Investigating the Utility of self-Explanation for Learning to Program through Translation Activities

Programming is challenging for novices to learn. Code tracing is a foundational programming skill that involves simulating a program's execution. Translating lines of code into plain English explanations of what it does should allow students to better understand a program while code tracing. Translations is similar to self-explanation, a general mechanism that helps learning by making inferences beyond the instructional materials. This thesis investigates whether self-explanation via translation of programming syntax to plain English helps students learn to code trace, in a between-subjects study (\emph{N} = 44). We created two versions of a computer tutor that guides novice programmers through code-tracing problems. In the experimental condition, students were asked to translate lines of code during code-tracing activities, while in the control condition students were only asked to code trace. The experimental group performed significantly better on translation and code-generation questions, but the control group performed significantly better on code-tracing questions