A Study to Define Current Practices, Attitudes and Values in the Software Development Community
Date of Award
Doctor of Philosophy (PhD)
Graduate School of Computer and Information Sciences
Steven R. Terrell
The objective of this study was to secure information from a wide distribution of software development practitioners about their development processes, their views about related processes and methods and their perceived and actual software process maturity. Research questions answered in this study were: (a) what are the current software development practices in industry? (b) Are there significant differences between the practitioner's perceived and actual development process maturity; (c) Is there correlation between professional and/or organizational demographics and perceived and actual development processes?; and (d) What do practitioners believe are significant success and failure factors as they relate to job performance? These questions and others were answered by this study.
A self-administered, mailed survey, the Software Development Process Survey, was developed for this study as the data gathering instrument. The survey experienced a return rate of 49.2% and 95.6% of the questions were answered by the 203 respondents. Due to the size, distribution and response to validity-check questions, the results of the survey were considered fairly representative of the targeted population.
Two metrics were also developed. The Perceived Process Maturity Level was the result of the subject's stated perceived, estimated, process maturity level. The Calculated Process Maturity Level resulted from calculating responses to 25 process-related questions. The two metrics were then compared to each other and to the actual distribution of the Capability Maturity Model developed by the Software Engineering Institute.
The data show that the subjects have defined a 0.3 level difference between their perceived and actual process maturity level. Additionally, the subjects generally felt that their present position, company and product quality were not as good as experienced in previous positions. The study also revealed that metrics and metric processes had low priority, neither helped or hindered the development process, could be minimized in today's competitive environment, and increased development costs. A fully annotated version of the survey is included.
William M. Hartman. 1998. A Study to Define Current Practices, Attitudes and Values in the Software Development Community. Doctoral dissertation. Nova Southeastern University. Retrieved from NSUWorks, Graduate School of Computer and Information Sciences. (569)