Distal Data and Indicator Interactions

Abstract

Although representative ranks are indirect indicators with intrinsic indefiniteness, they are particularly advantageous for determining whether there are cases having pronounced sensitivity to a change in one indicator. It may be that one indicator has especially good status, whereby deterioration in the indicator would degrade the overall status of the case substantially. Conversely, it may be that an indicator has especially poor status, whereby improvement in the indicator would upgrade the overall status of the case substantially. In the latter situation, some remediation effort focused on the weak indicator could pay large dividends (Newlin and Bhat 2007). Statistically, a sensitive situation involves a rank outlier producing an extended tail on the distribution of ranks for the case, which can occur on one or both ends.