August 27, 2025
Significant changes in teaching practices have already been observed in the first modules of the training offered by the Aprender Linguagem project, according to an analysis conducted by researchers from Labedu and Harvard University.
Alyssa Gandolph, a research collaborator from Arizona (USA), is part of the team responsible for analyzing data from study in progress, led by Paula Cruz Pereira, a researcher at Labedu. The study investigates the effects of continuing education in early childhood education, focusing on transforming pedagogical practices in the classroom.
In an interview with Labedu, Gandolph discusses the main findings of the research, the methodological challenges of evaluating educational projects, and the evidence that indicates whether changes in teaching practices are, in fact, reaching children.
***
From your perspective, where does real change in teacher training occur? Which elements of teacher training seem most powerful?
I believe that teacher training involves a double transformation. One of the surprises we had was the significant leap between the initial internship and the Module 1. Initially, we imagined that the performance indicators would gradually improve throughout the project, but what we saw was significant growth early on, followed by stabilization in the subsequent modules. This made sense, since the Module 1 was closer to the first exposure to the practices and content of the methodology, still fresh in the teachers' memories.
I'm very excited to see that Labedu continues to evaluate the impact of its projects, because I suspect a second wave of transformation is occurring over the long term. In the research, we used a variable composed of several indicators that we expected teachers to develop throughout their training. While we saw significant growth between baseline* and the first module, there are less measurable aspects of teaching practice that are also crucial. As teachers become more confident with the content, the positive effects of the project may become even more evident in subsequent years—similar to what we see in the development of beginning and experienced teachers. This suggests that a second, more profound change may occur over time.
One of the most powerful elements of the Aprender methodology is that Labedu begins with the training of pedagogical coordinators, who are already leading figures within schools. Educators often resist training, with thoughts like, "This may have worked elsewhere, but will it work in my school, in my classroom?" Labedu works with coordinators who understand the local context and have legitimacy with teachers. This helps overcome a common barrier in training projects: ensuring not only faithful delivery of the content but also that it makes sense and is applicable to the reality of each school.
What are the main challenges you see in analyzing the implementation of a teacher training project like Aprender Linguagem, especially when the goal is to transform everyday classroom practice?
One of the biggest challenges in evaluating educational projects is identifying which training components are truly the most transformative. When we observe improvements in pedagogical practices related to content, we are encouraged by the results. But the question remains: what exactly in the training generated this improvement? And how can we enhance this aspect?
Another important point is that, although the Aprender program's content was the same, each municipality had a different schedule and began the project at different times. Some networks worked on the content in a more condensed format, over the course of a year. Others, over almost two years. This flexibility is necessary to adapt the project to local circumstances, but it makes analysis more complex, as it requires evaluating not only the overall impact of the training but also how each specific format influenced the results.
How can we better understand the relationship between the training offered and effective teaching practice?
One of the things we learned throughout the project came from Maria Grembecki, Labedu's Methodology and Projects Coordinator. When analyzing the results, we noticed that some teachers experienced a drop in proficiency after the initial gains. Module 1. Maria reminded us of something essential: teachers are also learners.
Many training projects are structured as one-off initiatives. Labedu's proposal, on the other hand, recognizes that the way teachers teach is directly linked to the way they learn. When we think of children as learners, we know that the learning process is not linear. Upon first encounter with a concept, there is rapid progress because the starting point is zero. But as the child evolves from beginner to expert, obstacles arise—the concepts become more complex. In our data, we saw significant growth toward proficiency initially, but there was no linear progression to complete mastery of more sophisticated practices. Some teachers regressed. Module 2 and grew again in subsequent modules. Others maintained stable performance but struggled to master higher-order concepts.
These findings reinforce the importance of the Labedu model, which offers ongoing training across several modules. Teachers need time to assimilate the content, revisit concepts that are still unclear, and resolve questions. This maturation period is essential to consolidate real changes in pedagogical practice.
What kind of data or signals help us understand whether changes in teaching practices are reaching children and making a difference in their learning?
It's difficult to directly measure the impact of early childhood education on children. Fewer assessment tools are available, and even if a standardized pre- and post-assessment tool were available across all municipalities, it would be necessary to isolate the effect of training from children's natural growth over time.
Therefore, we look for signs in teaching practices that are known to be associated with improvements in children's learning and use these indicators as proxies.** For example, Paula Cruz Pereira, the person responsible for the research, and I carefully analyzed how teachers conducted book reading activities. It wouldn't make sense to administer a reading test to children in this context, since the goal is to promote contact with books and written language—factors that are associated with future gains in language and literacy. Instead, we observe behaviors such as: Does the teacher allow children to participate? Does she direct questions and provide explanations to them during reading? The literature shows that early engagement with books is linked to language development. Thus, we use these behaviors as signals that children are engaged, and we know that engagement and interaction are essential steps for learning.
*It is the set of data collected before an experiment begins; it helps measure the impact of a variable or intervention.
**Proxy, in the research context, refers to an indirect indicator used as a substitute to measure or represent a phenomenon that is difficult to observe directly. These indicators do not directly measure learning, but they function as reliable surrogates to infer that it is occurring.
Find out more:
Labedu presents research results on Language Learning at the VII CONBAlf