A Multiple Executions Technique of Visualization

Abstract

Algorithm visualizations can be structured or presented in different ways: animation, multiple algorithms, etc. In this paper, we present a presentation format that has hardly been explored, namely to display simultaneously visualizations of a given algorithm for several test cases. By similarity with other presentation formats, we call it the "multiple executions" technique. We illustrate this presentation format with a history-preserving graphical representation, namely recursion trees. Multiple executions composed of recursion trees provide at a glance a view of the behavior of an algorithm in different situations. We illustrate their application to a number of issues in algorithm courses. We also report on two evaluations conducted on the effects of structured multiple executions on students' comprehension of complex recursive algorithms. The results show some statistically significant differences on students' skills for some design tasks.