ABSTRACT
Natural language conversations have been utilized for their effectiveness in intelligent tutoring systems for over a decade. In the present chapter, we discuss expanding this effectiveness to the context of assessment. Conversation-based assessments allow for the assessment of deeper knowledge, skills, and abilities by requiring students to generate their own responses and providing follow-up questions that further probe students’ understanding of key constructs. In addition to expanding automated conversations to the assessment context, we also propose expanding the scope of what can be assessed. Specifically, we propose utilizing students’ rich, natural language responses to assess their communication skills.
