We tend to have a quick reaction to them, and they flush out differences in values quickly, too.
A few days ago, American journalist John Bohannon wanted to make a big splash with a post titled, “I fooled millions into thinking chocolate helps weight loss. Here’s how”.
It’s part of a publicity campaign started by two German journalists co-producing a documentary with German and French TV broadcasters. It’s a takedown of the diet industry, set to air next weekend.
As hoaxes go, this one is particularly complicated. And “hoax” is a pretty slippery concept, anyway.
At one extreme of a very broad spectrum, a hoax can be a straight-up crime – like fraud, and scams to swindle people out of money. Police “stings” hover towards this end of the spectrum, because they can veer into the crime of entrapment. (There’s debate about whether stings are as effective as other policing approaches, and whether they have only a very short-term effect or might increase crime.)
Way down the other end of the spectrum, a hoax can be a parody or performance art that’s snuck into the mainstream with just enough ambiguity that some people don’t realize it’s not “real”. It might be intending to jolt awareness, educate, or draw attention to ideas or a point of view. It’s not really intending to deceive.
This chocolate hoax is still unfolding, so we can’t have any idea where on the spectrum of seriousness or impact it’s going to land. We have been reacting to bits of a very stage-managed picture.
So far, the journalists have told us some details about four components:
- creating a website with an address and contact point for a fictitious research institute (in Germany);
- a real study (in Germany) on chocolate and weight loss designed to be rubbish science that would all but guarantee statistical artifacts that were marketable and misleading;
- a pseudoscientific article with authors’ names altered, claiming to show chocolate is a “weight-loss accelerator” in an English journal (pulled offline by the journal in the last few days, but available here); and
- a press campaign touting the study to bait journalists into covering it, including paying for testimonials.
Before we get to it, though, because it involves a genuine randomized trial, let’s look at what governs the conduct of this kind of research in Germany. Frameworks vary a lot internationally, even regionally. (Michelle Meyer, from Harvard Law School, has written a very interesting blog post on this hoax, from the perspective of the rules that would have applied if they’d done it in the US instead.)
Doctors are in a particular position of power and trust, with their patients, and in society. As long as they are acting in a medical capacity with an individual, they have considerable freedom.
However, when there’s an inherent possibility that another interest has the potential to come before the interests of the individual, then their wings may be clipped.
One of those areas is doctors doing human research. That’s not to say that research is always inherently against the individual’s interests. Just that the new dimension changes the “social contract”. It’s a position I’ve long agreed with, although many don’t.
In Germany, if a doctor (or certain others) is going to do research in people that might affect their mental or physical health, or with identifiable data, she or he must first lodge an application to an ethics committee. That details the study plan, numbers of patients, issues around privacy and data handling, how people will be indemnified against harm, and so on. There are additional research areas where this is needed, but that’s not relevant here.
With a non-drug study outside a university, the ethics committee responsible is from a state medical board like this one (called a “chamber”). German legislation makes the international Declaration of Helsinki binding for doctors too. (There are summaries of the German arrangement in English here [PDF] and in the background of this article.)
In this interview with Bohannon, he said there was no submission for ethical review (and in a later interview with Charles Ornstein from ProPublica he claimed there was no need for one). There’s been no public statement from anyone else involved on this question that I could find by internet searching in both German and English. Update: On 10 June, Gunter Frank, the doctor who according to Bohannon “ran” the study, confirmed in an email to me that there was no submission for ethical review, and it needed to be real people not actors: the broadcaster wanted real footage.
On the 28th, NPR reported asking Bohannon a follow-up question about ethics, but no answer has been reported yet. (Update: In the documentary that aired on 7 June, the journalists reported that they wrote the paper “totally without medical help”, although they had it reviewed by an ectotrophologist, a German type of nutrition scientist.)
Here’s how Bohannon and the two German journalists whose project this is, Diana Löbl and Peter Onneken, have described what happened. You can see Löbl and Onneken in action here in this short clip (warning, it’s disrespectful of overweight people to say the least).
Löbl and Onneken wanted to “reveal the corruption of the diet research-media complex by taking part”, according to Bohannon. They approached a medical scientist for advice about how you would do a scientifically rigorous trial, so they could make sure, she says, that they did the opposite.
Bohannon was called by the pair in December 2014. Löbl and Onneken “used Facebook to recruit subjects around Frankfurt, offering 150 Euros to anyone willing to go on a diet for 3 weeks. They made it clear that this was part of a documentary film about dieting, but they didn’t give more detail” (according to Bohannon). According to Löbl, she and Onneken did the recruiting, wearing white coats – because “clothes make the person”, she said – when they recruited 16 people in January. (Update: It was filmed for the documentary: you can see it here 5 minutes in.)
