Exploring CS1 Student's Notions of Code Quality

Abstract

Coding tasks combined with other activities such as Explain in Plain English or Parson Puzzles help CS1 students to develop core programming skills. Students usually receive feedback of code correctness but limited or no feedback on their code quality. Teaching students to evaluate and improve the quality of their code once it is functionally correct should be included in the curricula towards the end of CS1 or during CS2. However, little is known about the student's perceptions of code quality at the end of a CS1 course. This study aims to capture their developing notions of code quality, in order to tailor class activities to support code quality improvements. We directed students to think about the overall quality of small programs by asking them to rank a small set of solutions for a simple problem solving task. Their rankings and explanations have been analysed to identify the criteria underlying their quality assessments. The top quality criteria were Performance (64%), Structure (51%), Conciseness (42%) and Comprehensibility (42%). Although fast execution is a key criteria for ranking, their explanations on why a given option was fast were often flawed, indicating students need more support both to evaluate performance and to include readability or comprehensibility criteria in their assessment.