University of Minnesota
School of Physics & Astronomy


Computer Coaches

Qing Xu Ryan
Qing Xu Ryan
Alex Schumann

Qing Xu Ryan is a graduate student in Professor Ken Heller’s Physics Education program. She has participated in designing and building computer coaches to help with problem solving in physics and conducted experimental studies to assess the effectiveness of these coaches on over 400 students.

The computer coaches were implemented in introductory Physics 1301 to help students with some mechanics problems. The coaches were built within a cognitive apprenticeship framework. “In a traditional apprenticeship a tailor learns from another tailor by watching them do it, having them help, even holding their hands to coach the correct movements.” Qing says that the idea is to incorporate this framework into the modern classroom practice and the hope is to achieve some effective coaching using the computer coaches. In the current setting of the introductory physics, the professors model problem solving for students in the lecture and the Teaching Assistants often function as coaches. The computer will never substitute for the human element, Qing says, but it can be a beneficial supplement. “The computer coach has the advantage of being available 24/7 and it’s not judgmental, so you can try as many times as you like and not feel bad if you aren’t getting it.”
It is important to make the problem solving process explicit. “For a lot of physicists, the decision making they do while solving a problem is not explicit to novices,” Qing says, “however, they all more or less follow a coherent problem-solving framework, no matter if they do it consciously or not”. Such a problem solving framework is incorporated into the computer coaches and is made explicit to the student. The other element in the cognitive apprenticeship model is “fading”, like the process to helping a child learn to ride a bicycle. At a certain point the parent needs to let go and let them find balance on their own. Once the student gets better with problem solving, the amount of help is gradually reduced. The computer coach is flexible and can emphasize different stages/aspects in teaching problem solving.

The first version of the computer coaching software consists of 35 problems and was implemented on a large scale to over 400 students. With an average completion time of 30 minutes per problem, the coach is easy to use and not a burden on students’ time. According to the survey feedbacks, students perceive them as useful and ranked the coaches among the top three useful components of class. Qing says that students in introductory physics courses might see more coaches in classrooms in the near future.

A problem solving rubric was used in this study to assess the effectiveness of the coaches. “One of the challenges in this research is to conduct a good experimental study, like any other educational research”, Qing says, “for example, getting enough statistics and forming good experimental and control groups.” Qing says that further research will include improving the experimental design as well as developing better iterations of these coaches—a project in progress.

To see the problem solving coach in action, select one of the coaches from the list here