| Summary: | In this paper we describe an experimental study conducted at McGill University in which we assessed the performance of different types of team organizations in terms of their effect on key process variables such as rework, defect density, review quality, cost of quality, productivity, etc. We also analyzed the ability to rank product quality from process data, prior to the completion of the projects. The qualitative analysis showed the benefits and disadvantages in working on a hierarchical team or a collaborative team organization. === All those factors are also relevant when developers need to choose the best product, based on process data collected since earlier phases (prior to product delivery). Because of that, we also investigated and rank the processes in order to predict based on process data which product is most likely to be chosen by the client In order to rank key process variables, we (1) analyzed the quality features individually, (2) analyzed a selected set of variables and (3) did combinatorial (mathematical) analysis to see if any mechanical results concur with the qualitative analysis accomplished. === All the templates, and standards used in this empirical study followed the PSP[28] or Personal Software Process. (Abstract shortened by UMI.)
|
|---|
