Estudos sobre o desempenho acadêmico contábil: influência do mercado de trabalho e análise sobre diferentes níveis de desempenho

Abstract

This paper discusses accounting performance in two essays. In the literature on accounting education, the Accounting Sufficiency Exam (ESC) is commonly used as a performance proxy since it is a prerequisite for professional certification in the market. However, the ESC has a low approval rate, which encourages discussions about the factors that influence these results. The influence that institutional and educational factors can exert on the performance of students in the exam has already been identified and discussed. However, the first research gap pointed out concerns about the fact that there is no discussion about the influence that the labor market can have on performance in this exam. In this sense, the first essay aims to contribute to the literature and advancing in this perspective, with the objective of verifying if the economic situation of the labor market influences the performance of the students in ESC. In order to respond to the proposal, inflated beta regression models were used, considering two editions of the ESC that occurred in the same year. The findings show that the labor market situation influences the ESC approval rate, an unprecedented result in the literature. In a complementary way, the work also discusses the variables that influence the dispersion in the performance data, as well as the variables that influence the presence of 0 (total failure) and 1 (total approval) in the exam approval rates. The results obtained in the first essay are promising for future discussions about academic performance, presenting a channel of influence beyond the widely discussed characteristics of educational institutions. The second shortcoming concerning accounting education refers to the fact that the regression models used generally analyze the influence of several variables on the average performance, neglecting potential differences that exist along the different test scores. In view of this, no national studies were identified that deal with the measurement of accounting performance in ESC considering the different levels (low, medium and high) of performance, so the second essay aims to identify which factors influence the low, medium and high performance of students in the Accounting Proficiency Exam. We used quantile regression to obtain answers to the question raised and considered the combination of two editions of the ESC for an annual analysis of the exam. It can be discussed about the findings that the variables do not present a constant behavior at all levels of performance, that is, they have influence on a certain quantile of performance and do not present on others. These results help the discussion about accounting performance, expanding the disparate view on what has been used in the elaboration of educational policies, attesting that the form of influence of a variable is not the same at all levels of performance in ESC and, therefore, the benefits of this contribution to academia should not be overlooked.