Pfizer, a Big Pharma is ever there was one, is losing its patent on Lipitor, a cholesterol lowering drug.  When Big Pharmas lose their patents within one year virtually all market share will be lost to generic versions of the drug.  With Lipitor that is an $11 billion loss in sales for Pfizer, about 18% of total revenues for the company.  That’s the way it goes in the Big Pharma business, but Pfizer is trying to hold on to this brand and continue selling it against generic competitors.  So far Pfizer is:

Offering insured patients a discount card . . . Paying pharmacies to mail Lipitor patients offers for the $4 copay card and to counsel patients that Lipitor lowers bad cholesterol more than rival drugs and helps prevent heart attacks and strokes . . . Keeping U.S. marketing spending nearly level until the last minute . . . Negotiating unusual deals with some insurance plans and prescription benefit managers to block pharmacists from dispensing generic Lipitor.

Hey, Pfizer!  Where’s the persuasion in this?  You’re just pounding away on power with legal and financial maneuvers.  You know those discount plays burnout in the long run.  Why not practical persuasion, too?  Remember the Rule:

You Can Get Farther with a Kind Word and a Gun Than with Either Alone!

Mavens!  You have an interesting opportunity to help a struggling business AND do persuasion science.  If Pfizer is looking at an $11 billion loss, they might consider testing your persuasion plays for saving Lipitor.

How would you design a persuasion campaign that uses communication to keep old users and attract new ones to Lipitor?

$11 Billion.  $11 Billion.  $11 Billion.

Physical attractiveness is a great variable for demonstrating the classic answer for every psychological phenomena:  It depends.  Especially in persuasion, no variable has a monolithic, absolute, always and forever, simple, main effect.  All persuasion variables are contingent which can be expressed in my Rules:

There Are No Laws of Persuasion.

All Persuasion Is Local.

Nothing always works in a simple way and you need to know the local conditions, the Box and Play, before you can explain, much less operate.  So too with Beauty.

In persuasion, Shelly Chaiken’s 1979 paper remains foundational for its excellence in both theory and research.  If you’ve never read it, you must or else you’re just wearing Maven Badges and are little more than an Eagle Scout who still gets lost in the woods.  Chaiken’s work is the exemplar here.

Which brings me to a new book from an economist on the value of attractiveness as understood through dollars.  He find a lifetime earning advantage for beautiful people over ugly people of . . . what do you think?  Again the contrast is between the top and the bottom of attractiveness.  We’ll look at the rest of us near uglies in the middle in a moment.  How much more as a percentage do the Brad Pitts earn than the Plug Uglies?

How about 3%?

Remember the Windowpane here.  A Small Windowpane would be 10% in this application.  So, according to this economist, beautiful gets 3% more than ugly day after day and across their lifetimes.  The great majority of us in the middle – rather like that little girl’s soup preference of not too hot nor too cold – earn about 1.5% more than the Plug Uglies.

These are stupendously Little Effect Sizes and rise to importance only because economists employ those huge samples in their observational designs which guarantees a statistical sensitivity that borders on dysfunctional.  Only 3%?  And comparing the highest to the lowest?  What gives?

It depends, baby.  It depends.

If you read or remember Dr. Chaiken’s 1979 study you know that she found important interactions, the Local, in her research.  Beauty does not operate in a social vacuum, but moves in the mess of life where the presence and absence of This, That, or the Other modulates the obvious impact of attractiveness.

Friendliness, for one large example, is a crucial Local variable.  Beautiful, but Mean, strikes the beholder very differently than Beautiful and Likable.  Think now about a Local of great importance – your health and proper treatment of an illness – and contrast Beautiful, but Incompetent, against Beautiful and Skilled.  And . . .?

Of course!  The dual process models!  Chaiken’s HSM or the ELM!  No persuasion variable can possibly have a main effect because these theories describe, predict, and explain the mess of life in the Local and the interactions of all variables in the Box and Play.

The economist study here attempts Variable Analytic Research, the VAR, aka, Dustbowl Empiricism.  Take one variable, run it through everything, and find the average effect which more often than not flutters closely around zero, that daredevil moth near the candle’s flame, just past the heat, but always in danger of bursting into the null.

When your research is only skin deep, like observational economics, you miss the mess of life and the opportunity to bring it into study.  Treat everything but one variable as confounders, covariates, or mere annoyances, then partial, adjust, or fudge them and you get one third of a Small Effect that explains one thing to the gullible, but leaves mavens wondering.

P.S. That is a diamond body tattoo from Yair Shimansky.  It is the current world’s record for most expensive diamond tattoo.  And, no, Melanie.  You cannot have this.

You think you’re smart.  Handle this.

The Obama administration is wrestling with the thorny question of whether scientists should inject healthy children with the anthrax vaccine to see whether the shots would safely protect them against a bioterrorism attack.

You remember anthrax.  Some knucklehead released that hound from hell in 2001 shortly after the September 11, 2001 attacks demonstrating yet another weapon of mass destruction.  We’ve got a vaccine for anthrax that works pretty well with adults, but we don’t know about kids.  Consider these sequential practical persuasion problems.

1.  Persuade parents to volunteer their healthy kids right now before an attack.

2.  Wait until an attack then persuade parents why we didn’t test the vaccine on kids before and why we don’t know right now what will happen to their kids if they get the vaccine.

Don’t you just love the ambiguity, uncertainty, and risk?  This is just one more reason why academic research is sometimes so prissy.  You can avoid the consequences of situations like this; shootfire, you don’t even have to think about them much more than reading this blog post.  By contrast, if you’re in the Fed, you’ve got to deal with these terrible choices as an unavoidable part of your work.  If you think you’ve got the science to handle this, I’ve got a job for you!

