SBIR Phase I: The Teacher Practice Feedback Tool- real-time information for teachers who want to measure and improve their talk time, wait time, and question-asking skills.

Period of Performance: 07/01/2016 - 12/31/2016


Phase 1 SBIR

Recipient Firm

Earshot LLC
1100 NE Campus Pkwy
Seattle, WA 98105
Firm POC
Principal Investigator


This SBIR Phase I project seeks to improve teaching by providing teachers with real-time, personalized feedback. Most teachers receive little regular, actionable feedback about their teaching to help them improve. Questions and discussions are the currency of great teachers, especially as teachers help students develop critical thinking, collaboration, communications and creativity skills, those skills in highest demand for careers in science, technology, engineering and math. This project is to design, develop, test and refine a mobile app that provides powerful, real-time instructional data for teachers at all grade levels and subject areas. Through the automatic analysis of questions and discussion, teachers will have access to useful data about communication in their classrooms to help them continuously improve instruction. This project is designed to help teachers incorporate more inquiry-based instruction methods into their teaching practice. Inquiry is a research-based approach to increasing student competencies. The intended outcomes for using this application are for teachers to talk less and ask more questions, to allow sufficient wait time before calling on students, and for teachers to incorporate more and higher-level questioning into their teaching. The technology can be sold to directly teachers at schools and universities throughout the United States on a subscription basis, enabling teachers to control their own data and decide how it is used, as well as ensuring that the tool is used for improvement, rather than for evaluative purposes. The technology being developed for this project builds upon recent advancements in natural language processing and voice analysis. This project uses machine learning and algorithms to develop a user-friendly app that analyzes teacher voice data around four key data points: talk time, wait time, question level (higher-order and lower-order questions based on a proprietary taxonomy), and question frequency. The objective of this project is to design, test and refine the first version of the app, to be used on an iPhone using an external microphone. Teachers representing various grade levels, subject areas, levels of experience, and voice characteristics will pilot this technology and provide feedback to inform development. This project will employ research methodologies such as pre and post surveys, observations, interviews, and think-aloud sessions, to inform usability and feasibility.