Investigating software design measures as indicators of understandability

Early indications of the impact of different design / coding considerations on program understandability (or the difficulty or ease of it to understand) can help in reducing maintenance costs and consequently in improving quality. Our literature survey indicates insufficient attention of researchers to address this issue, using static analyses techniques. Although some work has been conducted in the past to validate different design measures and observe their impact on quality, the main focus' of these studies was directed towards quality attributes such as, maintainability and fault proneness. This paper presents an empirical study that aims at exploring the usefulness of thirteen software measures to obtain indications of their effect on understandability early in the software lifecycle. The study limits itself to only those aspects of understandability that can be captured by the above notion. Promising results that can be of great value to quality controllers to improve understandability of designs were obtained through the study.