Think about the dissonance properties for the scientists actually facing these choices.  You’re supposed to be smart and figure it out, but there are no correct answers and no matter what you decide you can be sure that the reality, when it arises, will play out differently than you expected.  But, you nonetheless made a thoughtful and deliberate choice that will nonetheless find aversive consequences.  That puts you on the dissonance path.  Dissonance reduction will then tuck you tighter to your earlier decisions, making it that much more difficult to see your failures and how to change.

And, of course, during all of this you are still trying to change the Other Guys, all those parents to either risk their kid now or to accept your refusal to test before the attack.  Jeepers, persuasion is hard enough when you only worry about what you are doing to the Other Guys without the situation doing something to you.

There Are No Laws of Persuasion.

But, sometimes you’ve still got to do it and be effective.

Odysseus can be understood as a persuasion maven who achieves more with how he thinks and speaks than with only how he acts.  The contrast with Achilles instructs here.  The Iliad shows him a relentless killing machine motivated by pride; it’s all about him.  Odysseus instead accomplishes the goal everyone seeks but cannot attain through careful observation of the Other Guys, a proper understanding of the Local, and a master’s grasp of the Rule, Persuasion Is Strategic Or It Is Not.

Past Odysseus as persuasion metaphor, I recommend Zachary Mason’s book, The Lost Books of the Odyssey, a wild reimagining of Homer’s tale.  Mason take events from the Odyssey and shifts perspective for dramatic and thoughtful consideration.  He takes the tale of the Cyclops and tells it from the point of view of the Cyclops rather than Homer’s perspective through Odysseus.  The tale becomes painful and poignant when you think and feel through the monster rather than the man.  Mason’s book is not a linear narrative retelling of the tale of Odysseus, but rather is what the title implies: Lost Books that could have been told or written. Mason amazes me with both his ranging imagination and his voice; the guy knows how to write with tones and structure from Borges.  If you enjoy literature, you would probably enjoy The Lost Books of the Odyssey.

To what end does Odysseus pursue his persuasion?  Tennyson closes Ulysses with this perception.

Come, my friends,
‘Tis not too late to seek a newer world.
Push off, and sitting well in order smite
The sounding furrows; for my purpose holds
To sail beyond the sunset, and the baths
Of all the western stars, until I die.
It may be that the gulfs will wash us down:
It may be we shall touch the Happy Isles,
And see the great Achilles, whom we knew
Though much is taken, much abides; and though
We are not now that strength which in old days
Moved earth and heaven; that which we are, we are;
One equal temper of heroic hearts,
Made weak by time and fate, but strong in will
To strive, to seek, to find, and not to yield.

When we ran the 1% or Less! Milk Campaign in the 1990s one problem shouted in our faces.  Low fat milk tastes awful!  How can you possibly persuade people to switch from that tasty, but unhealthy whole fat milk to that thin, blue, and ucky tasting low fat swill?  Being empirical, somebody in the project actually hit the field and did blind taste testing.  That unscientific (no randomization) formative research nonetheless and surprisingly demonstrated that people weren’t very good at tasting the difference in a truly blind taste test.  Somewhat reassured, we plowed on with the project, achieved huge success, fame and fortune, film at 11.  Or something like that.  Yet people insisted that low fat milk tastes awful and everyone knows it.

I was at one of those Cool Table conferences after we’d published our findings and I described the campaign and why it worked for all the other Cool Table experts like myself assembled in the room.  They hooted at me.  Just started laughing and blowing raspberries at me.  No way.  And, so, Kim Witte, a fabulous and talented researcher, took it upon herself to build and conduct a blind taste test among the assembled Cool Table experts.  Kim handled the event with skill and charm, ensuring the process created enough noise and error to muddy the results thus allowing everyone to hold their opinion with their heads held high.

But, my expert friends and esteemed insult artists were wrong then and are wrong now.  Here’s the data.

A research team in Arizona conducted blind taste tests in WIC Clinics (Women Infant and Children) across the state.  While waiting in those clinics people were recruited for the test on the spot.  Each person took a drink from a cup then stated whether they liked it and guessed whether it was whole, 2%, 1%, or skim milk.  The order of the types of milk was random for each participant.  Exactly 100 people participated in these blind taste tests.  Here are the results.

If this was a random variate, you’d expect about 20 counts in each cell, but clearly that didn’t happen.  A chi-square test shows a result of 55.3, p < .000000001, and a w effect size of .744, a very Large Windowpane.  There’s something very non-random going on here.  Now, given my hooting experts friends and the Wisdom of the Crowds (Low Fat Milk Is Awful!!!), you’d expect this non-random distribution to skew to the top with a lot of correct guesses, but you can obviously see this is wrong.  Fewer than 10% got most of their guesses correct.  Even if you say that 50% or more correct (2/4) is a better criterion, only 36% hit that mark even though Low Milk Is Awful!!!

If you are in the food and beverage business you are utterly unsurprised at this news.  Most people can’t tell Coke from Pepsi, diet from regular.  People rarely carefully taste what they put in their mouths and their response to the taste is highly dependent upon other factors in the context.  Thus, when they do a controlled comparison they are usually at a loss to discriminate one taste from another.  Turns out that milk is no different, thus, it was actually not as difficult for our campaign to overcome the Wisdom of the Crowds and the Cool Table.

You can read the paper for more gory details here (MilkTasteTest pdf), but let me express the Main Point with professional reserve, accuracy, and charm:

I’m Right and They’re Wrong!

Nah-nah.