Contabilidade.

Bardagi, M. P. (2007). Evasão e comportamento vocacional de universitários: estudos sobre o desenvolvimento de carreira na Graduação (Tese de Doutorado), Universidade Federal do Rio Grande do Sul).

Bean, J. P. (1980). Dropouts and turnover: The synthesis and test of a causal model of student attrition. Research in higher education, 12(2), 155-187. DOI: https://doi.org/10.1007/BF00976194

Braga, M. M., Peixoto, M. D. C. L., & Bogutchi, T. F. (2003). A evasão no ensino superior brasileiro: o caso da UFMG. Avaliação, 8(1), 161-189.

Biazus, C. (2004). Sistema de fatores que influenciam o aluno a evadir-se dos cursos de graduação na UFSM (Tese de Doutorado). Universidade Federal de Santa Catarina. Florianópolis, SC.

Bisinoto, G. D. S., & Arenas, M. V. D. S. (2016). Gestão da Permanência: um estudo sobre o perfil socioeconômico, permanência e evasão dos discentes do Curso de Bacharelado em Administração Pública da UAB/UNEMAT. In: XVI Coloquio Internacional de Gestión Universitaria - CIGU. Arequipa, Peru.

Camelo Neto, G. (2014). Análise dos fatores da evasão discente de uma IES privada de Fortaleza-CE (Tese de Doutorado). Universidade Federal do Ceará, Programa de Pós-graduação em Políticas Públicas e Gestão da Educação Superior, Fortaleza (CE). 118p.

Canziani, I. F. S. (2015). Evasão dos cursos de pós-graduação lato sensu (2010- 2014) da Universidade do Sul de Santa Catarina/UNISUL - Campus Sul, Tubarão/SC.

Castro, A. (2012). Evasão no ensino superior: um estudo no curso de psicologia da UFRGS. 2012. 115 f (Dissertação de Mestrado). Programa de Pós-graduação em Psicologia. Universidade Federal do Rio Grande do Sul, Porto Alegre).

Coordenação de Pessoal de Nível Superior (CAPES). (2017).

Cornelio, R. A., Vasconcelos, F. C. W.,& Goulart, I. B. (2016). Educação a distância: uma análise estatística dos fatores relacionados à evasão e à permanência. Revista Gestão Universitária na América Latina-GUAL, 9(4). DOI: https://doi.org/10.5007/1983-4535.2016v9n4p26

Cunha, J. V. A. D. (2007). Doutores em Ciências Contábeis da FEA/USP: análise sob a óptica da teoria do capital humano. São Paulo (Tese de Doutorado). Programa de Pós-Graduação em Controladoria e Contabilidade, Departamento de Contabilidade e Atuária, Faculdade de Economia, Administração e Contabilidade da Universidade de São Paulo.

Díaz, O. E., Guajardo, D. C., Fiegehen, L. G., Campos, J. L., & Grau, E. S. C. (2012). Fatores intraescolares associados ao abandono escolar no Chile: um estudo de caso. Revista Lusófona de Educação, (20), 47-64.

Durso, S. O. (2015). Características do processo de evasão dos estudantes do curso de Ciências Contábeis de uma Universidade Pública Brasileira. (Dissertação de Mestrado). Programa de Pós-graduação em Controladoria e Contabilidade. Faculdade de Ciências Econômicas, Universidade Federal de Minas Gerais, Belo Horizonte.

Feitosa, J. M. (2016). Análise de evasão no ensino superior: uma proposta de diagnóstico para o campus de laranjeiras. (Dissertação de Mestrado). Programa de Pós-graduação em Administração Pública. Universidade Federal de Sergipe.

Gama, E. N. K. (2015). Obstáculos à formação no ensino superior: um estudo da retenção discente nos cursos de graduação presencial do Centro de Artes da UFES (Dissertação de Mestrado). Programa de Pós-graduação em Gestão Pública. Universidade Federal do Espírito Santo).

Gilioli, R. D. S. P. (2016). Evasão em instituições federais de ensino superior no Brasil: expansão da rede, Sisu e desafios. Estudo Técnico - Câmara dos deputados.

