The Reasons for Retraction

Publications related to food habits are important if you’re in the weight loss field; I rely on them to help people achieve their weight loss goals. If the studies were poorly done, that’s unfortunate but behavioral science is an inexact science anyway. But if someone intentionally manipulated the data to get a specific outcome, that’s just not right. Let’s see what several scientists found when they examined Dr. Wansink’s data more closely. What were the problems?

It seems there were three. First, as I mentioned on Tuesday, he had a graduate volunteer continue to examine the data to come up with hypotheses that were significant. That means they organized the data differently and kept running statistical analyses until they came up with something that was statistically significant. As I said, that’s a no-no because of the potential of finding something by chance; you get the best answers to the questions you actually ask, so finding something by accident is not as valid among scientists.

Second, there were errors in the way data were displayed. The reviewers made a very big deal of granularity and how the means displayed weren’t possible. I’ll leave that to the people who specialize in statistics.

Finally, they accused him and his colleagues of plagiarism. If there were an absolute violation of science, that would be it.

But as you might expect, not everything is always as clear as people make it out to be, and I’ll explain that on Saturday. Until then, I would still keep the snacks out of sight and continue using that salad plate instead of a dinner plate in order to eat less.

What are you prepared to do today?

Dr. Chet

 

Reference: BMC Nutrition. doi.org/10.1186/s40795-017-0167-x.