Is it possible to replace classroom observation with problem-solving interviews to evaluate teachers? An experience of the Brazilian Teaching Practice Assessment

Abstract Information: This article explores and discusses the limits and possibilities of using problem-solving interviews as an alternative to classroom observation for capturing the teaching practices of elementary school teachers. The analysis is based on the experience of Itaú Foundation in building an instrument that aims to measure the quality of teachers' pedagogical practices and generate formative feedback, both individually and for educational systems.
Based on the construction and application of an instrument to approximately 1,200 teachers in Brazilian municipalities, this study presents an innovative methodology for assessing teaching practices. The instrument consists of two online mediated interviews based on simulations of real classroom situations. The objective is to assess teachers' pedagogical practices in four learning dimensions:
Articulation between prior and new knowledge
Development of disciplinary competence
Knowledge with understanding and autonomy
Ways of relating to the world and socio-emotional learning
The methodology includes questions about how the teacher would deal with each simulated situation and an analysis tool/rubric for evaluating each teaching action foreseen in the taxonomy, with five performance levels:
Incipient
Low development
Medium development
High development
Exemplary
Based on the assessment of the interviews, individualized formative feedback is provided for each teacher, and a diagnostic analysis of the quality of the teaching practices of a school network is conducted.
The results of the instrument application showed us that it is sensitive to identifying high and low development points in teachers' pedagogical practices. Some teachers reported that participating in the interview was very beneficial for reflecting on their teaching practices, as they had never stopped to reflect on some of the practices presented in the interview situations.
However, through the applications, we were able to identify some elements for improvement, such as the length of the interview. Since they are one-and-a-half-hour interviews, we had difficulty scheduling times to conduct them. Therefore, one of the points we are working on is reducing the length of the interview without losing the sensitivity of the instrument.