Understanding Conceptual Transfer for Students Learning New Programming Languages

Prior research has shown that students face transition challenges between programming languages (PL) over the course of their education. We could not find research attempting to devise a model that describes the transition process and how students' learning of programming concepts is affected during the shift. In this paper, we propose a model to describe PL transfer for relative novices. In the model, during initial stages of learning a new language, students will engage in learning three categories of concepts, True Carryover Concepts, False Carryover Concepts, or Abstract True Carryover Concepts; during the transition, learners automatically effect a transfer of semantics between languages based on syntax matching. In order to find support for the model, we conducted two empirical studies. Study 1 investigated near-novice undergraduate students transitioning from procedural Python to object-oriented Java while Study 2 investigated near-novice postgraduate students doing a transfer from object-oriented Java to procedural Python. Results for both studies indicate that students had little or no difficulty with transitioning on TCC due to positive semantic transfer based on syntax similarities while they had the most difficulty transitioning on FCC due to negative semantic transfer. Students had little or no semantic transfer on ATCC due to differences in syntax between the languages. We suggest ways in which the model can inform pedagogy on how to ease the transition process.