The Φ index and world university rankings

World university rankings pose several methodological questions, from using largely unnormalized counts of publications and researchers to relying on subjective reputation surveys and allocating arbitrary weights to various categories. Some of these ranking schemes use citation averages normalized for differences between fields, publication years, and document types. However, failing to standardize citation averages for size can introduce scale bias, an effect that has been shown [1] to skew Journal Impact Factor rankings. Previously, we developed the Φ index methodology to design standardized citation averages that produce unbiased journal rankings [2]. Here, we extend the Φ index to the rankings of research institutions and compare our findings to some well-known university rankings.