Improve Feedback Mechanism in Programming Assessment Systems with Progress Indicators and Reward to Foster Students' Growth Mindset: (Abstract Only)

When students learn programming, the assignment feedback information from current automatic programming assessment systems, such as Web-CAT [1] is often negative, objective, and unfriendly. These feedback information can easily frustrate students to lose interest in programming related activities. The negative feedback information can have possible serious consequences to students. We work to improve current feedback mechanism in mindset perspective: encourage students by positive feedback with a group of fifteen progress indicators and possible reward. The fifteen progress indicators were designed and implemented based on students' sequential programming submissions. These fifteen indicators include seven general purpose indicators about various aspects when students construct solutions for assignments; eight other software testing indicators concentrate on students' progress when students self-checking their code [3]. We did statistical analysis for these fifteen indicators' suitability to a collection of programming assignments data set including 257 students. In order to validate fifteen progress indicators' effectiveness, we also apply a student performance model: Recent-Performance Finite Analysis model (R-PFA) [4] to the same programming assignment data set we used before. We calculate R-PFA model's prediction accuracy and apply learning curves analysis. In learning curve analysis, eight software test indicators demonstrate students gradually learn positively when they work on their assignment submissions. Based on progress indicators information, we plan to give students possible reward when they make progress. We will research on reward mechanism, reward format, and timing, etc. In this way, moves students to growth mindset [2] - belief that hard work and practices can improve their skills and capabilities.