The Use of Artificial intelligence to Automatically Measure Children’s Communication during a Standardized Play-based Assessment

The Early Communication Indicator (IGDIs ECI website) is an internationally-used standardized observational measure with strong psychometric properties. It is designed to support screening and progress monitoring of children 6-42 months of age. Like all standardized observational assessments, the ECI requires training and staff time to administer and score each assessment, which most infant-toddler programs and pediatric offices cannot afford.

Through this proof-of-concept study, we aim to demonstrate the use of Artificial Intelligence (AI) to automatically score child communication during ECI assessments to minimize the need for human coding and provide nearly immediate scores. We are working with community partners to train and test the AI algorithm to score child behaviors during the ECI assessment. This will also include the development and initial testing of a web-based application where parents and practitioners will be able to upload a recorded ECI assessment and receive scored results much faster than using human coding.