Adam Marcus, the managing editor of Gastroenterology & Endoscopy News, and Ivan Oransky, distinguished writer in residence at New York University’s Arthur Carter Journalism Institute and vice president for editorial at Medscape, are co-founders of Retraction Watch.
Government scientists may have felt a chill after The Post reported in April that the Agriculture Department has been requiring its researchers to label their published studies as “preliminary.”
Given the Trump administration’s disdain for science that doesn’t fit its narratives or self-interest, it’s hard to view the USDA’s policy as anything other than a ham-handed political intrusion into scientific freedom. And yet the Trumpian edict might have gotten something right about science, even if for the wrong reasons.
Recall that Trump once nominated Sam Clovis, who isn’t a scientist, to be the USDA’s top science official. (Clovis withdrew after he was linked to the special counsel’s Russia investigation). Or that Scott Pruitt, Trump’s first head of the Environmental Protection Agency, wanted to prevent the publication of any studies for which the data weren’t publicly available.
Pruitt professed a desire for greater transparency in EPA-funded research, but critics saw the proposal as a Trojan horse that would give political appointees veto power over science they didn’t like — a case made stronger by the fact that much of environmental science involves studies of people who agree to participate in exchange for strict protection of their privacy.
In other words, no one should be shocked if Trump’s USDA uses the “preliminary” labeling requirement as a weapon to undermine research into climate change or other politically charged areas.
Not surprisingly, some researchers familiar with the policy expressed concern about the rule. Christine McEntee, director of the American Geophysical Union, told The Post that her group feared the disclaimer might interfere “with the dissemination of scientific findings that are important for the public.” Ed Gregorich, editor of the Journal of Environmental Quality, was blunter. Science that appears in the pages of his publication is “finalized,” Gregorich said. “There’s nothing preliminary about it.”
Yet even if the motivation for the “preliminary” disclaimer is wrongheaded, the message isn’t off-base. Just because a scientific finding appears in published form does not make it finalized — or even necessarily true. Wine is good for you until it’s bad for you. Ditto coffee. Violent video games lead to violent humans — until they don’t. A small but growing mountain of evidence shows that in certain fields, much of what is published in the scientific literature cannot be fully, or sometimes even partially, reproduced.
The problem of irreproducible results is bad enough in some areas, such as behavioral psychology and cancer research, to have been called a “crisis” by many experts in scientific rigor. That means, as scientists remind each other when they’re speaking privately, that all science is for practical purposes preliminary. And pretending otherwise helps neither researchers nor the public.
The world has seen that with vaccine research, as science writer Melinda Wenner Moyer said last year in a New York Times op-ed: “Scientists are so terrified of the public’s vaccine hesitancy that they are censoring themselves, playing down undesirable findings and perhaps even avoiding undertaking studies that could show unwanted effects.”
Yet research results need to be aired regardless of whether their meaning can be exploited for unscientific ends. None of this means that vaccines cause autism. They don’t. Or that climate change isn’t real. It is. Rather, we should be comfortable with the idea that a single study raising questions doesn’t upend an entire theory.
Scientists and policymakers need to take a long-term view, even if politicians on the fossil fuel dole are eager to embrace every opportunity to cast doubt on global warming. The strength of the scientific method — when it’s practiced as it should be — is in a transparent examination of data following the best experimentation possible.
Significant efforts are afoot to improve reproducibility, and there will continue to be some growing pains. But the threat of bad-faith attempts to exploit science reform shouldn’t be allowed to derail these attempts. To quote Marcia McNutt, president of the National Academy of Sciences: “Indeed there is always the possibility that those who would harm science would punish us for ferreting out our own weaknesses and correcting them. If so, shame on them.”