Designing boosters and recognition to promote a growth mindset in programming activities

Abstract

When one first learns to program, feedback on early assignments can easily induce a fixed mindset-where one believes programming is a fixed ability you either have or you don't. However, possessing a fixed mindset perspective has negative consequences for learning. The alternative is to foster a growth mindset, where one believes ability can be improved through practice, effort, and hard work. However, automated grading tools used on programming assignments currently focus on objectively assessing functional correctness and other performance-oriented features of students' programs. Unfortunately this encourages students to adopt performance-oriented goals, which are characteristic of a fixed mindset. By building on existing measures of “productive effort”, we design a new kind of feedback approach that focuses on recognizing, encouraging, and rewarding diligence and productive actions based on those indicators. The goal is to add such elements to existing feedback in an emotionally supportive way that recognizes the efforts of a student expending and valuing these practices. The feedback design presented here consists of two main components: textual/verbal feedback that recognizes productive effort students spend on a problem, or that encourages students to be strategic about expending effort to improve their own skills. The point of this feedback is to convey to the student that constructive practice to improve one's skills is valued and recognized, independently of the final product they are creating. In addition to the textual feedback, the feedback also includes boosters, or rewards in the form of perks that enhance parts of the student work experience. By taking inspiration from video game psychology and other sources, we designed a booster-based reward system that recognizes hard work without tacitly promoting performance-oriented (score-oriented) motivation. In addition to describing the design of the reward and recognition feedback strategy and the variable ratio reinforcement schedule on which the strategy is based, we also present a post hoc analysis of the results obtained when applying this strategy to existing student submission data. This allows investigating what feedback or boosters would have been earned by individual students in a real-life situation to validate the feedback design before live deployment.