Rapid Student Feedback 

Learning to program is difficult, especially with limited instructor time and resources. One way to combat these difficulties is to provide a lot of practice accompanied by timely feedback. However, giving large amounts of feedback is difficult because instructor time in classrooms is limited. Even in a classroom of 20 students, giving feedback on 5 or 6 programming problems can be a difficult, especially if we want to give timely and impactful feedback, and especially as classroom sizes continue to increase. So to address the scale of current classes and give students timely feedback, one way to approach it is to design impactful feedback based on observed student performance and automate it’s delivery so feedback can be given at scale at the moment when the student needs it as opposed to before or after.

Goals

1. Development of instructor tools where instructors can write template code that mirrors student code that can be used to deliver feedback for specific programming mistakes
2. Analysis of student programming mistakes into a compiled library of student mistakes
3. Quantification of the impact of different designs of feedback on student learning

Key Elements

Immediate Feedback; Computer Science Education; Abstract Syntax Trees; Instructional Design; Assessment

Skills

Python, Java, JavaScript, Docker, Linux, GitHub, Instructional Design (not necessarily required, but having more of these skills is better). 

Projects

• Develop a Java version of Pedal (Pedagogical libraries) for Python
• Develop a Knowledge Component Map for existing Computer Science curriculum
• Design and develop a meta-tagging system for feedback
• Design and develop feedback and feedback delivery for existing assignments and new assignments
• Design and develop reusable feedback libraries
• Analyze student coding data using Feedback patterns