Evaluating the Impact of Adaptive Parsons Problems on Students in Introductory Computer Science Courses

Abstract

It's no surprise that students in introductory computer science courses face numerous challenges: for many, it's their first time ever programming, and computer science courses tend to be difficult and time-consuming. Parsons problems, problems where students piece together given code to write programs, solve some of these student issues: they have been shown to be more time efficient, yet just as effective for learning as traditional programming problems. However, in introductory computer science courses, students come in with varying skill sets and backgrounds. In response, some researchers have explored the space of adaptive Parsons problems, problems that provide varying levels of scaffolding based on students' abilities, continuing to adapt in difficulty based on students' progress. This paper will extend the work done on adaptive Parsons problems in two ways: by applying it to block-based programming languages, which are often used in introductory computer science courses, and by assessing its impacts on student confidence and attitudes towards the material. This study will use both intra-problem adaptive Parsons problems, where the difficulty of the existing question can be adjusted, and inter-problem adaptive Parsons problems, where the difficulty changes across problems based on student progress.