Golde, C. M. (1998). Beginning graduate school: Explaining first year doctoral attrition. New directions for higher education, 1998(101), 55-64. DOI: http://dx.doi.org/10.1002/he.10105

Gonzalez, M. V. (2017). Estudio del abandono empleando un modelo de riesgos proporcionales. In: Congresos CLABES.

Gomes, A. A. (1998). Evasão e Evadidos: O discurso dos ex-alunos sobre evasão escolar nos cursos de licenciatura. 1998. 175 f (Tese de Doutorado). Programa de Pós-graduação em Educação da Universidade Estadual Paulista, Marília.

Gomes, K. A. (2015). Indicadores de permanência na educação superior: o caso da disciplina de Cálculo Diferencial e Integral I. Dissertação - Centro Universitário de La Salle, Canoas.

Lapini, V. C. (2012). Panorama da formação do professor em Ciências Contábeis pelos cursos stricto sensu no Brasil. (Dissertação de Mestrado). Ribeirão Preto, USP/RP.

Lima Júnior, P. (2013). Evasão do ensino superior de Física segundo a tradição disposicionalista em sociologia da educação. (Tese de Doutorado) Universidade Federal do Rio Grande do sul.

Lima, F. S., & Zago, N. (2016). Evasão no Ensino Superior: tendências e resultados de pesquisa. XI Anped Sul - Educação, movimentos sociais e políticas governamentais.

Machado, C., & Szerman, C. (2017). Centralized Admissions and the Student- College Match. DOI: https://dx.doi.org/10.2139/ssrn.2844131

Martins, G. D. A., & Theóphilo, C. R. (2016). Metodologia da investigação cientifica. São Paulo: Atlas.

Massi, L., & Villani, A. (2015). Um caso de contratendência: baixa evasão na licenciatura em química explicada pelas disposições e integrações. Educação e Pesquisa, 41(4), 975-992. DOI: https://doi.org/10.1590/s1517-9702201512135667

Miranda, G. J., Nova, S. P. D. C. C., & Cornacchione Júnior, E. B. (2012). Os Saberes dos Professores-Referência no Ensino de Contabilidade*/The Knowledge of Model Professors in Teaching Accounting. Revista Contabilidade & Finanças, 23(59), 142. DOI: https://doi.org/10.1590/S1519-70772012000200006

Nagai, N. P., & Cardoso, A. L. J. (2017). A evasão universitária: uma análise além dos números. Revista Estudo & Debate, 24(1). DOI: http://dx.doi.org/10.22410/issn.1983-036X.v24i1a2017.1271

Palharini, F. (2004). Elementos para a compreensão do fenômeno da evasão na UFF. Avaliação, 9(2), 51-80.

Pascarella, E. T., & Terenzini, P. T. (1980). Predicting freshman persistence and voluntary dropout decisions from a theoretical model. The journal of higher education, 51(1), 60-75. DOI: https://doi.org/10.1080/00221546.1980.11780030

Pereira, R. S., Zavala, A. A. Z.,& Santos, A. C. (2014). Evasão na Universidade Federal de Mato Grosso. Revista de Estudos Sociais, 13(26), 74-86.

Rafael, J. A. M., Miranda, P. R. D., & Carvalho, M. P. D. (2015). Análise da Evasão em um Curso de Licenciatura em Matemática da Rede federal de ensino nos seus primeiros cinco anos de implantação. Revista Paranaense de Educação Matemática, 4(6).

Rocha, C. S. (2015). Por que eles abandonam? evasão de bolsistas PROUNI dos cursos de licenciaturas. (Tese de Doutorado). Universidade do Vale do Rio dos Sinos, São Leopoldo.

Sales Junior, J. S. (2013). Uma Análise Estatística dos Fatores de Evasão e Permanência de Estudantes de Graduação Presencial da UFES (Dissertação de Mestrado). Programa de Pós-graduação em Gestão Pública. Universidade Federal do Espírito Santo.

