On How Novices Approach Programming Exercises Before and During Coding

Various tools and approaches are available to support undergraduate students learning to program. Most of them concentrate on the code and aim to ease the visualization of data structures or guide the debugging. However, in undergraduate introductory courses, students are typically given exercises in the form of a natural language problem. Deriving a correct solution largely depends on the problem-solving strategy they adopt rather than on their proficiency in dealing with the syntax and semantics of the code. Indeed, they face various challenges (apart from the coding), such as identifying the relevant information, stating the algorithmic steps to solve it, breaking it into smaller parts, and evaluating the implemented solution. To our knowledge, almost no attention has been paid to supporting such problem-solving strategies before and during the coding. This paper reports an interview and a sketching exercise with 10 participants exploring how the novices approach the programming exercises from a problem-solving perspective and how they imagine a tool to support their cognitive process. Findings show that students intuitively perform various actions over the exercise text, and they would appreciate having support from the development environment. Accordingly, based on these findings, we provide implications for designing tools to support problem-solving strategies