Adaptive Parsons Problems as Active Learning Activities During Lecture

Abstract

Adaptive Parsons problems could be used to reduce the difficulty of introductory programming courses and increase the use of active learning in lecture. Parsons problems provide mixed-up code blocks that must be placed in order. In adaptive Parsons problems, if a learner is struggling to solve a problem it can dynamically be made easier. This makes it possible for students to correctly solve a problem in a limited time, even if they are struggling. Previous research on the effectiveness and efficiency of solving Parsons problems for learning has been conducted in controlled conditions or in lab/discussion. We tested the efficiency of solving adaptive Parsons problems versus writing the equivalent code as lecture assignments through three between-subjects experiments. The median time to solve each Parsons problem was less than the median time to write the equivalent code for all but two of the problems, both with complex conditionals. However, that difference was significant for only six of the 10 problems. Our hypothesis for why two problems had a higher median time to solve as a Parsons problem than as a write code problem was that the problem instructions did not match the Parsons problem solution and/or they also had a large number of possible correct solutions. Results from student surveys also provided evidence that most students (78%) find solving adaptive Parsons problems in lecture helpful for their learning, but that some (36.2%) would rather write the code themselves. These findings have implications for how to best use Parsons problems.