If you can’t trust the evidence, there’s nothing left

I hope others are as concerned as I feel about this article, from The Chronicle of Higher Education (“Journals Find Fakery in Many Images Submitted to Support Research”), giving examples of where Photoshop or similar programs had been used to fabricate or manipulate images to support research conclusions.

Editors at scientific publications are, it seems, having to act more like detectives to test the images supplied with articles for signs of tampering. The root of the problem, I guess, is the ease with which modern software allows images to be manipulated and ‘enhanced’, coupled with researchers’ desire to present compelling evidence for their conclusion (even when it doesn’t quite exist).

It’s not an isolated phenomenon either. The author of the article, Jeffrey R. Young, says:

Ten to 20 of the articles accepted by The Journal of Clinical Investigation each year show some evidence of tampering, and about five to 10 of those papers warrant a thorough investigation. . . (The journal publishes about 300 to 350 articles per year.)

We all tend to assume that scientific papers are paragons of objectivity and accuracy. After all, human lives might depend on what they report, given the tendency for other scientists to draw on previous work to substantiate their own claims.

The reality is that some ‘cleaning up’ of data is nearly universal and the temptation to clean up some fuzzy image may not seem wrong to a generation reared on the use of software to improve even family snaps. In the past, I encountered several reputable scientists who would cut ‘outliers’ from data without a second thought, based on the comforting grounds that they were “almost certainly” random mistakes. Were they? Maybe, but there’s no doubt they made the data presented look rather more compelling than if they had been present.

Unless there are some thing in life we can trust, more or less implicitly, there will no longer be any route to the enhancement of knowledge. Trust is such a fragile thing that even a fear of being mislead can destroy it. Like Caesar’s wife, our scientists need to be above suspicion or they will forfeit much of their credibility and claims for funding.

Technorati Tags: honesty, credibility, trust in science, ethics, ethical research standards, truthfulness, trust