[Update] On 31 May, an interview with Frank, appeared in his local newspaper. In answer to reporter Ingrid Thoms-Hoffman’s question on whether the patients know they were part of a hoax, he said: “No, they had no idea when they came to my practice in January. Four days ago, I wrote to them and explained in full.” He reported they were recruited over a TV station’s Facebook page, and they were mostly from Heidelberg. (Note: all translations from German in this post are my own.)
When researchers intend to deceive people about the real purpose of a study, there are additional ethical hurdles. They are expected to justify why they can’t be told the truth. And to demonstrate there’s a potential for benefit that outweighs the burdens of inconvenience and potential harm people will be signing up for under false pretenses.
Well, there were no benefits to accrue from the study itself for the participants: when the ones who had to change to a “low carb” diet (2 out of 3 of the groups) had to eat more fat and protein and less carb every day, when they were all getting their blood tested, taking urine samples every day for weeks, logging their sleeping, eating, weight, moods etc every day – the data could never advance knowledge about dieting. Here’s an old post of mine on the risks of multiple testing. Each time you test, you run a small risk of striking a data anomaly. But the more you hunt, the more this risk snowballs.
The study couldn’t tell us anything we don’t already know about its true objective either, the diet research industry and journalism. We already know how easy it is to publish bad science, and that there’s low quality journalism. And actually, comparatively few journalists fell for it.
[Update] When asked by Thoms-Hoffman what he needed to prove with this study, since he had long been critical of the way dubious research is used to manipulate, Frank replied: “For me, there wasn’t anything more to prove. Nevertheless I was happy to join in. We live in a modern age, and a smart TV report makes this con trick easier for people to see”.
If you’re going to deceive people, you have an ethical obligation about how you reveal the deception to them afterwards. That’s the subject of the question NPR has posed to Bohannon: just how are the 16 people finding out they were duped, if that’s what happened? (Here’s hoping it’s not by first seeing that YouTube clip.) [Update – Frank reported that he wrote to the participants four days before the 31 May interview. However, in the documentary, he’s filmed explaining it was a trick to one of the participants, and reporting that he told each of them one-on-one.]
[Update] And what about the data analysis? Bohannon wrote about it this way: “One beer-fueled weekend later and… jackpot!” When Thoms-Hoffman asked Frank what the result of the study was, he answered, “The one we planned to have, that is, that chocolate helps weight loss”. “And that was a lie?”, she asked. Frank’s reply: “Not directly. We used the usual research tricks to get the outcome we wanted. For example, we would ‘overlook’ an inconvenient result in a second measurement, some data were processed…the usual thing.”
[Update] Further details were reported in the half-hour version of the documentary shown on German television on 7 June. One of the research participants interviewed has diabetes. All 16 showed up to the follow-up session: they report that the data from 2 were dropped. They describe this is as done typically in scientific studies: drop-outs, they explain, means “we threw out the results that didn’t suit us”. In addition, before the final weigh-in, the people in the 2 non-chocolate groups drink a large glass of water to ensure they’d weigh more.
So was the study ethical or wasn’t it? One of the reasons many researchers feel very negative about these ethical review systems is that there’s often no consistency in the answers people come to on these questions. Meyer’s post goes into this important point in some depth. Maria Godoy’s report at NPR found contradictory opinions here too.
My personal position is that if we don’t agree with current standards for professional and ethical conduct, we can try to change the system, but we have to abide by it. If the standards are too low, we should aim higher – but we shouldn’t do less than is expected of us. To breach professional standards can be serious. And we can’t afford to erode public trust in doctors and health care professionals who approach people to participate in clinical trials.
Much was made in the publicity and documentary about conflicts of interest of others in the diet and media industry. Yet the participants in this episode demonstrated very little awareness of their own interests, be they in the success of the documentary itself, personal publicity-seeking, and book sales for one of them. And that’s the point of ethical review of research, isn’t it? To ensure that someone who does not stand to benefit from the exercise has checked that appropriate safeguards are in place. [Para added on 10 June.]
The argument in favor of this hoax is that it will have a benefit in terms of better journalism and perhaps even statistical literacy for enough members of the public to make it justifiable. Gary Schwitzer wrote: “I don’t know if Bohannon’s latest stunt will have any positive impact. Journalists and bloggers have already moved on to the next, much-quicker-than-24-hour, news cycle.” Meyer wrote: “What we need are feasible solutions to make these groups aware of this problem, not more evidence of the problem that perversely contributes to the problem itself.”
I agree with both of them. And there is some data to back that point of view: the real obstacles to better reporting are issues that won’t be affected by this, and even intensive training may not change behavior. Anyway, education has moved on from models based on shaming.
For what it’s worth, as a blogger who plugs away at statistical literacy writing and cartooning, I didn’t find the interest in reading my relevant posts as high as it gets when a media story challenges something people actually believed. (Last year’s spate of interest in resveratrol, for example.)
