Exploring the Effects of Contextualized Problem Descriptions on Problem Solving

Abstract

Prior research has reported conflicting results on whether the presence of a contextualized narrative in a problem statement is a help or a hindrance to students when solving problems. On the one hand, results from psychology and mathematics seem to show that contextualized problems can be easier for students. On the other, a recent ITiCSE working group exploring the “problem description effect” found no such benefits for novice programmers. In this work, we study the effects of contextualized problems on problem-solving in an introductory programming course. Students were divided into three groups. Each group was given two different programming problems, involving linear equations, to solve. In the first group both problem statements used the same context while in the second group the context was switched. The third group was given problems that were mathematically similar to the other two groups, but which lacked any contextualized narrative. Contrary to earlier findings in introductory programming, our results show that context does have an effect on student performance. Interestingly depending on the problem, context either helped or was unhelpful to students. We hypothesize that these results are explained by a lack of familiarity with the context when the context was unhelpful, and by poor mathematical skills when the context was helpful. These findings contribute to our understanding of how contextualized problem statements affect novice programmers and their problem solving.