The Leiden Manifesto for research metrics

Abstract

Data are increasingly used to govern\rscience. Research evaluations that\rwere once bespoke and performed\rby peers are now routine and reliant on\rmetrics1\r. The problem is that evaluation is\rnow led by the data rather than by judgement.\rMetrics have proliferated: usually\rwell intentioned, not always well informed,\roften ill applied. We risk damaging the system \rwith the very tools designed to improve\rit, as evaluation is increasingly implemented\rby organizations without knowledge of, or