Teaching Mental Models in Introductory Programming

Abstract

In this article, an evaluation of syllabi, books, teaching materials, and examinations concerning introductory programming revealed that the subject disproportionately focuses on teaching and learning language features instead of mental models. It is demonstrated that the shift from low-level to high-level languages resulted in the "language-trap" that leads to an emphasis on language features at the expense of mental models. To mitigate the language-trap effect, a novel instructional approach called MTL three-tier that combines low-level syntax, Memory Transfer Language, and high-level syntax is proposed. Results from two experiments show that using assembly codes in combination with the MTL-three-tier approach at the beginning of the course assists instructors in avoiding the language-trap. For novices, the cognitive load is reduced, consequently, increasing the ability to form viable mental models. Results from the first experiment show that novices in the experimental group were 2.17 times more likely to form viable mental models than those in the control sample. From the second experiment, results show that novices from the experimental group were 14.50 times more likely to avoid common errors in introductory programming than were the novices in the control group.