Visualization Techniques for the Integration of Rank Data.

In consumer preference studies, in Web-based meta-search or in metaanalysis of microarray experiments, we are confronted with ranked lists representing the same set of distinct objects. All these applications have in common that one is interested in the top-ranked objects with considerable overlap in their rankings across the lists. This requires the estimation of the truncation point beyond which the ordering of the objects is dominated by noise. The point of degeneration into noise can be obtained from an inference procedure due to Hall and Schimek (2008) which even works for large or huge data sets. Before its execution, it is essential to specify the distance parameter delta. In this paper, graphical approaches for the delta-choice, as well as for the integration of the top-ranked objects, are introduced for the rst time. Finally, the new graphical tools are applied to the integration of microarray data from several experiments.