Student Question-Asking Patterns In An Intelligent Algebra Tutor

Cognitive Tutors are proven effective learning environments, but are still not as effective as one-on-one human tutoring. We describe an environment (ALPS) designed to engage students in question-asking during problem solving. ALPS integrates Cognitive Tutors with Synthetic Interview (SI) technology, allowing students to type free-form questions and receive pre-recorded video clip answers. We performed a Wizard-of-Oz study to evaluate the feasibility of ALPS and to design the question-and-answer database for the SI. In the study, a human tutor played the SI's role, reading the students' typed questions and answering over an audio/video channel. We examine the rate at which students ask questions, the content of the questions, and the events that stimulate questions. We found that students ask questions in this paradigm at a promising rate, but there is a need for further work in encouraging them to ask deeper questions that may improve knowledge encoding and learning.