Silva, G. P. (2013). Análise de evasão no ensino superior: uma proposta de diagnóstico de seus determinantes. Avaliação: Revista da Avaliação da Educação Superior, 18(2). DOI: https://doi.org/10.1590/S1414-40772013000200005

Silva, F. C. D. (2016). O desempenho acadêmico e o fenômeno da evasão em cursos de graduação da área da saúde. Congresso Brasileiro de Ensino em Engenharia. Natal. DOI: http://dx.doi.org/10.26512/2015.04.D.19947

Slhessarenko, M., Reis Gonçalo, C., Carlos Beira, J., & Cembranel, P. (2014). A evasão na educação superior para o curso de bacharelado em sistema de informação. Revista Gestão Universitária na América Latina-GUAL, 7(1). DOI: https://doi.org/10.5007/1983-4535.2014v7n1p128

Spady, W. G. (1970). Dropouts from higher education: An interdisciplinary review and synthesis. Interchange, 1(1), 64-85.

Spady, W. G. (1971). Dropouts from higher education: Toward an empirical model. Interchange, 2(3), 38-62. DOI: https://doi.org/10.1007/BF02282469

Tinto, V. (1975). Dropout from higher education: A theoretical synthesis of recent research. Review of educational research, 45(1), 89-125. DOI: https://doi.org/10.3102/00346543045001089

Tinto, V. (1993). Building community. Liberal Education, 79(4), 16-21.

Tinto, V. (1997). Classrooms as communities: Exploring the educational character of student persistence. The Journal of higher education, 68(6), 599-623. DOI: https://doi.org/10.1080/00221546.1997.11779003

Tontini, G., & Walter, S. A. (2014). Pode-se identificar a propensão e reduzir a evasão de alunos? Ações estratégicas e resultados táticos para instituições de ensino superior. Avaliação: Revista da Avaliação da Educação Superior, 19(1). DOI: https://doi.org/10.1590/S1414-40772014000100005

Vergidis, D., & Panagiotakopoulos, C. (2002). Student Dropout at the Hellenic Open University: Evaluation of the Graduate Program," Studies in Education". The International Review of Research in Open and Distributed Learning, 3(2). DOI: https://doi.org/10.19173/irrodl.v3i2.101

Villar, S. F. (2014). Evasão escolar no âmbito do Instituto Federal de Brasília. Revista Eixo, 3(2). DOI: https://doi.org/10.19123/eixo.v3i2.160

Vitelli, R. F. (2013). Evasão em cursos de licenciatura: perfil do evadido, fatores intervenientes no fenômeno. (Dissertação de Mestrado). Universidade do Vale do Rio dos Sinos, São Leopoldo.

Zewotir, T., North, D., & Murray, M. (2015).The time to degree or dropout amongst full-time master's students at University of KwaZulu-Natal. South African Journal of Science, 111(9-10), 01-06. DOI: http://dx.doi.org/10.17159/SAJS.2015/20140298

Appendix A - Profile of graduate students in the sample

Source: elaborated by the authors.

Variable	Master’s degree		Doctoral degree	Total
	Scientific	Professional
Gender	390	57	172	619
Male	172 (44%)	30 (53%)	80 (47%)	282 (45%)
Female	217 (56%)	27 (47%)	90 (52%)	334 (54%)
I do not wish to state	1 (0%)	0 (0%)	2 (1%)	3 (1%)
Age	390	57	172	619
Between 20 and 25 years old	115 (29%)	1 (2%)	18 (10%)	134 (22%)
Between 26 and 30 years old	112 (29%)	11 (19%)	51 (30%)	174 (28%)
Between 31 and 35 years old	76 (19%)	17 (30%)	38 (22%)	131 (21%)
Between 36 and 40 years old	38 (10%)	7 (12%)	24 (14%)	69 (11%)
Between 41 and 45 years old	24 (6%)	5 (9%)	24 (14%)	53 (9%)
Over 45 years old	25 (6%)	16 (28%)	17 (10%)	58 (9%)
Where do you live	390	57	172	619
In the program location city	264 (68%)	32 (56%)	111 (65%)	407 (66%)
Out of town of program location	126 (32%)	25 (44%)	61 (35%)	212 (34%)
Granting of scholarships	390	57	172	619
Scholarship holders	163 (42%)	8 (14%)	68 (40%)	239 (39%)
Non-scholarship holders	227 (58%)	49 (86%)	104 (60%)	380 (61%)
Works or does not work	390	57	172	619
I works	235 (60%)	49 (86%)	120 (70%)	404 (65%)
I do not work	155 (40%)	8 (14%)	52 (30%)	215 (35%)
If the student works, the company’s funding is:	235	49	120	404
Private	127 (54%)	39 (80%)	36 (30%)	202 (50%)
Public	108 (46%)	10 (20%)	84 (70%)	202 (50%)
Family group monthly income	390	57	172	619
Up to a minimum wage	0 (0%)	1 (2%)	0 (0%)	1 (0%)
Between one to two minimum wages	40 (10%)	0 (0%)	10 (6%)	50 (8%)
Between two to five minimum wages	131 (34%)	3 (5%)	38 (22%)	172 (28%)
Between five to ten minimum wages	113 (29%)	23 (40%)	65 (38%)	201 (32%)
More than ten minimum wages	106 (27%)	30 (53%)	59 (34%)	195 (32%)
Participate or not in family economic life	390	57	172	619
I participate	283 (73%)	56 (98%)	147 (85%)	486 (79%)
I do not participate	107 (27%)	1 (2%)	25 (15%)	133 (21%)

