The combination of modern computing power, the interactivity of web applications, and the flexibility of object-oriented programming may finally be sufficient to create computer coaches that can help students develop metacognitive problem-solving skills, an important competence in our rapidly changing technological society. However, no matter how effective such coaches might be, they will only be useful if they are attractive to students. We describe the design and testing of a set of web-based computer programs that act as personal coaches to students while they practice solving problems from introductory physics. The coaches are designed to supplement regular human instruction, giving students access to effective forms of practice outside class. We present results from large-scale usability tests of the computer coaches and discuss their implications for future versions of the coaches.
Ryan, Qing X., Evan Frodermann, Kenneth Heller, Leonardo Hsu, and Andrew Mason. "Computer problem-solving coaches for introductory physics: Design and usability studies." Physical Review Physics Education Research 12, no. 1 (2016): 010105.