Success rates by software development methodology in information technology project management: A quantitative analysis

Despite over half a century of Project Management research, project success rates are still too low. Organizations spend a tremendous amount of valuable resources on Information Technology projects and seek to maximize the utility gained from their efforts. The author investigated the impact of software development methodology choice on ten measures of project success in an effort to further understanding. Responses from a web-based survey were gathered for 94 United States software development Project Managers. The software development methodology used on the software project, the independent variable, was classified as either agile, structured, or with some degree of hybridization. The ten critical success factors, the dependent variables, were derived from existing research. The author utilized Spearman's rank correlation coefficient and post-hoc power analysis to examine the impact of software development methodology choice on ten measures of project success. The results of the study indicate that for supplier satisfaction agile projects exhibit slightly higher success rates than structured projects at the .05 significance level, albeit at lower than desired power. For the other nine measures of success, software development methodology choice does not appear to impact the success rates. This suggests that practitioners should make software development methodology choices without concern about the impact on the ten measures of success. It is noted that a statistically significant relationship was found at the .05 level between software development methodology choice and both years of experience with the software development methodology and Project Management Institute membership; indicating the possibility of confounding variables.