Appendix B - Scores given by different student profiles considering gender, age and scholarship

Source: elaborated by the authors.

Reason	Female		Male		Age equal to or less than the median (31 years old)		Older than median age		Scholarship holder		Non-scholarship holder
	Mean	Propensity degree	Mean	Propensity degree	Mean	Propensity degree	Mean	Propensity degree	Mean	Propensity degree	Mean	Propensity degree
1	4.5255	Medium	4.1844	Medium	4.4971	Medium	4.2130	Medium	4.3891	Medium	4.3579	Medium
2	6.1351	High	5.7660	Medium	5.7427	Medium	6.2491	High	5.2469	Medium	6.4237	High
3	5.0601	Medium	4.6525	Medium	5.0439	Medium	4.6968	Medium	5.1967	Medium	4.6947	Medium
4	4.0390	Medium	3.7730	Medium	3.8099	Medium	4.0578	Medium	3.8117	Medium	3.9895	Medium
5	4.0360	Medium	4.3759	Medium	4.0848	Medium	4.3610	Medium	4.0837	Medium	4.2868	Medium
6	5.0450	Medium	4.4681	Medium	4.9737	Medium	4.5848	Medium	4.9205	Medium	4.7237	Medium
7	4.8018	Medium	4.0177	Medium	4.5175	Medium	4.3935	Medium	4.5021	Medium	4.4368	Medium
8	3.1231	Medium	2.7624	Low	2.9942	Low	2.9458	Low	3.1381	Medium	2.8684	Low
9	2.8228	Medium	3.1809	Medium	2.8830	Low	3.1408	Medium	3.0000	Low	2.9974	Low
10	4.0480	Medium	4.1206	Medium	4.1491	Medium	4.0108	Medium	4.3264	Medium	3.9368	Medium
11	3.0030	Medium	3.2482	Medium	3.1842	Medium	3.0542	Medium	3.2469	Medium	3.0500	Medium
12	2.5616	Low	2.4291	Low	2.5058	Low	2.5451	Low	2.4937	Low	2.5421	Low
13	4.1832	Medium	3.9397	Medium	4.1140	Medium	4.0650	Medium	4.1381	Medium	4.0632	Medium
14	3.1892	Medium	2.4929	Low	3.0643	Medium	2.6715	Low	3.2510	Medium	2.6605	Low
15	3.8408	Medium	3.6560	Medium	3.8187	Medium	3.7256	Medium	3.7322	Medium	3.8053	Medium
16	4.0751	Medium	3.8546	Medium	4.0614	Medium	3.8917	Medium	3.8787	Medium	4.0526	Medium
17	3.3333	Medium	2.9965	Low	3.4006	Medium	2.9711	Low	3.4059	Medium	3.0842	Medium
18	2.4024	Low	2.2518	Low	2.3567	Low	2.3141	Low	2.5732	Low	2.1895	Low
19	5.8468	Medium	5.5922	Medium	5.4035	Medium	6.1661	Medium	4.6569	Medium	6.4289	High
20	3.2432	Medium	3.2589	Medium	3.2632	Medium	3.2671	Medium	3.2636	Medium	3.2658	Medium
21	4.6066	Medium	4.0638	Medium	4.2690	Medium	4.5126	Medium	4.7573	Medium	4.1395	Medium
22	4.2643	Medium	3.9184	Medium	4.3830	Medium	3.8014	Medium	4.4142	Medium	3.9395	Medium

Appendix C - Scores given by different student profiles considering work, income and residence

Source: elaborated by the authors.

