2023-09-08 04:02:23
It’s a simple two-page opinion in the British Journal of Medicine, August 17, 2023. This is the summary of a May 2023 presentation by Ivan Oransky at Stanford University. How much is a quote worth? Three, six or 100,000 US$? All three numbers are wrong, but it all depends on the appraiser.
Biased publications and citations to assess science!
Depending on the indicators, publications and citations represent 20 to 60% in the calculation. These indicators are used by agencies, governments to allocate resources, by juries for the promotion of careers! What an idea, because these indicators can be easily manipulated…. All tricks to manipulate are known, including paper mills.
Not all actors want retractions that harm calculations, careers… and newspapers don’t want to end up in court. So retractions are slow, opaque and inconsistent when they occur. An English report by a House of Commons science committee recommended that retractions are made within two months of publication !!!!
Publishers, institutions and researchers are caught in an infernal loop: “Following the money”. Publishers respond to requests by creating an astronomical number of special issues, the paper mills target these special issues.
No solution without changing incentives
Ivan’s words are stern and realistic: there will be no solutions without changing the incentives. It’s simple, reduce the importance of citations in these indicators… It’s easy to say… Why not reward behaviors like data sharing and literature corrections. As proposed by the Research Excellence Framework in England, quotes should represent 10% in the weight of an indicator!!!
In short, it would be better to reward the virtuous behavior of researchers rather than the manipulators of impact factors… although DORA, Leiden and other recommendations have been signed, few of the signatories are implementing them. AMAZING, everyone agrees but who will take the first step?
‘Science must be made more rigorous’ is R Lévy’s message in Nature reviews physics.
This short article contains 34 well-selected quotes.
1694146290
#science #believed #result #bibliometric #indicators #race #universities