Technologies like automatic speech recognition (ASR), natural language processing (NLP), and voice-based analytics are increasingly embedded in educational tools and research platforms. Speech and audio data are opening new opportunities for how researchers study reading, language development, and classroom interaction. The data systems and opportunities for research are new and hold great promise for greater insight and programming in the future.

Supported by the Institute of Education Sciences, the U-GAIN Reading Center hosted Speech Data for Learning Scientists 101, a free, two-day virtual convening intended to help graduate students explore emerging uses of speech technology and build capacity to use large-scale datasets. For the convening, the U-GAIN team brought together a wide range of speakers, including university researchers, edtech chief AI scientists, school district leaders, and learning scientists. Presenters examined how speech data can move beyond the basics toward understanding student expressiveness, prosody, and real-time engagement.

This convening is part of the U-GAIN Reading national leadership mission to build research capacity at the intersection of AI, the Science of Reading, and literacy instruction.

Playlist for Day 1: Speech Data for Learning Scientists 101

Day 1 focused on the following questions: What are speech data? How are they being used in education? What research opportunities do speech data support? Access the youtube playlist.

Playlist for Day 2: Speech Data for Learning Scientists 101

Day 2 focused on the following questions: How do I actually build a research project that uses speech data? What does it look like to work with speech models? Where are the most promising opportunities now across research and industry? Access the youtube playlist.