Boston Globe: Surveyed scientists admit misconduct

[Guest guest]

THIS STORY HAS BEEN FORMATTED FOR EASY PRINTING

Surveyed scientists admit misconduct

One-third cite research tactics

By Gareth Cook, Globe Staff  |  June 9, 2005

A third of American biomedical scientists have engaged in questionable

research practices, according to survey results released yesterday that raise

questions about the integrity of the nation's multibillion-dollar quest to

understand the human body and cure diseases.

The study, based on a survey of about 3,000 government-funded scientists, is

the first broad, quantitative examination of misconduct that asked researchers

to admit their own misdeeds. The scientists, who participated anonymously,

were asked whether they had done any of 33 actions in the three years before the

2002 survey. Asked about the most serious misconduct, 0.3 percent said they

had falsified data, and 1.4 percent said they had used another's ideas without

gaining permission or giving credit. In addition, 15.5 percent said they had

changed how they conducted an experiment or its results in response to pressure

from a funding source, raising the prospect that companies are influencing

scientific papers to support their commercial interests. The scientists also

admitted a range of other misdeeds, such as circumventing the rules on using

human subjects in experiments, and not properly disclosing ties with companies.

''We found a striking level and breadth of misbehavior, " said lead author

C. son, a researcher at HealthPartners Research Foundation in

Minneapolis. ''I think this really causes us to call into question the

assumption

that it is just a few bad apples. "

There is no way, said son, to gauge how much of the nation's research

was compromised by the misconduct. And several specialists on scientific

conduct said that it was difficult to know from the study how common scientific

misbehavior is because many of the questions were worded vaguely, and could

include behavior that is not objectionable. For example, a scientist might have

changed the design of an experiment after a legitimate suggestion from a

government funding source.

But the specialists welcomed the work, which was published in the journal

Nature, saying more research like it is needed at a time when science is

becoming

increasingly commercialized.

Trust and integrity lie at the heart of the scientific process, with

published experimental results making careers, determining whether scientists

win

research grants, and shaping spending priorities in the nearly $30 billion

budget

of the National Institutes of Health. At a time when scandals have shaken the

worlds of business, politics, and journalism, the authors of the new report

said that similar factors -- such as intense competition and human failings such

as greed and cynicism -- threaten the fundamental working of science. They

said the problem goes well beyond the egregious cases that the government is

authorized to investigate.

Editors of prominent medical journals have been increasingly vocal about

financial conflicts of interest that they say are hampering science, causing

researchers to hype positive results and downplay negative ones. Yet the topic

of

misconduct tends to make scientists uneasy, and that has led to a dearth of

research on the subject.

''I think this is a very important step, " said C. K. Gunsalus, a special

counsel at the University of Illinois at Urbana-Champaign, and one of the

nation's

leading specialists on research integrity. ''The discomfort means that we

don't like to talk about it, and that means we don't have good data. "

Surveying misconduct was controversial even before the current study was

done. In 2002, the US government's Office of Research Integrity proposed

conducting a survey of scientific misconduct, but several scientific groups,

including

the Association of American Medical Colleges, objected. They said that the

survey questions were vague and might be misused, and that the federal

government's role should be restricted to policing fabrication, falsification,

and

plagiarism.

The government study was eventually canceled, and that same year the editors

of Nature harshly criticized the scientific groups for their role in stopping

it, saying they gave ''a good impersonation of aged, out-of-touch special

interests with something to hide. "

The survey reported yesterday was done with government funding, including

money from the Office of Research Integrity, but it was conducted by an

independent scientific team. An official with the Association of American

Medical

Colleges, which represents some of the nation's leading biomedical research

institutions, said the group had no objection to the survey being done this way,

but

she declined to comment on the results of the study, saying she had not had

time to review it carefully.

son said his team designed the survey based on interviews with

scientists about the kinds of misbehavior they believe are most common. In these

interviews, he said, he was surprised at how candid scientists were in

describing a

wide range of problems. Some said they felt guilty crossing ethical lines,

but that they needed to in order to succeed. One scientist, he said, described

coming across a case where his own work had been systematically plagiarized,

but the scientist did not report it because the person who had done it was a

powerful figure in the field.

The team designated 10 of the behaviors as the most serious types of

misconduct, based on interviews with officials at universities who oversee

research

integrity. Thirty-three percent of scientists admitted to at least one of these

10 behaviors in the three years before the survey, according to the paper.

In the report, titled ''Scientists Behaving Badly, " the most common

misbehavior was making changes in response to pressure from a funder. There have

been

cases, now public, where drug firms have pressured scientists to rewrite or not

publish papers because they would harm the market for one of their products.

Two of the most common practices found in the survey are likely to raise red

flags because they hint at a breakdown of the basic checks and balances that

are supposed to correct the scientific record. Of the scientists surveyed, 12.5

percent admitted to ''overlooking others' use of flawed data or questionable

interpretation of data, " and 6 percent admitted to ''failing to present data

that contradict one's own previous research. " The paper also reported other

behaviors beyond what it called the ''top 10 " most serious offenses. Ten percent

admitted to ''inappropriately assigning authorship credit " and 15.3 percent

admitted to ''dropping observations or data points from analyses based on a gut

feeling that they were inaccurate. "

But because of the vagueness of many of the questions, it is impossible to

know how serious an infraction the scientists were admitting to, or even if it

was an infraction, said Dr. Drummond Rennie, a deputy editor of the Journal of

the American Medical Association who has been a longtime advocate for more

study of misconduct. Rennie said that he welcomed the work and hoped there would

now be more rigorous study of the issue.

Another problem, Rennie and others said, is that the survey relies on

scientists to report on themselves, and even with the promise of anonymity, the

results depend on the honesty of the people filling it out. Also, only about

half

of the scientists responded to the survey, which was mailed.

son said that he agreed there were flaws in the study, but that he

hoped it would inspire more discussion of the problem.

''I don't have all the answers, " son said. ''What I think I have here

is some evidence that suggests we need to begin a more broad-based

conversation. "

Gareth Cook can be reached at cook@...  

© Copyright 2005 The New York Times Company