Peer Review violation

Published by the National Science Foundation, Office of Inspector General on 1997-06-20.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                           CLOSEOUT FOR M96090029

       The complainant' received from a scientific journal for publication review a
manuscript2 (the manuscript) that reported work that was similar to preliminary results
discussed in an NSF proposal3 (the proposal). The complainant was concerned about
the possibility that the ~ u b j e c t the
                                       , ~ manuscript's first author, may have reviewed the
NSF proposal and then decided to work o n the research problem.
       Both the manuscript and the proposal describe a series of experiments using
similar techniques; both documents drew similar conclusions from the data discussed.
We determined that NSF had mailed the subject a copy of the proposal for peer
review, but had not received a proposal review from the subject.

       At our request, a knowledgeable, independent scientist assessed the degree of
similarity between the ideas presented in the manuscript and the proposal.' That
scientist said, "[slinck both [the complainant] and [the subject] have developed [these]
methods and have worked in the area for years, it is not surprising that they would use
the same m e t h o d ~ l o g ~ .The
                                " ~ scientist noted that although this type of experiment
was not uncommon and the subject had previously published these specific types of
measurements, the subject and the complainant had worked with the technique in
different subfields of their discipline.' The scientist also noted that, in a discussion of
preliminary results in the proposal, the complainant had "anticipatedn the manuscript's
conclusion that the "previously reportedn value of a parameter "is in error."' "While
the experimental results could have made [the anomalous value of the parameter] an
obvious topic to discuss, it is also not difficult to speculate that the discussion might
have been more limited had not interest in this point been invited by" the discussion
in the complainant's p r o p ~ s a l . ~

*   Id.
                                      Page 1 of 3                           M96-29
                            CLOSEOUT FOR A496090029

       We informed the subject that we had received allegations of misconduct in
science against the subject, specifically, allegations of intellectual theft and violation of
confidential peer review. We asked the subject to explain and date the origin of the
subject's interest in the specific research problem and to explain the manuscript's focus
on the anomalous behavior of the parameter. In his response to us, the subject (1)
asserted that he had never read the proposal before receiving it from OIG in
connection with this inquiry, (2) provided documents purportedly establishing his
interest in the specific research problem' prior to the date NSF mailed the subject the
proposal for review, and (3) explained the manuscript's discussion of the parameter as
"address[ing] a series of obvious and apparent scientific issues that fall right out of the

        The subject's assertion that the subject had not read the proposal before
receiving it from OIG is credible. The NSF jacket for the proposal does not contain a
review from the subject. The subject told us that the subject had "no recollection of
reading" the proposal prior to receiving it from us. The subject noted that because the
proposal discussed one of the subject's published papers, the subject "would have
remembered." The subject also said that the subject's own records contained no
evidence of a review for the proposal during the relevant year and provided us with a
list, compiled from the subject's own records, of four other NSF proposals the subject
had reviewed during that year. The NSF reviewer database identifies six proposals
from the relevant year that were sent to the subject for review, in addition to the
complainant's proposal: according to the database, NSF received reviews from the
subject for four of these proposals-the same four proposals identified by the subject.

        Enclosed with the subject's response was a copy of a paper, authored by one of
the subject's graduate students and the subject, that was published about 3 years
earlier than the proposal. The subject told us that the subject's investigation of the
specific research problem using the methodology described in that paper began well
before the date that NSF mailed the subject a copy of the proposal for review. The
subject provided us with copies of pages from laboratory notebooks recording two
experiments performed by a n undergraduate, a co-author o n the subject's manuscript,
4 months and 2 months prior to the date that NSF mailed the subject a copy of the
proposal for review. We have n o reason to doubt the validity of these documents,
which demonstrate the subject's active interest in the specific research ~ r o b l e mprior to
the date that NSF mailed the subject a copy of the proposal for review.

       Finally, the subject adequately supported the contention, which had been
recognized as a possibility by the knowledgeable, independent scientist, that the
discussion of the parameter in the manuscript addresses an obvious scientific issue.
The subject stated that the parameter was needed by the subject in "standard

                                        Page 2 of 3
                          CLOSEOUT FOR M96090029

'training/testing'" and that any observed anomalous behavior should be addressed in a
scientific analysis. The subject explained:

      [Wlhat we did was to take literature data and plot it according to a
      method in the paper we cite. When we looked at the plot, we felt the
      literature data, fit by a smooth curve showed the value [of the parameter]
      was off. . . .

      . . . Because we are 'rejecting' just one of .the . . . data points from the
      literature, we also go through an error analysis to show what impact that
      rejection has on the ultimate value we derive.

       We conclude that there is no substance to these allegations and close this

cc: IG, AIGO

                                     Page 3 of 3