Principal investigator: Adriano Valladão Pires Ribeiro
Article title: The Effect of Teacher Training on Proficiency at the Beginning of Elementary School
Article authors: Geraldo Andrade da Silva Filho
Location of the intervention: Brazil
Sample size: 71.758 classrooms (around 320 students per subject)
Main theme: Education
Type of Intervention: Teacher training/development
Main Variable of Interest: Gains in proficiency in Portuguese Language and Mathematics
Evaluation method: Experimental Evaluation (RCT)
Policy Problem
A student's academic and professional success is linked to the quality of the teachers who taught them. Factors determining teacher quality can be related to observable characteristics, such as experience and education, and unobservable ones, such as interpersonal skills and classroom management. INEP researcher Geraldo da Silva Filho contributes to this discussion by measuring the impact of higher education among fourth and fifth grade teachers in public elementary schools on their students' proficiency.
Evaluation Context
Primary school teachers can be qualified in three ways according to their training: they can have higher education compatible with the area they teach, another higher education qualification, or no higher education qualification. The National Education Plan 2014-2024 foresees a national policy for teacher training, ensuring that all basic education teachers have specific higher education training in their area of expertise, that is, that all teachers have a degree in the specific area or in Pedagogy. This led federal, state, and municipal governments to launch incentive programs in order to comply with the public policy guideline. Therefore, measuring the impact of teacher higher education on student proficiency becomes even more relevant.
Intervention Details
To study this issue, several INEP databases were combined. The sample consisted of all students enrolled in public schools who took the 2015 Prova Brasil (PB) and the 2013 National Literacy Assessment (ANA). Additionally, students who had more than one teacher, either in the fourth grade in 2014 or the fifth grade in 2015, were excluded so that the student's results could be attributed to the teacher's educational background. Finally, for each student in this description, the higher education level of their Portuguese Language and Mathematics teachers was associated. Teacher information was obtained from the 2014 and 2015 Basic Education School Censuses (CEB), respectively for the fourth and fifth grades.
The Prova Brasil is administered to fifth and ninth grade elementary school students and aims to evaluate students in Portuguese language and mathematics. Along with the test, there is also a socioeconomic questionnaire to link context to student performance. The tests for the National Literacy Assessment, in 2013, were applied to third grade elementary school students. Half of the students in each class answered the Reading exam and the other half the Mathematics exam, with the distribution for each test done through random selection.
Methodology Details
The effect of teacher training on their students' proficiency in Mathematics and Portuguese Language was captured through joint learning throughout the 4th and 5th years of elementary school, based on the scores obtained in the 2015 PB (Brazilian Portuguese Assessment) applied to fifth-grade students. In other words, the aim is to estimate the added value of teachers with higher education in student learning. However, there are difficulties in carrying out such measurement. The main problem lies in isolating the effect of teacher training itself from other factors, that is, confusing what is caused by teacher training with other reasons such as student characteristics, school characteristics, and even other teacher characteristics (experience, postgraduate studies, etc.).
In particular, two factors deserve discussion. First, the student's learning process is cumulative and built up over the years; the performances obtained in the fourth and fifth grades should be conditioned by their entire school, family, and individual skills history. In practice, however, it is not possible to observe all these factors, so the ANA score applied in the third grade is used, and it is assumed that it is sufficient to aggregate all the student's previous information up to that point. Second, the assignment of teachers to students may not be random; that is, the allocation of the teacher could be based on their skills and compromise the measurement of the effect of the teacher's training on student performance. The SAEB Principal's Questionnaire helps to resolve this limitation by providing information on the criteria used to form a class and assign its teacher.
Having information about students' grades in the 2015 PB and 2013 ANA when they were in the fifth and third grades, respectively, their socioeconomic characteristics, the characteristics of the class and school they attended, and their teachers' training, it is possible to capture the effect of the teacher's higher education on student progress between the two assessments. This exercise is done separately for the subjects of Mathematics and Portuguese Language, the two subjects assessed in the ANA and PB, investigating, initially, only the effects of higher education and, subsequently, the effects of specific higher education training.
Results
The first investigation, into whether any higher education affects the academic proficiency in Mathematics and Portuguese Language of students in the public school system, concluded that there are no significant effects; that is, students who had teachers with higher education in the fourth and/or fifth grade performed similarly to those who had classes with teachers without higher education. This result considers the average among all teachers with higher education; the following exercise disaggregates higher education into specific training in Mathematics or Portuguese Language, training in Pedagogy, and training in another higher education course.
Students who studied mathematics with teachers trained in that area had an average performance 4 points higher on the SAEB scale in the Prova Brasil exam when compared to students who had classes with teachers without higher education. The second largest positive effect occurs for those who had at least one of the two years of study with a teacher trained in mathematics. For the analogous case of Portuguese Language, the same cannot be observed, with zero proficiency gains for all training levels. In other words, the disaggregation captured an effect of improved student performance only when the specific training is in mathematics.
Lessons in Public Policy
There is a cost to encouraging and training elementary school teachers to pursue higher education in accordance with the guidelines of the National Education Plan. However, higher education does not guarantee better teaching, as quality is also linked to factors such as knowledge of the content taught, the pedagogical practices adopted, and classroom management. In the case of the study presented, only students with teachers trained in mathematics managed to improve their performance on the mathematics section of the Prova Brasil exam.
Reference
Silva Filho, Geraldo Andrade da. “Effect of Teacher Training on Proficiency at the Beginning of Elementary School.” Brazilian Journal of Economics, v. 73, n. 3, p. 385-411, 2019.
