A Case Study of Writing to Learn to Program: Codebook Implementation and Analysis
Mahnas Mohammadi-Aragh, Phyllis J. Beck, Amy Barton, Bryan A. Jones · 2020
Abstract Introductory computer science courses continue to be a barrier to engineers who are pursuing a computing degree pathway. Learning to program requires students to self-monitor and self-assess their programming metacognition. Based on the idea that writing is a visual representation of thinking, Writing-to-learn (WTL) activities are usually short, low-stakes writing assignments that are designed to promote reflection, analysis, synthesis, and deeper understanding of course material. WTL activities can promote metacognition in any discipline. The research discussed in this paper focuses on answering the following two research questions: RQ1) What do source code comments tell us about novice programmers’ organizational and thinking processes while coding? RQ2) How do WTL activities impact novice programmers’ organizational and thinking processes? In this paper, we present a case study that demonstrates how we used WTL techniques and the literate programming paradigm in an introductory programming course. Next, we detail our analysis using three laboratory assignments to describe how we analyzed student submissions to create a qualitative codebook for students’ Visual Organization Strategy and Thinking Processes. Students’ Visual Organization Strategy is a top-level overview of the organizational structure of students’ code and the underlying characteristics such as white-space, commenting patterns, and blocks of code that students use to visually communicate the structure of their code. We identified five Visual Organization Strategies: Block-level, Unitization, Every-line, insufficient and None. Student’s Thinking Processes are a line-by-line determination of the level of programming metacognition and strategic knowledge. Our efforts resulted in six categories of Thinking Processes: Organizational, Reflective, Conceptual, Literal, Insufficient, and None. We conclude the paper with initial classification results for WTL sections, and a description of future applications of our research.