DIMBERU: The right kind of science

In the past couple of months, we’ve seen two high-profile lessons about how science, as a way to uncover truth, is supposed to work. The first case was inspiring, the second disheartening. Both have something to teach us about doing science.

Last fall, a team of researchers from the OPERA Collaboration in Gran Sasso, Italy shocked the world with their report of clocking neutrinos — electrically neutral subatomic particles — traveling faster than the speed of light. As the speed of light had long been considered the universal speed limit, this announcement prompted equal shares of excitement and skepticism. Could Einstein’s theory of relativity actually be wrong? When a second attempt of the experiment showed the same 60-nanosecond difference between the time it should have taken the neutrinos to travel and the time it did, it looked like the physics world would be turned upside down.

But as the saying goes, “If it’s too good to be true, it probably is.” A few weeks after the paradigm-shifting results were first announced, the researchers found a small error in their measurement that could account for the 60-nanosecond discrepancy. To the layperson, this may have looked like a huge failure. No doubt the scientists were embarrassed. However, this is precisely how science works. Experimental results are provided to the scientific community (usually in the form of a publication in a peer-reviewed journal) for confirmation and eventually to inform new lines of research. The researchers that made the discovery did not entirely believe the results themselves, and they wanted to have the greater scientific community help them validate the findings. The point is not whether the results were correct or not (though I’m glad Einstein wasn’t wrong), but that the truth won out and the scientific method prevailed.

Juxtapose the neutrino story with one that continues to unfold from the University of Tokyo. Prominent molecular biologist Shigeaki Kato is currently under investigation by the university for allegedly manipulating data from dozens of publications over more than a decade. These include several articles in top-tier scientific journals including Cell, Nature, and the Proceedings of the National Academy of Sciences.

Suspicion of Kato’s alleged scientific misconduct came to light when a 2009 publication was corrected in late 2011. While the authors claimed that the correction would not affect the results and conclusions of the original publication, they admitted to having “inadvertently duplicated or erroneously created” images. This admission raised more than a few eyebrows, and when one curious researcher pored over 24 different publications from Kato’s group dating back to 2001, he discovered a myriad of images that appear to have been doctored. The anonymous researcher brought the allegations against Kato by making a nearly six-minute video showing the manipulated images and posting it on YouTube. After watching the video, I can say that the evidence against Kato appears convincing. However, it would be prudent to wait for the university’s investigation to conclude before drawing any conclusions. But Kato’s group has already retracted two publications from 2004 and 2007, citing unintended plagiarism in both papers.

If the allegations of doctored images are true, it would be a monumental fraud. Science is always based on previous work, whether it is one’s own previous work or that of others in the same field. Publication of false data wastes the time and money of others who may then base their research on the false data. There are certainly times when an honest mistake ends up in a paper, and most researchers will publish a correction when they or someone else picks up on the error. But a decade’s worth of data manipulation is not a mistake. It is a deliberate masking of the truth to advance one’s career.

Science does not need punitive measures against those who are wrong, such as the team in Switzerland, because being wrong can still teach us something. However, intentional falsification of experimental results harms everyone involved, and it harms the integrity of the scientific method. This is not the right way to do science, and I hope that, as it did in the neutrino story, the truth will prevail here as well.

Comments

  • CharlieWalls

    Seems like apples and oranges to me. One involved numbers from machines and the other “images that appear to have been doctored.” In the first case, the last I’ve read is there was a loose connection causing erroneous readings. Pretty much hard science v. soft stuff (topic not even mentioned). Secondly, a more interesting focus might have been the validity or morality of non-refereed attacks by YouTube.