Software visualizations for improving and measuring the comprehensibility of source code

Experiments were performed to measure the relationship of two software visualizations to software comprehension. The Control Structure Diagram (CSD) was examined for its influence on aiding the comprehension of software source code, and the Complexity Profile Graph (CPG) was examined for its ability to measure source code comprehensibility. Test subjects were asked to respond to questions regarding a code specimen written in Java. Statistical analysis of the data collected from the first experiment revealed that the CSD was highly significant in enhancing the subject’s performance in this program comprehension task. Similar analysis of data collected from the second experiment showed that the measure of complexity calculated by the CPG correlated with the amount of time required to correctly understand the source code. The CPG was found to cue the software developer to complex portions of the code that might require special attention; the CSD was found to aid in understanding specific details of the code. The results of these experiments along with the planned follow-on experiments promise to shed light on fundamental questions regarding the role software visualizations can play in program comprehensibility.