Abstract
Given the increasing popularity of across-sample R2 comparisons in accounting research, this paper illustrates why the regression R2s are incomparable across sa