A statistical framework for analyzing the duration of software projects
Empirical Software Engineering
| Hi-index | 0.00 |
This paper proposes a new approach that can discriminate risky software development projects from smoothlyor satisfactorily going projects and give explanation forthe risk. We have already developed a logistic regressionmodel which predicts whether a project becomes risky ornot. However, the model returned the decision with the calculated probability only. Additionally, a formula was constructed based on the risk questionnaire which includes 23questions. We therefore try to improve the previous methodwith respect to accountability and feasibility.In new approach, we firstly construct a new riskquestionnaire including only 9 questions (or risk factors)Q1, Q2, \cdot\cdot\cdot, Q9, each of which concerns with the projectmanagement. We then apply multiple regression analysis toactual project data PJ1, PJ2, \cdot\cdot\cdot, PJ32, and clarify a set offactors which contributes essentially to estimate the relativecost error and the relative duration error, respectively. Wethen apply the constructed formulas to another project dataPJ33, PJ34,\cdot\cdot\cdot, PJ40. The analysis results show that boththe cost and duration of risky projects are estimated fairlywell by the formulas. We can thus confirm that our new approach is applicable to software development projects in order to discriminate risky projects from appropriate projectsand give reasonable explanations for the risk.