Review of Dominance Analysis: An Approach for Determining the Relative Importance of Predictors

El-Houssainy Abdel Bar Rady; Ahmed Amin EL-Sheikh; Mohamed Rashed Ezzeldin

Abstract

A lot of methods had been developed to explain the unclear concept of relative importance for independent variables. One of the most important methods for determining the relative importance of predictors is dominance analysis, which is a technique that determines variable importance, based on comparisons of unique variance contributions of all pairs of variables involving all possible subsets of predictors. The aim of this paper is to take a closer look at dominance analysis including assumptions, methodology, techniques and the types of data used and provide a review of recent developments of the dominance analysis approach and its applications. The strengths and weaknesses of the dominance analysis have been clarified. We focus on some areas where there has been an intense methodological development during the past years, linking dominance analysis to a theoretical framework that demonstrates the complementary roles they play in regression findings. In relation to the goal and scope definition and based on current research we split the paper into sections including dominance analysis and the main areas and approaches used. We also discuss recent developments in relation to some of the strengths and weaknesses of dominance analysis. Finally, we end with several recommendations and future research suggestions

Abstract

Keywords