Individual research performance: a proposal for comparing apples to oranges

TL;DR

This paper proposes a method to compare individual research performance across different scientific fields by testing various scaling factors on productivity distributions, using Italian university data.

Contribution

It introduces a novel scaling approach for cross-field research performance comparison, addressing a key challenge in research evaluation.

Findings

The average productivity of researchers with non-zero output is the most effective scaling factor.

The method was tested on data from 174 scientific fields in Italy.

Results facilitate fairer comparisons of individual research performance across disciplines.

Abstract

The evaluation of performance at the individual level is of fundamental importance in informing management decisions. The literature provides various indicators and types of measures, however a problem that is still unresolved and little addressed is how to compare the performance of researchers working in different fields (apples to oranges). In this work we propose a solution, testing various scaling factors for the distributions of research productivity in 174 scientific fields. The analysis is based on the observation of scientific production by all Italian university researchers active in the hard sciences over the period 2004-2008, as indexed by the Web of Science. The most effective scaling factor is the average of the productivity distribution of researchers with productivity above zero.