Examining the Accuracy of a Conversation‐Based Assessment in Interpreting English Learners' Written Responses

Substantial progress has been made toward applying technology enhanced conversation-based assessments (CBAs) to measure the English-language proficiency of English learners (ELs). CBAs are systems that use conversations among computer-animated agents and a test taker. We expanded design capability prior instructional assessment developed CBA designed language skills mathematics knowledge middle school ELs. The prototype simulates an authentic engaging classroom where taker interacts with two virtual solve math problems. embedded feedback supports triggered by how interprets students' written responses. In this study, we administered ELs (N = 82) residing in United States. examined extent which system was able consistently interpret responses (722 for 82 students). study findings helped us understand factors affect accuracy system's interpretations shed light on improve incorporate scaffolding.