Previous Chapter Chapter 28: Dominance Analysis Next Chapter
Chapter 28: Dominance Analysis
In a stepwise regression approach, the relative importance of a set of variables is dependent upon that variable's unique contribution to [Page 92]prediction after partialling out any variance shared with variables entered previously. When correlated predictors are compared with each other, all sensible measures of relative importance will be affected by the other predictors. Inspired in decision theory, dominance analysis is an analytic strategy that assesses the contribution each variable makes to the prediction of a dependent variable, considering both its unique contribution and its contribution when combined with other predictors (Budescu, 1993; Azen and Budescu, 2003; Budescu and Azen, 2004). It should be noted that dominance analysis described in this entry is not related to stochastic dominance analysis.
Typically, the relative importance ...