Faye Flam questions the grandiosity of Bohannon’s “I fooled millions” claim:*
Unless there was a suspicious spike in chocolate sales, it seems more likely that most people said “meh” or maybe even “huh?” but did not rush out to start a chocolate diet. For all we know from the evidence presented, 27 Americans were fooled. Bohannon’s claim plays to one of the most common and powerful biases – The desire to believe we’re smarter than other people. I’d never heard of the original study, but the exposure of the hoax is causing delight among my Facebook friends. People can’t resist reading about all those millions of inferior thinkers. Perhaps Mr. Bohannon fooled millions, or perhaps he fooled himself.
This gets us to the next slippery issue about this hoax: the journalists’ side. I think on balance the world’s journalists don’t get such a bad grade in terms of jumping to the string-pulling of the first campaign. But I wouldn’t give them high marks for their coverage this week.
First time around, the trio got some journalists from the infotainment end of the industry to dance the tune. This time, they’re getting the high end of the street to tell the story they want and promote their documentary, with very few sparing any words for the ethical issues.
I’ve mentioned some of the exceptions: in addition, Rachel Ehrenburg at Science News, Michael Hiltzik’s coverage at the LA Times, and Chris Lee at ArsTechnica really stand out. And Ed Yong led the charge on journalist ethics on Twitter. (The first I saw, anyway.) Seth Mnookin made a strong statement about falling for the click bait. (Update: And it turned out, Emily Willingham was posting about the same time I was.)
Hoaxing and the American press is the subject of a dissertation by Mario Castagnaro. It’s a fascinating look at the way the community’s attitude to hoaxes and journalists’ roles changed across several decades.
One of the things that struck me was that many see journalists’ hoaxes as a clever assault on the powerful. Yet, the more common hallmarks of the genre may be the ego and ambition of hoaxers – along with underlying disrespect for the public and other journalists.
Orson Welles’ famous War of the Worlds radio broadcast in 1938 was a cultural touchstone – and an unusually well-studied hoax. A revealing Welles’ quote relevant to his willingness to intentionally cause distress to bolster his career stood out for me: “I didn’t go to jail. I went to Hollywood.” Castagnaro singles it out as one of the turning points for society becoming more critical of hoaxes in the media – and expecting journalists to expose hoaxes, not undertake them. It was, he argues, key to the professionalization of journalism.
Clifford Irving’s Howard Hughes biography hoax in the early 1970s was another, changing direction from more gonzo journalism that had begun to encourage sensationalism, ego trips, and little regard for “objective distance” between the journalist and subject. (Irving did go to jail, because his hoax involved perjury and larceny.) Over time, hoaxing in journalism, Castagnaro argues, “came to be more readily associated with manipulation and the distortion of ‘truth’…a dangerous betrayal of public trust”.
I’m a cartoonist, so it’s no surprise that I am a big fan of parody and satire as a vehicle to communicate ideas. I don’t think there’s a fundamental breach of public trust there. Sure, occasionally people “fall” for an Onion story, but that’s not an abuse of power and a position of trust. We really need journalists to know the difference, and to call out abuses of power – not generate them.
* Note added after original posting: Millions would have seen the original story – the front page of Bild Zeitung alone would be millions. Millions seeing is not the same as millions being fooled. (Bild Zeitung has since pulled the story from its online version, noting that the science was a hoax.)
The first version of the post did not specify the detail of the “low carb” diet intervention (updated 1 June 2015). I also added a link to an earlier report addressing ethics that I hadn’t seen when writing this post (by Chris Lee), and one written while I was writing mine (by Emily Willingham).
On 8 June I added a reference to an interview with John Bohannon by Charles Ornstein from ProPublica.
On 10 June I included Gunter Frank’s communication to me that there was no ethical review. I deleted a paragraph about there needing to be confirmation that someone other than the participants had ensured that people were treated ethically, and added one (noted).
Disclosures: I was (and remain) an outspoken fan of a journalist’s April Fool parody in 2006 (for an invented drug that of course could not be used, embedded with educational materials about overdiagnosis). My experiences bias me towards the regulatory side of this fence. I have spent time on national ethics committees in Australia, including working on revisions of guidelines for medical research. I’ve been a member of two ethics committees for journals: The BMJ several years ago, and PLOS One’s Human Research Advisory Group now. I’ve also participated in the regulation of professional conduct for medical practitioners in Australia. When I lived in Germany, I had close personal and professional ties with Ingrid Mühlhauser, the scientist these journalists consulted about research methodology.
* The thoughts Hilda Bastian expresses here at Absolutely Maybe are personal, and do not necessarily reflect the views of the National Institutes of Health or the U.S. Department of Health and Human Services.