Use of Factor Analysis to Develop OOP Software Complexity Metrics
Al Lake, Curtis R. Cook · 1994
A large number of software metrics have been proposed for measuring the complexity Object-Oriented programs. These metrics were grouped into system, tree, and class metrics. Factor analysis was applied to each group which to obtain a small number (5 or less) of complexity domains by identifying collinear metrics. This reduction in complexity domains should faciliate the validation of the metrics and should provide a better understanding of OOP complexity factors. A side effect of this reduction is a plausible explanation of the controversy about the effectiveness of traditional metrics in measuring OOP complexity. Keywords: Software complexity metrics, object-oriented programming, factor analysis