Motivation: Explore the trade-offs in learning efficacy between completing fewer problems with guidance from a tutored problem solving system compared to seeing more worked problems without guidance.
Tutored problem solving: Computer systems can tutor students who are solving problems by providing them with hints and feedback at each step of the problem solving process. These kinds of systems, such as intelligent tutoring systems, generally improve problem solving performance. Using tutoring systems, however, is time consuming because they require students to attend to each step of the problem in depth, even if the student is not struggling with that step. This paper argues that the tutored problem solving approach can be helpful for students who need help, but it might hinder the progress of students who are proficient on their own. To address this question, the study compared low and high proficiency students’ learning gains on math problems between tutored problem solving (i.e., seeing fewer problem solutions but with step-by-step guidance) and untutored problem solving (i.e., seeing more problem and solutions but without step-by-step guidance) while controlling for time on task. The study used the ASSISTments system, which I am using for a current project and am loving.
Results: High proficiency students learned more when they saw more worked problems than when they used tutored problem solving. Low proficiency students were the opposite. They learned more when they saw fewer problems in tutored problem solving than when they saw more worked problems.
Why this is important: The results for the low proficiency students make intuitive sense. Students who would struggle with math performed better when they received guidance through computer-based tutoring than when they did not receive guidance. The results for the high proficiency students, however, are more interesting. These students performed worse when they received guidance on fewer problems than when they did not receive guidance but saw more worked problems.
This paper contributes to a growing body of evidence that providing too much guidance can be as detrimental to learning as providing too little guidance. Creating the best environment for learning means finding the correct balance of providing support so that students don’t flounder and withholding information so that students generate information on their own. As this paper points out, the correct balance can change from student to student. I expect that pre-tests and knowledge tracking will become critical for prescribing learning environments that are most beneficial to students, and thus, developing learning technologies.
Razzaq, L. & Heffernan, N. (2009). To Tutor or Not to Tutor: That is the Question. In Dimitrova, Mizoguchi, du Boulay & Graesser (Eds.) Proceedings of the 2009 Artificial Intelligence in Education Conference. IOS Press. pp. 457-464.
For more information about the article summary series or more article summary posts, visit the article summary series introduction.