Ambiguity is persuasion’s best friend.  When you are uncertain, persuasion principles drive your thinking, feeling, and acting.  If you doubt this, consider health and safety, or more properly nowadays, Health and Safety.  I’m going to take one example, use it to kick open a door, then drive a truck through it.

Statins.

Those wonderful little pills that reduce cholesterol, have virtually no side effects, and help everyone to live longer and healthier lives.  Except that’s not true.  Not even close.  The original research demonstrated that among people with existing cardiovascular disease, statins did help, but at a rather small effect.  Now add in fear (of death) and confusion (among patients and doctors), and soon almost anyone was taking a statin.  Instead of acting as a curative agent that reduced disease in sick people, statins became preventative agents that reduced the risk of disease ever starting in healthy people.

But the data never supported that shift from cure to prevention and today even the medical establishment is beginning to say it loudly out loud.  Journal Watch reports a nice summary of recent summary studies on statins.  Let me quote the headlines and the lead paragraph.

Statins and Primary Prevention for Patients at High Risk for Heart Disease.

The benefit, if any, is very small.

Statins – which clearly confer an all-cause mortality benefit in patients with known cardiovascular disease (CVD) – are prescribed extensively for primary prevention in patients at high CVD risk but without CVD.

The article then nicely summarizes results from several recent studies that demonstrates “preventative” statins had not even statistically significant effects on extremely large samples (68,000 cases!) in carefully controlled experiments.  In other words, among people who were not sick with CVD, but at risk (smokers, hypertensives, obese, etc.) the statins had no effect on mortality.  And, doctors prescribe and patients take statins like the pills were breath mints.

And this is not news.  If you doubt me, I’m not going to make the case here because you won’t believe me in a blog post.  Read the research yourself.  And read it.  Look at the numbers.  Think about the numbers.  Find the absolute rates, not those Sophistical Statistics like relative ratios that persuasively inflate outcomes.  And don’t look just at statins.  Look at the persuasion on:  vitamin D, omega-3, hormone replacement therapy, obesity, prostate cancer, screening tests.  For people who are already sick, various interventions do produce helpful, real, if small, effects.  But past that limited application, most of the Health (and Safety) research over the past 20 years operates more like persuasion than science.

Consider again just the instance of statins.

Even many children know that many people die of heart problems so that anything that reduce the chance is a Good Thing.  And, if a little statin pill helps sick people, won’t it help not so sick people?  If you don’t think carefully about it, the answer seems to be yes.  So, when you visit a doctor and she says you might consider this little pill that is covered in your health plan, tastes like nothing, and is easy to swallow with your breakfast juice, why not?

Now, here’s the persuasion analysis.

The pill has no effect.  You don’t know that, but because somebody in a white coat gave it to you, you think you are doing the right thing.  And the right thing here is not thinking for yourself, but doing what someone else suggests.  Thus, you are making external attributions (my doctor) rather than internal attributions (me) for your health.  You don’t have to think.  You can proudly stroll along the Peripheral Route cue-ing along with Credibility (your doc) and Comparison (everyone’s doing it).

Worse still, you are now also cue-ing with Commitment/Consistency meaning the next time your doc recommends something, say a screening test, you are more likely to comply simply because you did it in the past, so you need to stay consistent with that past commitment – taking a statin.  And, maybe that next New New Thing comes in a pill and your kindly doc gives you a free sample which you use without thinking, then thoughtlessly continue with in a Reciprocity cue (your doc gave you something – the free sample – and you need to give something in return – a paid order covered under your health plan).

Small effects, fear, and confusion combine to make health and safety not a science problem, but a persuasion problem.

You might remember an earlier post when you see the lineup.

Yeah, that’s right.  The pretty one is in red.

A little over a year ago Andrew J. Elliot and Daniela Niesta published a series of fun experiments aimed at testing the Romantic Red effect.  Let me quote myself:

In all five studies, men always rated the same woman more positively when her picture appeared against a red background.  (With women, the rating did not vary – a girl is a girl is a girl to other girls in this application.)  What surprised me was the size of the effect.  Expressed as the Windowpane, the effect ranged from a high of 75/25 to a low of 62/38 with an average across all five of 68/32, or a strong medium effect.

But, as noted at the time, these studies only tested the effect of red with women targets, not men.  Well, now Elliot and colleagues have run a similar set of experiments with men as the target and guess what?  Romantic Red still applies.

This new series of experiments from Elliot et al. employed a major interesting wrinkle compared to the first set of studies of female targets.  In the experiments with men, Elliot and colleagues varied the ethnicity of the male in the photo.  Some featured a Chinese male, others a Latino male, and others a Caucasian male – all raters were always of the similar ethnicity.

In general, the Romantic Red effect applied with male targets the same way it did with female targets.  Women rated a man either on a red background or wearing a red shirt as more attractive than the same man on a non-red background or non-red shirt.  The effect sizes were also big coming in from medium to large to jumbo.  And, finally, when men did the rating, color made no difference – a boy is a boy is a boy to other boys just as a girl is a girl is a girl to other girls.

Here are the grains of salt . . .

We’re talking pictures, folks, and not even talking pictures.  The raters never interacted face-to-face with the experimental targets in this research.  Would the Romantic Red effect apply even when that boy or girl started talking?  Would it apply when it was live and not mediated?  This research suggests you might want to wear a red sweater on your first date, but who knows?  If you’ve been around the block a couple of times, you know even the most gorgeous red-wrapped face and body you’ve ever seen can become repulsive when he or she starts talking.  Furthermore, you don’t need to read the experimental research to know that Red sells more stuff than just girls or boys.  Red is a hot marketing color for all manner of nonsexual goods and services.  So, does Red make people sexy or is Red a hot color.  Finally, realize that Red may get attention, but . . .

