Model to Predict Schedule Variance in Software Application Development Projects

Metamorphic changes in today’s business are inevitable due to advancement in technology. For a business to either flourish or perish technology plays an important role. In Digital era Information Technology plays a pivotal role in complex business transformations. IT organizations have its own challenges of delivering right the first time, delivering on time and on budget, sustaining profit margin and customer satisfaction. A research, with at most focus on software development project schedule and on-time delivery was administered. The primary data was dissected and analyzed using suitable statistical techniques. Objectives: (1) To develop a predictive model which measures schedule variance on the basis of multiple independent variables (2) To describe the skills required for project managers to effectively handle on time delivery of project using factor analysis. (3) To measure the strength of association between the schedule variance and demographic variables of a project manager and project. Methods:  Data from 70 different projects and additional responses from the project managers managing those projects are collected using mail survey techniques. Findings: The output of multiple linear regression (MLR) technique helped to predict change in software application project schedule with a degree of accuracy of R-Value, the coefficient of correlation 90.1%. The research analysis indicates that the relationship between the dependent variable, which is project schedule variance and independent variables, of the developed model is good. The model will fit the real time project environment. Using multivariate factor analysis, sixteen key skills needed for a project manager for efficient on-time delivery has been identified. These are highly loaded into five different factors. Chi-square goodness of fit which is a non-parametric test helped to identify the strength of association between schedule variance and demographic variables.Conclusion: The analysis has resulted in a predictive model with eleven independent variables (excluding constant) impacting schedule variance. The five different competencies are the factor labels which are estimate and re-estimate, managing risks, people management, requirement management and time management and control over runs required for project managers, to have a control on the project schedule. It is also inferred that there is no significant association between the project manager’s demographic variables and ability to complete the project within the agreed time in statement of work (SOW)