Reason	I work		I do not work		Income of up to five minimum wages		Income above five minimum wages		Resides in the course location city		Resides outside the course location city
	Mean	Propensity degree	Mean	Propensity degree	Mean	Propensity degree	Mean	Propensity degree	Mean	Propensity degree	Mean	Propensity degree
1	4.3292	Medium	4.4465	Medium	4.7354	Medium	4.1641	Medium	4.2678	Medium	4.5660	Medium
2	6.3366	High	5.2791	Medium	5.8341	Medium	6.0455	High	5.8698	Medium	6.1604	High
3	4.5594	Medium	5.5070	Medium	6.0807	High	4.2172	Medium	4.5479	Medium	5.5425	Medium
4	3.9554	Medium	3.8558	Medium	3.8610	Medium	3.9545	Medium	3.8452	Medium	4.0660	Medium
5	4.3317	Medium	3.9767	Medium	4.1525	Medium	4.2399	Medium	4.1867	Medium	4.2500	Medium
6	4.7104	Medium	4.9674	Medium	5.3857	Medium	4.4697	Medium	4.7641	Medium	4.8679	Medium
7	4.4381	Medium	4.5070	Medium	4.6592	Medium	4.3510	Medium	4.3587	Medium	4.6604	Medium
8	2.7847	Low	3.3256	Medium	3.4619	Medium	2.6970	Low	2.9017	Low	3.1085	Medium
9	2.9455	Low	3.0977	Medium	3.0807	Medium	2.9520	Low	3.0713	Medium	2.8585	Low
10	3.9851	Medium	4.2791	Medium	4.2018	Medium	4.0227	Medium	4.2187	Medium	3.8349	Medium
11	2.9307	Low	3.4930	Medium	3.3139	Medium	3.0202	Medium	3.1351	Medium	3.1085	Medium
12	2.4381	Low	2.6837	Low	2.7758	Low	2.3813	Low	2.4447	Low	2.6745	Low
13	4.0842	Medium	4.1070	Medium	4.2511	Medium	4.0025	Medium	4.2138	Medium	3.8585	Medium
14	2.6361	Low	3.3628	Medium	3.4574	Medium	2.5682	Low	2.8133	Low	3.0330	Medium
15	3.8020	Medium	3.7302	Medium	3.9462	Medium	3.6818	Medium	3.8378	Medium	3.6604	Medium
16	4.0322	Medium	3.8977	Medium	4.2422	Medium	3.8409	Medium	3.9705	Medium	4.0142	Medium
17	3.0173	Medium	3.5674	Medium	3.5516	Medium	3.0152	Medium	3.1499	Medium	3.3208	Medium
18	2.2376	Low	2.5256	Low	2.5650	Low	2.2096	Low	2.2752	Low	2.4575	Low
19	6.4307	High	4.4558	Medium	5.2691	Medium	6.0126	High	5.6978	Medium	5.8349	Medium
20	3.2797	Medium	3.2372	Medium	3.3498	Medium	3.2172	Medium	3.1622	Medium	3.4623	Medium
21	4.0965	Medium	4.9070	Medium	4.7758	Medium	4.1540	Medium	4.1597	Medium	4.7972	Medium
22	3.8738	Medium	4.5907	Medium	4.6009	Medium	3.8535	Medium	4.0418	Medium	4.2783	Medium

Notes

Author notes

In Memoriam

The longing that Victor leaves us is enormous and equivalent to his academic brilliance and his human fullness, especially in the pleasant integration and maturity in relationships.

Corresponding author: Tel. +55 (31) 3409-7275 E-mail: jvacbr@yahoo.com.br(J. V. A. da Cunha); bcavelino@gmail.com(B. C. Avelino); edgardbc@usp.br(E. B. Cornacchione Junior) Universidade Federal de Minas Gerais. Av. Antonio Carlos 6627, Pampulha, Belo Horizonte/MG - 31270901, Brazil.