Scientific Fraud

See allHide authors and affiliations

Science  22 Dec 2006:
Vol. 314, Issue 5807, pp. 1853
DOI: 10.1126/science.314.5807.1853

One year ago, as Science was assembling its 2005 Breakthrough of the Year issue, the need for a last-minute change became uncomfortably clear. A shadow was creeping across one of this journal's landmark papers, in which a team of South Korean and American researchers, led by Woo Suk Hwang at Seoul National University, claimed to have created the first-ever human embryonic stem cell lines that matched the DNA of patients. After anonymous allegations of irregularities in that paper appeared on a Korean Web site, South Korean authorities launched an investigation. As the story unfolded, Science's news editors hastily pulled an item about the Hwang achievements from the issue's roster of runners-up.

See Web links on scientific fraud

Today, the fallout from the Hwang case is plain. Multiple inquiries discredited two papers Hwang published in Science in 2004 and 2005, which claimed some of the greatest accomplishments to date with human embryonic stem cells. The papers were retracted. But the scientific fraud, one of the most audacious ever committed, shattered the trust of many researchers and members of the public in scientific journals' ability to catch instances of deliberate deception.


The unraveling of Hwang's stem-cell papers was the first and worst of the year's research scandals.


As it turned out, the Hwang debacle marked the beginning of a bad year for honest science. Incidents of publication fraud, if not on the rise, are garnering more attention, and the review process is under scrutiny. In June, European investigators reported that the bulk of papers by Jon Sudbø, formerly a cancer researcher at the Norwegian Radium Hospital in Oslo, contained bogus data. Those included two articles in The New England Journal of Medicine that described a new way of identifying people at high risk of oral cancer, a strategy that many clinicians were keen to apply to patients.

Eric Poehlman, formerly a menopause and obesity researcher at the University of Vermont in Burlington, garnered perhaps the most dubious distinction of all: He became the first researcher in the United States to go to jail for scientific misconduct unrelated to patient deaths.

The Hwang case, however, was unique for its combustible mix of startling achievements in a high-profile field and publication in a high-visibility journal. Manipulated images, purportedly of distinct stem cells matched to patients but in fact showing cells drawn from fertilized embryos, handily fooled outside reviewers and Science's own editors. “The reporting of scientific results is based on trust,” wrote Editor-in-Chief Donald Kennedy in a January 2006 editorial explaining why journals are not designed to catch fraud. It's a comment echoed often by journal editors facing the nightmare of faked data in their own pages.

But the shock of the Hwang deception, along with other recent fraud cases, is jolting journals into a new reality. Five scientists and a top editor of Nature examined Science's handling of the Hwang papers, at the journal's request. Their report, published on Science's Web site earlier this month (, concluded that operating in an atmosphere of trust is no longer sufficient. “Science must institutionalize a healthy level of concern in dealing with papers,” the group wrote. It recommended “substantially stricter” requirements for reporting primary data and a risk assessment for accepted papers. Science and some other journals are also beginning to scrutinize images in certain papers, in an effort to catch any that have been manipulated.

Stem cell researchers, meanwhile, endured deep disappointment as a remarkable scientific advance evaporated before their eyes. Cloning early-stage human embryos, and crafting customized stem cell lines, is not the cakewalk some scientists hoped Hwang's papers had shown it to be. Stem cell researchers are backpedaling to more modest goals, just as Science and other journals consider how to prevent a breakdown of this magnitude from striking again.

See Web links on scientific fraud

View Abstract

Stay Connected to Science

Navigate This Article