Category Archives: Health

all things morbid and mortal

Persuasion and Preventing Alzheimer’s

One of the more interesting observations I learned from a teacher is this:  If you think you understand something, change it.

The power of the observation is apparent.  Your knowledge should permit you to move an idea forward and backward, up and down, inside out and outside in, if you truly understand it.  You can make it do your bidding because you understand.

Push this observation and you realize that a persuasion orientation is a crucial part of science.  If you know something, then you can take your knowledge and change things.  The proof of what you know is not simply you know it, but that you can change it and change with it.  If you think you know something, but you cannot change it or change with it, then you don’t know after all.  As in the case of preventing Alzheimer’s.

Here’s the table.  Click to enlarge if needed.

According to the analysis that produced this table, 50% of Alzheimer’s cases worldwide are preventable.  Preventable.  That means they don’t happen.  Ever.  People don’t get Alzheimer’s.  Ever.  Half of all cases of Alzheimer’s don’t happen.  Talk about knowing something that makes a change!

Now, just take a moment and reflect.  First, realize that this is not original research, but a quantitative review of the literature.  Second, virtually all of that review is from Observational Research with no randomization, comparison, or control, in other words, little experimental data.  Third, see the Small Windowpane effects for all the risk factors.  Fourth, and hold on to your head for this.

Barnes cautioned that the findings assume that the individual risk factors actually cause Alzheimer’s, which has never been proved. “We are assuming that when you change the risk factor, you change the risk,” she said in a statement. “What we need to do now is figure out whether that assumption is correct.”

Can you imagine standing in front of your family or coworkers or neighbors and saying something like that?  “There’s no proof that this works, but dammit, I think we should all pull together and . . .”

Barnes is more than willing to assert that she understands something and understands it so well that she can change half of the outcomes from a negative state to a positive state all the while asserting that her knowledge has no proof.  Furthermore, realize that all of these numbers are pure speculation, that back of the envelope economics you read in advocacy columns with assumptions, historical trends, available data sets, and what ifs.  There is no good scientific evidence in this report that a particular factor, like physical activity, is actually causal for Alzheimer’s prevention.  And, the authors admit this.

Think about this report from another direction.

You are the persuasion maven tasked with an intervention to prevent 50% of Alzheimer’s case based on this research.  Barnes and colleagues confidently assert they understand how to prevent Alzheimer’s which means we know how to change it.  Okay.  What you would do to create this change?

Low education is a risk factor.  Think about that intervention.

Now, diabetes.  Think about that intervention.

See any difference in the scale at which you’d plan and execute?  Education requires a total organized effort at the national level and in the US at the state level.  Changing education is essentially changing the political and cultural institutions of the nation.  And, countries vary on that.  Diabetes, by contrast, is slightly more under the control of specific individuals, particularly lifestyle behavior like diet and exercise.  Changing that is essentially what we’ve been doing already since we know that diabetes is a serious health problem and does have controllable elements.

The list of seven factors operate at different levels of abstraction (global, cultural, individual, behavior, heredity), yet are all in the same basket called Preventable Risk Factors.  The factors are so different from one another that it makes a mockery of the term, Risk Factor.  There’s no rigor, clarity, or unity in the classification.  Thinking about the persuasion implications of this report makes you realize just how bad the science is behind it.

What also worries me as the persuasion guy charged with making the Life Saving Intervention, is that bad science like this gives people false hope.  I have no trouble with this getting published in the peer review literature.  People smarter than me are thinking the same kind of critical thoughts about the vaporous analysis, but at least we’re thinking about the problem and how to solve it.  This weak work might stimulate better research from someone else.  Publication in peer review literature is good here.

But, this crew went out of its way to attract media attention and did a notably bad job of explaining the nuance.  This is a fantasy, What If?, kind of research projection that frames a problem.  Seen in that light, it might be useful, although you clearly see that I don’t walk in that light.  The authors contribute to public confusion with their persuasion efforts for media attention.  Various Cool Table FauxItAlls will now confidently assert that if society would wake up, We Could Cut Alzheimer’s By 50% In Our Lifetime!  Of course, you could put a trillion dollars on Alzheimer Prevention and it would probably have little effect on the number of cases because this research is so weak.

