Ranking Projects Using Multivariate Statistics

Organizational evaluation is a complex activity which demands a lot of effort. In order to accomplish this activity it is necessary to work with process and project indicators. Looking for project indicators it is necessary to identify good and bad projects to treat problems and to consider improvements. This paper presents a new methodology to rank projects using principal component analysis, a simple and efficient multivariate statistical method. This method aggregates project measurement indicators in just one score during the execution of the Organizational Evaluation macroactivity defined in the Organizational Software Measurement Process (OSMP).