Loquat - Classroom Talk Time at Teacher’s Fingertips

• In Event: Harnessing AI for teacher empowerment: Teacher development, self-reflection, and lesson planning

Proposal

Currently, most approaches to teacher coaching are time-consuming and costly, particularly in contexts with scarce resources. Education systems, especially in low-and middle-income countries (LMIC), often do not have the capacity or the tools to provide consistent, targeted feedback to teachers. In addition, while teacher-led lectures and quiet work by students have their place in the classroom, a healthy amount of verbal interaction is essential. This applies in the early years when students are learning to read or when learning a new language, and in the adolescent years when they are developing personal identity, self-efficacy, and peer relationship skills. Yet, time-on-task studies conducted in several LMICs suggest that lessons continue to be dominated by teacher talk. To help address this challenge, RTI developed Loquat, a mobile app that allows teachers to autonomously monitor and better balance their talk time with that of their students.

With Loquat, teachers set their own talk time goals in advance of a lesson. They then capture audio from their lesson, using any available smartphone device. After the lesson, teachers upload the audio file to the Loquat server for analysis. Using machine learning, Loquat detects and classifies verbal interactions between teachers and students. Loquat does not analyze the content of the communication, but measures the time spent talking by children and adults. This allows the app to be used in different countries and settings, regardless of the instructional language. Loquat then translates these data into simple visualizations and feedback reports. Each of these reports aims to aid the teacher in initiating self-reflection, applying immediate corrective actions, and improving talk time balance in their classroom (e.g., proportion of teacher talk time versus student talk time, individual talk time versus group talk time).

As such, Loquat is designed as a tool that empowers the teacher to take immediate action to increase active learning, student participation, and inclusion in classrooms. The app can also be used as part of a standard coaching session where the coach, or peers, help the teacher with techniques in improving their talk time balance in the classroom. 

This presentation will present findings from pilots in Ghana as well as implementation under USAID’s Basic Education Quality and Transitions project in Guatemala and USAID’s Progresa con Educación project in Honduras. Discussion will focus on the qualitative feedback provided by teachers that speak to aspects of trust in technologies such as Artificial Intelligence and machine learning, invisible pedagogical mindsets, and self-reflection and feedback in the use and adoption of Loquat. In addition, the presentation will explore the technological aspects of deploying Loquat in these country contexts – for example, in the case of Honduras, where the software needed to analyze a range of child and adolescent voices in grade 6-9 classrooms. Finally, potential approaches to scaling the use of this software for teacher coaching nationwide, to support bilingual education, and to monitor instruction against scripted lesson plans will also be discussed.

Authors

• Carmen Strigel, RTI International

• Hétter Núñez