. . . you probably could take it too far.

The New New Thing asserts:

You think your car radio is broken because it doesn’t display the name of the song and the artist. You tap a word on a paperback and wonder why the definition doesn’t automatically pop up. You swipe a digit across the screen of your cell phone and all you get are fingerprint smudges . . . Our touchscreen existence has literally rewired our brains . . .

Yes!

What shall we name this New New Thing that literally reWires our brains?

iPavlov.

Hey, kids, it’s not the technology that makes you; it’s the thoughtfulness.

reWired is only Low WATT zombies ambling along the Peripheral Route, Cue-ing along, Ding-Donging a song.

Again, Whitehead says it nicely.

Civilization advances by extending the number of important operations which we can perform without thinking about them.

Even if you are the least interested fan about the 2010 World Cup soccer tournament, you know that there were serious problems with the referees.  They made a lot of whopping bad calls that even the most casual of observers could spot.  What to do?

Nudge, baby.

1.  Add Referees.

2.  Add Technology.

3.  Increase Scoring.

4.  Redefine “Offside.”

5.  Rethink Penalties.

6.  Reduce Faking.

Really.

This is Nudging.  Those little things that make big differences.  Little things like adding more key personnel, changing primary rules, spending more money.  Subtle.  Nuanced.  Bright in a deep way.  Deep in a bright way.  Nudge.

And, not that there’s any bias in Nudge:

Consider the Consumer Financial Protection Bureau now being established. Above all, I’d urge the head of this agency to devise rules under the assumption that, someday, he or she will be succeeded by a nitwit.

Gee, the Obama appointee to this new Bureau (yet to be named and confirmed) won’t be a nitwit?

Nudge as Oracle!

I’ll probably lose my setting at the Cool Table for this, but I have access to the Nudge for Democrats in the November elections.  If you are a Republican or some other nitwit who isn’t voting for Democrats, stop reading this, otherwise Nudge Along!

1.  Add Democratic Voters.

2.  Delete Republican Voters.

3.  Hire Nudge Consultants.

All Bad Persuasion Is Sincere.

Persuaders Can Either Be Famous or Effective, But Not Both.

There’s a Difference between Persuasion, and Smoke and Mirrors; With Persuasion the Illusion Lingers.

A consumer research firm, My Type, has broken through the chatter today with its scientific conclusion that iPad owners are “Selfish Elites.”  Of course, this conclusion is probably not true, but it’s a great headline that hits a prejudiced sweet spot.  (And, while I’m probably a Selfish Elite, I don’t own an iPad.)  Let’s look at details.

My Type uses a research method that is not scientific, but sounds like it is.  It’s kinda like kids dressing up in their parents clothes and playing Mommy and Daddy.  My Types dress up like scientists by using science sounding words, but all they are doing is poaching Facebook.  For example, they offer free Personality Quizzes that require you to give them complete and eternal access to your Facebook profile.  They save this information in a huge database.  Then they do “research” by comparing different groups – say Facebookers who own an iPad versus Facebookers who do not own an iPad – on the Facebook profile data.  In fact, this particular “research” on selfish iPad owners is done exactly like this.

Thus, if you want to be honest rather than cunning, you’d have to say today’s infamous results are based on:

1. people who’ve taken a My Type Personality Test (or other such product) and;

2. who’ve given permission to My Type to access their Facebook profile data.

Not exactly what you’d call a random sample or even a representative sample.  Now, it is a big sample.  My Type says they looked at data from over 20,000 people.  A number that big sounds useful, but it is not.  First, even stupidly small differences from foolishly big samples are “statistically significant” but practically useless.  Second, a big and biased sample is not better than small and biased sample – the sample is still biased.  And third, there’s no replication other than the pinging sound you hear when a dishonest shot hits the prejudiced sweet spot.

If you look at the “report” from My Type which is available from my old friends at Scribd, you’ll find tables and figures and numbers.  Download them.  Study them.  Repeat them at parties.  Just don’t bet any money, reputation, or consequence on them, because they are not honest, accurate, reliable, trustworthy, scientific, . . . gee, you get the point.

Now, even more interesting is how My Type is reacting to the coverage of this contrived report.  They are clearly monitoring the conversation and jumping in to clarify as needed.  Here, for example, someone claiming to be with My Type offered comment and amplification to the Wired reporting.  Let me quote:

3) We’re not claiming to be ultra-scientific here. The data is rigorous, the interpretation of the data is up for debate. Again, look at the report, lots of the raw affinities are given there (and “affinity” is defined). You can come to your own conclusions.

Ultra-scientific?  It that like Ultra-Pregnant or Ultra-Cool?  I haven’t taken or taught a research methods course in ten years, so is Ultra maybe a new technique?  And the “data is rigorous” but the “interpretation . . . is up for debate?”  Huh?  This sounds like an answer you write on an essay exam you forgot was scheduled for today.   (It’s also fun to read the rest of the comments from Wired readers.  Many of them smell the rat.)

Now, are there demographic and personality differences between people who own iPads and those who don’t?  Sure.  And can you market iPads differently for different people?  Sure-sure.  And, can you believe My Type’s results to guide your marketing?  No-no-no!

On this one, I’ve got to hand it to the folks at My Type.  All Bad Persuasion Is Sincere and there is absolutely nothing Sincere about this “study.”  But it brought them attention which might mean clients which might mean money.  However, I wouldn’t hire them to research anything beyond their own foolishness.