Risk factors in software development projects: Analysis of the Chilean software industry
Javier Pereira, Narciso Cerpa, Mario Aldea Rivas · 2008
Here, we present some results and future perspectives of an investigation looking for identification and description of risk factors in the Chilean software development industry. Data has been obtained by questionnaires asking for seven risk categories identified in the international literature. Two experimental questions encourage our research. Firstly, are these risk categories present in the Chilean case? If not, for those relevant and new categories found in the Chilean case, what are the relevant variables describing them? Secondly, is it possible to define the software project success in terms of risk factors and, consequently, find a synthetic cause-effect model, in the Chilean reality, where a software project manager could search for answer elements in front of a specific project? Actually, we have already questioned a number of software development enterprises, groups and professionals. Using data analysis we found some interesting results indicating characteristics of a successful project, as perceived by people: 1) improvement of professional expectations, 2) enhancement of satisfactory labor relationships and 3) detailed project scheduling. In the next step, we are currently searching for regression models which could explain the project success by variables as methodology, estimating, management commitment, sponsorship, and others. Further, a Bayesian belief network is being built to have a causal model of project success in terms of risk factors. The final stage of our investigation is that network implemented in a decision support system.