CIS faculty Dr. Paul Smaldino recently published a paper on the “natural selection of bad science.” This work demonstrates that incentives for productivity can drive questionable scientific practices, and may then hinder scientific progress. In the paper, he and his collaborator analyze existing data, in addition to developing an exciting computer model showcasing this kind of dynamic in the scientific community. The computer model is called an “agent-based simulation,” and forms the bedrock of Dr. Smaldino’s research. In the model, simulated “laboratories” compete for influence in a world in which published results are the coin of the realm. The resulting incentives to produce publishable findings can lead to a propagation of questionable research practices, because simulated “students” of the computer model then develop their own research laboratories, which continue these practices, akin to an evolutionary dynamic. Click here to see the paper.
From the abstract, they note…
As in the real world, successful labs produce more ‘progeny,’ such that their methods are more often copied and their students are more likely to start labs of their own. Selection for high output leads to poorer methods and increasingly high false discovery rates. We additionally show that replication slows but does not stop the process of methodological deterioration. Improving the quality of research requires change at the institutional level.
The work gained significant attention in the press, for example:
The Atlantic: The Inevitable Evolution of Bad Science
Times Higher Education: ‘Bad science’ spreads through natural selection, says study
The Economist: Incentive Malus