PPIG 2022 - 33rd Annual Workshop
Mental Models of Recursion: A Secondary Analysis of Novice Learners’ Steps and Errors in Java Exercises
Natalie Kiesler
Abstract: Teaching and learning recursive programming has been the subject of numerous research projects and studies. However, few research publications focus on learners’ steps while solving recursive tasks and the corresponding identification of mental models. It is the goal of this secondary analysis to identify the challenges novice learners of Java encounter in recursive problem solving and to map them to the mental models and conceptions from the literature. The investigated dataset was collected via thinking aloud experiments with eleven first-year-students of computer science in a professional usability laboratory. Students had to recursively compute the factorial of n and the Fibonacci sequence in a learning environment. By using deductive categories from the literature, the students’ performance was evaluated in terms of their programming steps, their challenges/errors, and thus their ability to generate a recursive function. The results show that mental models can partially be identified via the analysis of students’ problem solving steps and errors. Moreover, recursive tasks with more than one recursive call are more challenging for novice learners. The passive flow of control along with the end of the recursion chain also seem to be counterintuitive for learners. The lack of viable, complete mental models implies the need for further educational research on instructional methods addressing these challenges of first-year students. Learning to program may be easy, but novices require fine-grained, step-by-step scaffolding and instruction, as well as time to understand and apply the more abstract concepts, which include recursion.