Psychologist Brian Nosek believes that reproducibility is a core principle of science. To promote the idea, he co-founded a nonprofit organization in 2013 that allows scientists to publish a description of their experiments before they conduct them. This week Nosek’s Center for Open Science (COS) went a step further, offering $1000 to every scientist who preregisters their protocol with COS.
The payment is meant to be a carrot leading to greater transparency and accountability in research, says Nosek, a professor at the University of Virginia in Charlottesville. “Preregistration increases the credibility of hypothesis testing by confirming in advance what will be analyzed and reported,” says the center’s website in describing the rationale behind the prize challenge. Advocates of preregistration say it could also reduce the number of “file-drawer studies,” in which scientists decide not to publish anything because of negative results.
It’s a limited offer. Only 1000 scientists will receive the money, which will be awarded once they have met all the requirements. The research must appear in a journal that has agreed to practice many of the open-science principles that the center espouses. And scientists don’t receive a cent until after publication. There are 460 approved journals across several disciplines, and the center is looking for greater participation from those in the social, behavioral, and economic sciences. Most researchers running clinical trials are already required by law to preregister on a government website.
The new prize isn't intended to support any research, says David Mellor, a project manager at the center in Charlottesville. Rather, he says, it's meant to be a marketing tool for COS in convincing investigators to try something new. The prize money comes from a $1.2 million gift from the Laura and John Arnold Foundation.
Preregistration may seem like a time sink, Nosek acknowledges. But he insists it saves resources in the long run. "My lab now preregisters everything we do,” he says. The process has led to better-designed studies, he adds. “It's helped to identify problems that we would have only discovered after collecting data and wasting time,” Nosek says. “Ultimately, preregistration makes us more efficient."
Nosek also hopes to put the registration process itself under a microscope. "Can we find evidence of whether [preregistration] is yielding an increase in the credibility of the research?" he asks. "That is a research question."