There is plenty written in the research literature about bad research. Sometimes that bad work is cited by a number of other researchers, before it gets corrected or retracted (if ever). What this page is about is when bad research gets picked up by journalists and their eager editors who cook up click-worth headlines.
“Cheese really is crack. Study reveals cheese is as addictive as drugs” – Los Angeles Times, October, 2015;
“Science Says Cheese is Basically Cocaine” – GQ, February, 2016;
“Cheese Is Addictive as Cocaine, Science Says” – Men’s Health, December, 2018.
After the story spread, the L.A. Times followed up with a modest measure of invitation to skepticism,
“Is cheese really like crack? Drilling down on the science behind the viral cheese study” – Los Angeles Times.
An op-ed from Time was more explicit,
“No, Cheese Is Not Really ‘Like Cocaine'” – Time, October, 2015
The problem is that the study cited in these stories found no such thing.
Researchers at the University of Michigan asked participants to report which foods in a list were most difficult to cut down on or eat in a controlled way.
Yes, the actual study was a survey. Cheese by itself actually got a middling score on the rankings.
Still the damage is done. The idea has not gone away. The actual Mayor of New York City, a pseudo-vegan of some influence, cited the idea and went as far as to get sciencey with it.
“Mayor Eric Adams says it’s hard to tell the difference between a person ‘hooked on cheese’ and someone ‘hooked on heroin'” – Business Insider, February, 2022.
In these parts we appreciate when a politician speaks in declarative testable statements instead of regurgitating trite platitudes! But in this case Mr. Adams has gone full-on Donald Trump on this one.
“You take someone on heroin, put them in one room, and someone hooked on cheese, put ’em in another room, and you take it away, I challenge you to tell me the person who’s hooked on heroin and who’s hooked on cheese.”
The hypothesis is explicitly stated and experimental method is proposed. I say we lock Mr. Adams in a room with recovering heroin addicts as a combination discussion session and null control. I’ll kick in twenty bucks and a few pizzas to help fund the study.
Technical section: As noted, the original study is based on surveys, of non-random participants [NEWS FLASH: College kids like pizza!]. Foods listed were then cross-correlated using naive statistics based on criteria pre-selected by the authors. Long debunked terms like BMI and p-factors were key to the math involved in the rankings. Some discussion is given to theories on how different ingredients and secondary chemicals affect humans. No practical data is presented or reviewed.
“Which Foods May Be Addictive? The Roles of Processing, Fat Content, and Glycemic Load” – Schulte, et al, 2015, PLoS One.
The good news is the published research doesn’t actually claim to prove anything at all.