Anderson, Christopher. "How Not To Publicize a Misconduct Finding," Science 263 (25 March
1994), p. 1679.
Here is an interpretation of the Bernard Fisher, Pitt, National Surgical Adjuvant Breast and
Bowel Project (NSABP) mess. According to Anderson: ORI is adopting new procedures to
ensure that misconduct findings are reported to the scientific community. The "problem" is here
being defined as ORI's: they did not get the word out. The publication of their findings of
misconduct in the Federal Register is not enough.
ORI is adopting procedures to insure that this will not happen again.
At the time that ORI issued its report in 1993, ORI had reason to believe that a correction from
Fisher and Pitt was in the works. It was to be sent to the NEJM which had published the original
report on mastectomy and lumpectomy in 1985. The problem was, apparently, that Dr. Fisher
wanted to do more than just a reanalysis without the Poisson-supplied data, he wanted to talk
about the statistical issues involved! That discussion of statistical issues is, purportedly, the
reason that Fisher did not publish his emendations.
ORI, it is reported here, did not notify NEJM of the misconduct it had uncovered and the editors
learned only from Crewdson's work that Poisson had winged it. ORI responds that it believed
Fisher would publish the story. He never did.
The National Cancer Institute is reported working with its legal staff to insure that corrections be