Humboldt-Universität zu Berlin - High Dimensional Nonstationary Time Series

SFB649DP2016 020

Academic Ranking Scales in Economics: Prediction and Imputation

Alona Zharova
Andrija Mihoci
Wolfgang Karl Härdle

Abstract:
We address the problem that often hampers decision making in academic institutions – incomplete research profiles. We suggest a framework for collating ranking data of scientists for comparison purposes. As the result of an analysis of the interconnectedness between HB sub-rankings through quantile regression, we propose a HB common score for scholars within the HB community. The cross-ranking dependence analysis of Handelsblatt, Research Papers in Economics and Google Scholar ranking schemes shows that researcher age and field of specialization – mapped onto the JEL classification codes – have a substantial impact on the resulting scores.

Keywords:
ranking, prediction, quantile regression, Handelsblatt, RePEc, Google Scholar

JEL Classification:
C81, C53, C21, M10