Exploring the Use of Auto-Grading Systems to Improve the Efficacy of Feedback through Small, Scaffolded Programming Assignments
Angela A. Siegel, Tavis A. Bragg, Alexander Brodsky, Eric Poitras · 2021
In this panel, we will explore the use of auto-graded programming assignments to support timely and effective feedback to learners via small, scaffolded programming tasks. Program comprehension involves processing at different levels [5,10]. As students proceed through a program, associative processes take place and are described as information as the current statements activate information from the previous statements and from memory of prior knowledge. The most frequently inferred relations by the students are those that provide a coherent understanding of the state changes and outputs of a program [3] as well as the purpose of a piece of code [9]. The resulting, interconnected representation of the program goes beyond the syntax of tokens and statements. The outcome of successful comprehension is a representation that captures the meaning of each statement as students infer the operations of a statement, in terms of the underlying data and control flow, given its function in the context of solving a problem [4]. Although models of program comprehension generally agree regarding the processes by which a student arrives at a mental representation of a program, it is less clear how student-initiated processes play a role in program comprehension, and how they combine with such passive processes to result in comprehension.