Information cascades in scientific practice

Here’s an in­ter­est­ing re­cent pa­per in the Bri­tish Med­i­cal Jour­nal: “How cita­tion dis­tor­tions cre­ate un­founded au­thor­ity: anal­y­sis of a cita­tion net­work”. (I don’t know if this is freely ac­cessible, but the ab­stract should be.)

From the pa­per:

Ob­jec­tive To un­der­stand be­lief in a spe­cific sci­en­tific claim by study­ing the pat­tern of cita­tions among pa­pers stat­ing it.”

Con­clu­sion Ci­ta­tion is both an im­par­tial schol­arly method and a pow­er­ful form of so­cial com­mu­ni­ca­tion. Through dis­tor­tions in its so­cial use that in­clude bias, am­plifi­ca­tion, and in­ven­tion, cita­tion can be used to gen­er­ate in­for­ma­tion cas­cades re­sult­ing in un­founded au­thor­ity of claims. Con­struc­tion and anal­y­sis of a claim spe­cific cita­tion net­work may clar­ify the na­ture of a pub­lished be­lief sys­tem and ex­pose dis­torted meth­ods of so­cial cita­tion.”

It also in­cludes a list of spe­cific ways in which cita­tions were found to am­plify or in­vent ev­i­dence.