Let’s shout “Cure!” in a crowded hospital waiting room.  Some researchers think that their scientific speech functions the same way in public that it does in peer review literature when obviously it doesn’t.  Just folks are in no position to critically evaluate this report and they assume that the Shouting Expert has enough sense, good taste, and prudence to not holler when she should whisper.

All Bad Science Is Persuasive.

P.S.  By dumb luck a new issue of Journal Watch (Aug 1, 2011, vol 31, no 15) arrived with a couple of studies on diabetes reduction.  One tested for more intensive management versus normal clinical care for asymptomatic diabetics and found no improvement in outcomes for the intensive management over 5 years of testing.  The other study tested the impact of adding daily exercise (walking) for type 2 diabetics over 1 year and found no benefit over a control group.   Do you have any other ideas for reducing diabetes to reduce Alzheimers?


Persuasion with Ratios – Sophistical Statistics


What, you say?  What?  Your statistical analysis came back null and void and unpublishable?  And, you followed the Observational Tooth Fairy and her huge convenience samples, huger number of variables for adjustments, and infinite logistic regression?  And, it’s still null?


Try analyzing ratios instead.


Yeah, instead of taking just one variable, combine two variables as ratios, then play the Epi Statistics Guitar.  For example, let’s say you want to prove that salt causes heart attacks, but your Tooth Fairy dataset resists that Fable.  (Here’s a surprisingly accurate news report.)  Now, make a ratio of salt consumption divided by potassium consumption and use that ratio to predict heart attacks.  How’s that work?

For sodium-potassium ratio, the adjusted HRs comparing the highest quartile with the lowest quartile were HR, 1.46 (95% CI, 1.27-1.67) for all-cause mortality; HR, 1.46 (95% CI, 1.11-1.92) for CVD mortality; and HR, 2.15 (95% CI, 1.48-3.12) for IHD mortality.

Isn’t that better?  Those HRs are actually Small Windowpane effects at 1.46.  Jeepers, when you looked at your first statistical analysis and found that salt consumption by itself was NOT related to CVD (yeah, heart attacks and such), but only related to all cause mortality (salt causes suicide?  car wrecks?  lung cancer?) you had a heart attack yourself (hmmm, Note to Self: next epi study focusing on effect of null results on heart function).  But, now when you create the ratio of salt over potassium, voila!

Of course, there’s not any literature in biology to prove this, but don’t worry. The Food Police echo chamber will amplify your message. Better still, your results will motivate a hungry cave dweller to run rat studies that vary salt and potassium intake on Mickey and Minnie.  That will take a few months and even if those studies disconfirm your ratio findings, you’ll be on to something else.

Persuasive Science never looks back!

All Bad Science Is Persuasive.

Drive with Science, Putt with Poetry.

P.S. And this ratio persuasion play is not limited to CDC persuasive science. Check out this report on BAC and the Injury Severity Ratio. And here’s the USA Today take on it. Notice how the writer entirely misunderstands the meaning of the Severity Ratio and confuses it with death and injury when that’s not what the report found. The writer’s confusion is exactly the intended outcome with Ratios and Sophistical Statistics. “Yeah, you could put it that way.”

Persuasion Pointers for Adultery

Want to motivate your partner to cheat on you? Here’s how.

Have your partner look at a computer screen that displays two pictures of individual people, the two pictures stacked one over the other. After a 500 millisecond view, the pictures disappear and in the location of one of them, a letter appears. Your partner then presses the key on the keyboard that corresponds to the letter on the screen. Do this task 10 times.

Got that? Two pictures of individuals. Look for a half second. The screen goes blank and a letter appears in the location of just one of the pictures. Press the key for the letter. Repeat 10 times.

The two pictures always show a person from the opposite sex (we’re doing this with heterosexuals, but the same thing should work with other sexual orientations with appropriate changes). One of the two pictures shows an attractive opposite sex person while the other shows an average looking opposite sex person.

