Understanding algorithms by means of visualized path testing_

Abstract

Visualization of an algorithm offers only a rough picture of operations. Explanations are crucial for deeper understanding, because they help the viewer to associate the visualization with the factual meaning of each detail. We present a framework based on path testing for associating instructive explanations and assignments with a constructive self-study visualization of an algorithm. The algorithm is divided into blocks, and a description is given for each block. The system contains a separate window for code, flowchart, animation, explanations, and control. Assignments are based on the flowchart and on the coverage conditions of path testing. Path testing is expected to lead into more accurate evaluation of learning outcomes because it supports systematic instruction in addition to more free trial-and-error heuristics. A qualitative analysis of preliminary experiences with the prototype indicates that the approach helps a student to reflect on her own reasoning about the algorithm. However, a prerequisite for an successful learning process with the environment is a motivating introduction, describing both the system and the main idea of the algorithm to be learned.