Research quality metrics as a complement to bibliometrics: an opinion paper

Substantial proportions of research investment may be wasted due to flawed study design, biased reporting, and publication bias, raising concerns about how research performance is evaluated. Despite this, current assessment systems rely heavily on bibliometric indicators such as publication counts, co-authorship, and citation metrics, which do not reliably reflect research quality or scientific rigor. In this opinion paper, we argue that such indicators may reward practices that undermine the accumulation of reliable knowledge, including p-hacking, selective reporting, and underpowered study designs, while undervaluing essential contributions such as replication research. We propose that bibliometric measures should be complemented by transparent and auditable research quality metrics, including indicators of reproducibility, statistical power, risk of bias, and adherence to reporting standards. We further suggest that recent advances in large language models may facilitate the scalable implementation of such metrics. Adopting quality-focused indicators could improve research evaluation and help reduce avoidable research waste.