Is Quality a Casualty in the Race to Publish?

WASHINGTON—Last spring’s newspaper stories that described how IBM researchers had boosted the critical current density of a superconductive thin-film crystal by a factor of 100 were also bringing news of the discovery to most scientists. Not until six weeks later were the details published in Physical Review Letters. Increasingly, scientists in fast-paced fields are announcing breakthroughs at meetings or press conferences. Long before results appear in scientific journals, they circulate in the popular press, by word of mouth, as preprints or in various newsletters, databases and other sources set up to convey information quickly. (See THE SCIENTIST, July 13, 1987, p. 1.) Some scientists worry about this departure from the established practice of announcing results only as they are published in a journal. “I find the trend a bit disturbing,” said M. Brian Maple, a physicist doing superconductivity research at the University of California at San Diego. When details of experimental procedures are withheld in such announcements, the results often are “of very little scientific value,” he said. “By short-circuiting the traditional ways of announcing results, it’s very difficult to assess what’s right and what’s wrong.” New England Journal of Medicine Editor Arnold Relman terms it “irresponsible” for scientists to announce their results without also providing the supporting data on which they are based. “In science, there’s no news without the data,” he said. “Until then it’s just gossip.” Yet others defend such tactics as vital to rapid advances in crucial areas of science. “In a field like AIDS, there is a need to get information out quickly,” said George Galasso, an associate director at the National Institutes of Health. “Waiting for it to come out in a journal isn’t good enough; even if.you’re fast, it still takes six months.” Many journals have relaxed their policies on prior disclosure and now permit scientists to announce their findings well before their articles are published. At the same time, some journals have also taken steps to get the information out more quickly. The Astrophysical Journal, for example, is expediting publication of articles on Supernova 1987A, according to Managing Editor Helmut Abt of Kitt Peak National Observatory. “Our papers normally go through the mail 14 times,” Abt said. The journal’s lengthy review process means that at least eight months normally elapse between an article’s submission and its publication. That interval has been cut to three months by having supernova articles typeset in the United States instead of Britain and by asking for quick turnarounds from reviewers, Abs said, adding that “we insist on the same standards” as for papers that take the usual amount of time. At N&JM, Relman said, “we’ve told authors well make every effort to accelerate the review process” for papers on AIDS and other particularly timely topics. In some cases, the normal six-month lag between submission and publication has been trimmed to less than three months. But balancing competing needs for quick publication and “our obligation to review manuscripts as carefully as possible” is a tough job, he added. Eat, Sleep and Read To cope with the flood of articles on superconductivity research submitted for publication in Physical Review Letters and Physical Review B, the editors formed an anonymous review panel to screen all papers on the topic. This step was intended to speed articles into print and ensure some consistency in publication decisions. The panel was recently expanded from five to nine scientists,’ said Assistant Editor Reid Terwilliger. Each paper is submitted to four of them for review. “All they’re doing, I imagine, is eating, sleeping and handling those papers,” said Terwilliger, who noted that the two journals have received nearly 400 papers on the topic this year. “The turnaround time (for the outside review] is quite good—one or two weeks, which is quite unusual.” Many scientists believe that the overall quality of published papers has declined as researchers rush to submit articles and editors rush them into print Yet most describe that tradeoff as inevitable. One member of the special review panel, who asked to remain anonymous, acknowledged that he is “less inclined to quibble” in screening superconductivity articles. He agrees that an erosion in quality is sometimes the price for timeliness. “There’s a tendency to give an author the benefit of the doubt because it’s such a new field,” he observed. “If-someone baa a new resuit and it’s suspicious, you don’t want a reviewer to (reject the papen, because it might be right.” “Ordinarily, scientists like to take their data and chew on it a while,” said Donald Murphy, head of solid-state chemistry research at AT&T Bell Labs. “It’s unheard of to start a project, do the research and write up the data within a couple of weeks.” When that process is compressed, he said, “the data are going to be more preliminary than usual. But in the beginning stages of a field, that’s acceptable.” Despite the pressure it generates, the new approach can yield a clearer picture of the overall direction of the field. Several researchers commented on the usefulness of the special April and May issues of the Japanese Journal of Applied Physics, which together contained about 200 non-reviewed papers reporting on superconductivity research in Japanese labs. The collection of articles “is quite valuable,” said Robert Dynes of Bell Labs,. who called it “almost a barometer of what people are thinking.” Other scientists insist that archival journals must maintain their review standards, even in hot research fields. “Some people say, ‘Just publish everything and let the reader sort it out.’ I think that would be a mistake,” said physicist Alex Zettl of the University of California at Berkeley. “I would like some filtration to be there.” Terwilliger said Physical Review editors rejected the idea of publish.. ing a special issue of papers that had not been reviewed. Instead, they are republishing in a single volume the 110 or so papers that appeared in both journals in the first six months of the year. Zettl said that, although he disagrees with some decisions by the Physical Review reviewers, the special panel is a necessity. “I don’t see any alternative,” he said. “I thought the older way of refereeing was more thorough, but you just can’t do that here.” Greene is on the staff of THE SCIENTIST.

Is Quality a Casualty in the Race to Publish?

Interested in reading more?

Become a Member of