Data reduction techniques and hypothesis testing for analysis of benchmarking data

This paper proposes a data reduction and hypothesis testing methodology tha t can be used to perform hypothesis testing with data commonly collected in benchmarking studies. A reduced-form performance vector and reduced-form s et of decision variables are constructed using the multivariate data reduct ion techniques of principal component analysis and exploratory factor analy sis. Reductions in dependent and exogenous variables increase the available degrees of freedom, thereby facilitating the use of standard regression te chniques. We demonstrate the methodology with data from a semiconductor pro duction benchmarking study.