>>> Posting number 316, dated 13 Jul 1996 14:57:46 Date: Sat, 13 Jul 1996 14:57:46 -0400 Reply-To: Discussion of Fraud in Science Sender: Discussion of Fraud in Science From: "W. R. Gibbons" Subject: Re: anyone could win Comments: To: prg@REDACTED.clas.virginia.edu, ljipps@REDACTED.berkeley.edu In-Reply-To: <199607130029.UAA110642@faraday.clas.Virginia.EDU> On Fri, 12 Jul 1996, Paul R. Gross wrote: > To Gibbons, et al.: Nothing in my question had anything to do > with concern about sloppy, bad, or dishonest science. The reasn > I follow this list -- the only reason -- is that I am as > concerned about it as are you. My question was about how much of > "science" is bad, sloppy, dishonest. That is NOT a trivial > question. Let me try again; I jumped on you and Professor Lipps a bit harder than I should have, and indeed I did quote him out of context. I apologize for that. Friday was not a very good day, in part because of issues at least peripherally related to this discussion. During my scientific lifetime, I seem to have encountered more than my fair share of dishonest science and shady tactics. Certainly I have seen more than many others on this list say they have seen. Science may be self-correcting if, as Professor Lipps says, the scientific process is allowed to run its course. However, dishonest or misrepresented results should bother us whether or not we think the scientific process will be allowed to run its course. Even if truth does eventually win out, it will only be after considerable expense in money, time, and perhaps careers. I would judge that very high profile results, if they are immediately recognized as important, will generally be tested fairly rapidly as others attempt to build on them. Large clinical trials may be an exception, because they are particularly difficult to replicate no matter how important the data may be. My impression is that a significant percentage of sloppiness and dishonesty occur in such trials, where the results are particularly difficult to correct and may affect the lives of patients. While high profile results will be tested in many cases, results that seem more pedestrian may not be tested, but may influence thinking for years. If they are wrong, it may take a long time for the truth to out. While we are waiting for the scientific process to unmask dishonest, misrepresented, or sloppy results, those who take the pains to do honest and careful work and present it fairly are at an extreme disadvantage. In today's environment, they may perish while the corner cutters prosper. If the corner-cutters win, will they police each other? It seems unlikely. PRG asks what percent of science is dishonest? I don't know; I suspect it is pretty small, albeit not as small as some would like to think. What percent is misrepresented? Again, one cannot say. It depends in part on what you mean by misrepresented. What percentage is sloppy? There, I think we may be getting into some rather significant numbers. But maybe a more serious question than absolute numbers is this: where are present forces pushing us--toward better science and more honesty, or in the other direction? I suspect tenure will slowly erode in higher education over the next decade, until it provides little security and little academic freedom. That, I think, will be an incentive to cut more corners and will push more people over the fuzzy line into dishonesty. In our lives, we have seen cases where people cheated for recognition and advancement in a fairly non-threatening and secure environment. What will they do when the issue is survival? I think I was accurate when I pointed out there is, today, little money for replication and little recognition for those who discover errors or dishonesty in the work of others. The emphasis everywhere is on positive outcomes, and that's where the money is. That is another disturbing trend. We have overproduced Ph.D.'s for years in many fields, in part because of our hunger for cheap and dedicated labor. Is the generation of a pool of underemployed Ph.D.'s conducive to greater honesty? I find that unlikely. So even if I might agree that the present level of outright dishonesty is low, I anticipate it will increase. As for the future of sloppiness, have you noticed that Baltimore and Imanishi-Kari are saying they were fully exonerated, because they were only sloppy? I will close with one personal anecdote. I once developed a hypothesis based on work that had been unchallenged for a decade or so. Trouble was, I could not replicate the original work. This led to a year of work to find that the previous work was erroneous, and correct it. When the result was submitted to what is probably the most prestigious physiology journal in the US, one reviewer remarked that the subject was significant, and our work was done with considerable care. The reviewer felt we were undoubtedly right, whereas the original work was undoubtedly wrong. He did not recommend publication, however, because publication would "needlessly create problems in the minds of physiologists." If I have concerns about the self-correcting nature of science, maybe I can be forgiven. WRG Ray Gibbons Dept. of Molecular Physiology & Biophysics Univ. of Vermont College of Medicine, Burlington, VT gibbons@REDACTED.med.uvm.edu (802) 656-8910