Developing Indicators of Code Quality For Block-Based Programming

• In Event: Electronic Board Session IV

Abstract

We present an automated approach to analyzing code structures of block-based programs and developing indicators of code quality: complexity, content, and structure. Using sample responses for debugging tasks from the Scratch repository, we validate how our indicators reflect a priori categorizations of four debugging patterns and present preliminary findings.

Authors

• Tianying Feng, UCLA/CRESST

• Gregory Chung

• Charles Parks, CRESST/UCLA