We are proposing to record data of in-class interactions among preschoolers and their teachers at the Child Development Laboratory, in order to train better AI models of preschool speech, language, and gesture. If the IRB proposal and Data Use Agreement are approved, they will be posted here, so that participants in the study can easily find them and review them.
Children with speech or language related concerns risk falling behind in their academic and social-emotional development. Many children in need of specialized speech-language instruction never receive it because of the shortage of speech and language pathologists (SLPs). The National AI Institute for Exceptional Education seeks to empower SLPs by creating improved artificial intelligence (AI) technologies that can help the SLPs to rapidly identify children in need of specialized instruction, and to track the effectiveness of different types of interventions. AI recognizes spoken words and visual communicative behaviors by comparing the new examples to examples in its training database, therefore an effective AI screener for preschool communicative disorders needs to have a very large training database of normal preschool communicative behavior. This research project is recording a training database of normal preschool communicative behaviors and will make it available to researchers under a data use agreement that guarantees participant privacy while maximizing the benefit to science and education.