T he Investigation
This case study has many similarities to the recent cases of fraud that have occurred in physics research. Because the events transpired at Sara Bridges's institution, it is the responsibility of that institution to instigate an investigation into the case. Thus, Bob Smith must notify the appropriate people at Bridges's institution, which is responsible for following the federal guidelines in putting together the process for investigating the case. That institution must also notify the funding agencies. It should be noted that although Bob Smith uncovered the possible fraud, it is not his--or his company's--responsibility to investigate the case further.
One of the problems in a case like this is the large amount of publicity that surrounded the publication. Because of this--and because the possible fraud was discovered outside the research group's home institution--the case must proceed in a public fashion. In other cases, where the possible fraud is discovered within an institution, it is my opinion that the investigation should be carried out in a fairly private way. This is to protect the individuals involved if it turns out that there is no fraud. One of the problems with very public cases is that no matter what the outcome, they tend to do considerable damage to the reputations of the people involved. Many institutions have guidelines as to how to proceed in such internal cases. Even so, the funding agencies must be kept informed.
Who Is Responsible?
The case clearly brings into focus the issue of responsibility of co-authors. The six authors all played different roles. Mary Greene and Toby Marshall contributed only research material and were not responsible for the primary data or conclusions. Thus, they should not be held responsible for the paper, except for their specific contributions. The roles of Sara, Tina, and William were all different, and each bears some responsibility for the total content of the paper. William was the supervisor of the research, so he had a major responsibility to ask pertinent questions regarding the data. Sara obviously got much of the credit for the work, so it was also her responsibility to ask hard questions about the data. Tina was more like a reviewer, as she simply helped with the writing.
The key question in determining responsibility is this: Did these authors have any reason to suspect the data? In the recent Schön/Bell Labs case in condensed-matter physics, the results were so phenomenal that most people believe that some of the co-authors should have been asking harder questions. Here, it is not clear that this is the case, although it is said that the results are dramatic. Surely, if you believe that your results are going to change the world, you ought to be more careful than usual--and Sara, William, and Tina should have been so.
That said, one could argue that it is too much to expect that any of these three would be able to catch this type of fraud and that, in fact, the scientific process did work. After all, the fraud was detected as soon as attempts to reproduce the data were made. On the other hand, what does it mean to supervise a graduate student if you are not making sure of the quality of his or her results? We, as scientists, tend to simply trust other scientists until we are explicitly shown that we should not (then, we completely ignore them as crackpots). We clearly do not want to create an environment in science in which we are suspicious of one another. At the same time, having your name as co-author on a paper and taking much of the credit for the work mean that you must also take responsibility.
But what does it mean to be responsible? Sara, William, and Tina did not commit fraud. This question is the hardest issue in such cases and usually is determined by the future actions of the institution, funding agencies, and the community. Because scientific research communities place enormous trust in their colleagues, perhaps the loss of trust by your fellow colleagues is the primary punishment.
A final point that some people raise is that authors' individual responsibilities are typically not indicated clearly in the published paper; in most journals, each author's contributions are not explicitly listed. Clearly, it is not hard, in cases of fraud, to delineate the various roles, but some believe that these roles should be made more apparent to the readers of the publication. This might actually be a useful step, and the American Physical Society is discussing allowing the authors of a paper to insert a footnote describing each author's role.