Assessing research effectiveness: a comparison of alternative nonparametric models

In this paper, we examine three alternative a posteriori weighting schemes with variable, common and restricted weights in order to assess research productivity by means of two seemingly similar nonparametric models: the Benefit-of-the-doubt and the Kao and Hung (2003) model. Our empirical results, based on different types of faculty members’ publications, show that there is more variability in the estimated effectiveness scores among alternative weighting schemes within each model rather than between models for any particular weighting scheme. In addition, we also found that the effectiveness scores from the BoD model are greater than or equal to those from the K&H model for the variable- and the restricted-weights schemes while there is no clear pattern between the BoD and the K&H effectiveness scores from the common-weights scheme.