Health Partners Research Foundation published a study on ethical research conduct with the University of Minnesota in this months journal Nature (subscription) (435, p737). The researchers surveyed 3,247 scientists who received support from National Institute of Health (NIH) grants. Gross scientific conduct is considered plagiarism, fabrication or falsification of results. The studied confirmed previous data that showed that this type of misconduct is infrequent, however the study found that there are other behaviors that are considered 'less problematic' that nevertheless seriously compromise the integrity of science:
"Thirty-three percent of our survey respondents admit[ted] to one or more of the top-10 behaviors. [T]he scientific community can no longer remain complacent about such behaviors..."
The "top ten behaviors" included changing data, failing to present oposing data, and unethical use of ones own data. Other behaviors in the 'top 16' included inappropriate research design and inappropriate assignment of authorship as well as inadequate record keeping. 27% of study respondants said they kept inadequate research records.
The study has been criticized by several scientists for asking questions where the responses were difficult to interpret, for instance the San Francisco Chronicle quoted David Magnus, director of the Center for Biomedical Ethics at Stanford who was critical of two questions, as were other scientists.
His point is well taken. Since the respondants were limited to "yes" or "no", its hard to evaluate certain answers. For instance about fifteen percent of the total cohort (about half of those surveyed returned usable surveys) said they had changed "the design methodology, or results of a study in response to pressure from a funding source." But what exactly does this mean? Science research is a collaborative process. Since research is peer reviewed, it is rare that a study is published without revision since often additional experiments are required to further test or validate a result. Editors and peer researchers review studies according to variable critera, but the process is generally rigorous. As well, grants come under tremendous scrutiny before approval. Again, this *can be* political, however the intense competition makes well thought out grant writing essential. Sometimes grants need to be scaled back to accomodate funding restrictions, or sometimes researchers will decide to come at a problem slightly differently due to feedback or if they get a particular preliminary result. However this is part of the process. It shouldn't be looked at askance.
On the other hand, all the behaviors in the survey potentially skew the presentation and interpretation of research, so while scientists may understandably defend their discipline, the results are problematic. One interesting trend in addition to the fact that 33% of the scientists admitted to at least one of the behaviors is that there wer significant differences between the younger and older cohorts, in the number of scientists who admitted to each behavior. 38% of the mid-career scientists admitted at least one of the behaviors, compared to 28% of the early career scientists.
