| The problem addressed in this study was the high failure rate of software development projects. There has been a trend in software development which has moved away from traditional approaches towards hybrid models tailored to the specific project and type of organization. A number of authors have argued for the use of hybrid models to help overcome weaknesses and limitations of traditional models, but the evidence base in support of this remains relatively weak. In this quantitative survey study, the extent to which project success varies as a function of IT managers' software development approach was examined. In addition, possible differences based on whether or not Six Sigma has been implemented in the organization, and possible relationships between project success and project size and managers' years of software development experience were explored. Data were gathered by the means of a survey questionnaire distributed in electronic format over the Internet to IT managers. The results indicated that (a) both agile and hybrid methods produced higher project success scores than other, non-waterfall methods; (b) project success was higher when Six Sigma was implemented than when Six Sigma was not implemented; (c) when team size was small, project success scores tended to be lower than when group size was large, with project success scores for projects having medium-sized teams not differing from those for small or large teams; (d) project success scores did not differ between low cost projects, moderate cost projects, and high cost projects; (e) project success scores for short projects, moderate projects, and long projects did not differ; and (f) more experienced managers led more successful projects. Based on these conclusions, it was recommended that agile and hybrid methods should be the preferred methods of software development for most projects and that more experienced IT managers should be sought. The researcher recommends organizations implement Six Sigma on information technology projects. In addition, it was recommended that this study should be replicated with a broader group of IT managers to examine the generalizability of the conclusions from this study to IT managers who are not members of the AITP, that future researchers should provide a more thorough assessment of the projects being rated by the participants, and that future researchers should explore team size under various software development scenarios. |
