Sunday, October 28, 2012

Misleading the people. What else is science for?

Brigham and Women’s Hospital recently retracted a press release regarding a new study on the link between drinking diet soda with aspartame and blood cancer.  This is a great illustration of the crazy state of health-related journalism, science publications, and manipulation.

What should BWH have expected when a press release stating this link was released?  A couple of things are obvious:
  • Any media coverage would be overly simplistic.  It would omit important details that don’t fit into a short headline.  It would omit important details that aren’t as good at increasing media ratings.  It would omit important details that might be over the heads of the readership, even if they are important to the applicability of the study.
  • Once the mainstream media gets it, it moves to the media’s social channels.  Even reputable and well-meaning journalists are limited to 140 characters in their tweeted headlines.  Many of them use a blog strategy that uses a short format so they can get out several per day and serve their readers with broad news coverage.
  • Then the general public gets their hands on it and reblogs, retweets, and otherwise circulates it.  Their headlines are shorter, omitting more details, unrepentantly inciteful rather than insightful. 
  • Soon the meme is viral and the intentional manipulation comes in.  How hard is it to take a scary medical finding or a promising health link and create the “New Alaska Diet” plan, blog, eBook, and then bestseller?  Or the snake oil to ward off blah blah mutations for only five easy payments of $29.99? 
 What was really the finding? 
  • The finding was only found for men.  The researchers don’t know how aspartame can affect one gender over another.  Without a scientific explanation, the statistical association is questionable.
  • There are lots of ingredients in diet sodas.  Any other one of them could also be the cause.  Or it could be a combination of several ingredients that are all required to elicit the effect.  There is no way to know.  Each one has different implications for what should be done about it.  Aspartame might be fine as a sweetener if you take out some other chemical that is not necessary.
  • The size of the increase was lost in communication.  What if your risk of blood cancer (non-Hodgkin’s Lymphoma to be exact) goes from 1 in a million to 2 in a million?  Would you care? 
  • There are other possible effects of switching sodas.  What if you switch to regular soda as a result and your risk of Type II Diabetes goes up to 10 in a million and complications of Diabetes to 3 in a million?  On net, you made yourself worse. 

Behavioral science in public safety

A company in India has come up with some great ways to leverage behavioral science to improve public health and safety.  The company is called FinalMile.  This is a great analogy for their strategy because they intervene at the proverbial sharp end of decision making, just when the person is about to make an unsafe choice.  Here are a few examples that I find particular imaginative.

To keep people from walking on train tracks they changed the way train horns are designed and used.  There used to be the usual long slow constant booming tone. This is easy to hear and pretty clear what it means.  Train operators blew it from far away to give people plenty of time to get off the tracks.  But these things violate what we know about behavioral psychology.  If I know I have a long time to get off the tracks, I can keep walking along them because they are more convenient – which is why I was there in the first place.  The long slow tone was also not very visceral.  So they changed it to a series of changing, sharp quick bursts.  This creates a cacophony that is instinctively scary, even if your conscious brain knows what it is.  They also reduced the lead time, so when you hear it, it is louder. And you know you have too little time to think about it, so you get off.

They also replaced the danger signs along the tracks with a photo of a real person running screaming in terror from an oncoming train.  It still tells you what you are supposed to do, but is much more visceral.

They also recognized that a lot of people stop taking medication when their symptoms go away, even when they are still sick.  We have heard of this frequently with antibiotics, but there are other diseases as well, including highly infectious ones.  So these patients don’t get better, spread the disease to others, and develop drug resistance in themselves and drug-resistant disease strains in the population.  Not good.  So they used digital printing and personalization – adding a photo of the patient him or herself looking sick.  It reminds you that you are not really better yet.  I think they could add a second photo (not in the same photo because you don’t want to reduce the clarity of its single focus) that shows you getting other people sick as well.  This exploits our social bonding instincts – none of us want to feel guilty or be blamed by our friends and family (or boss) for getting everyone else sick. 

To improve the use of public trashcans, they used photos of pleasant-looking neighbors throwing trash in the public cans and bad-looking strangers throwing it in street.  This leverages our self-identity resonance (I blogged about this recently here or a technical psych definition here), our social bonding instincts (we want to be as good as our in-group), and our competitive juices (if they can do it, I sure as hell can!!).