Now, here’s the adultery part.

Everytime the screen goes blank, that following letter is placed over the location of the average looking picture. Thus, your partner is made to look away from the location of the attractive image and learns to anticipate having to look at the average face to complete the letter identification task.

This is happening so fast (every half second) and your partner is instructed to be quick and accurate on that letter pressing task. Your partner will never pick up on the implicit training of forcing more attention toward a less attractive image and away from the more attractive image.

If you think just a minute, you realize that this task creates Forbidden Fruit. The more attractive image is the Fruit and the letter pressing task Forbids It. Forbidden Fruit is another way of saying a Perceived Unfair Restriction On Behavior which is simply Reactance Theory. Unfairly deny something from someone and you create Forbidden Fruit and Reactance.

So, your partner should be falling off the log and in the thrall of persuasion gravity after this manipulation. How we would see this? Well, we cannot in good conscience now provide your partner with an opportunity to cheat on you in the lab, so we’re left with more ethical demonstrations. Consider these two self reports. First, relationship satisfaction and commitment.

We first examined the effect of the attention manipulation on participants’ perceptions of their current relationship. An analysis of variance (ANOVA) showed a significant effect of experimental condition, F(1, 40) = 4.54, p < .04, d = .65. Participants in the attention-limiting condition (M = 31.90, SD = 9.48) reported less satisfaction and commitment than participants in the control condition (M = 36.77, SD = 4.79).

Now, attitudes toward infidelity.

We next tested the hypothesis that the attention manipulation would cause participants to report more positive attitudes toward cheating on their partner, as assessed with responses to the ATI. An ANOVA revealed the predicted effect of experiment condition, F(1, 40) = 6.86, p = .01, d = .80. Participants in the experimental condition (M = 26.15, SD = 4.07), compared with control participants (M = 28.68, SD = 1.91), reported more positive attitudes toward infidelity.

Thoughtfully, DeWall et al. provide those d effect size estimates. They fall into the Medium to Large Windowpane size, between 35/65 and 25/75, so they are pretty obvious differences in key adultery attitudes. Ever so briefly, this task provokes more discontent with the relationship and more acceptance for adultery.

Quickly realize that there’s no behavior in this persuasion. No one acted on their bad attitude. But, there is that instant, if fleeting, change in attitude towards fidelity that here and now predisposes your partner’s desire for another person.

DeWall et al. add two more experiments that demonstrate other cognitive effects for this Forbidden Fruit test. They find that your partner has better recall for the Forbidden Fruit picture compared to the other picture and that they devote more attention to the Forbidden Fruit during that half second picture display. Thus, implicitly denying access to attractive partners provokes a coherent and organic cognitive response. Your partner wants to look at Forbidden Fruit, remembers it better, and finds more desire in the denied alternative.

Realize that this task operates on the Peripheral Route of the General ELM. We see no evidence of that High WATT, Long Conversation in the Head, but rather the quick and repetitive Ding-Dong of Forbidden Fruit, activating and suppressing desire over just a few seconds of observation. If we could retest your partner just a day later on just those attitudes about the quality of your relationship and the desire for adultery, it’s likely those bad attitudes would return to their previous state.

Take the practical lesson from this study. Actively suppressing your partner’s safe and acceptable awareness for another person produces the perverse effect of increasing desire. If adultery is outside your tolerance, you should say that, but this research explains why you do not need to ride herd on your partner’s normal observation of others. The difference between setting boundaries and denying normal freedom is your concern here. A little desire goes a long way for loyalty.

See General ELM in what you think of as interpersonal or relational issues. Loyalty and fidelity in intimate relationships is not typically what you think of when you consider persuasion theory. But the Forbidden Fruit of Reactance applies in all domains.

DeWall, C. N., Maner, J. K., Deckman, T., & Rouby, D. A. (2011). Forbidden fruit: Inattention to attractive alternatives provokes implicit relationship reactance. Journal of Personality and Social Psychology, 100(4), 621-629.


Point and CounterPoint: Water

Here’s a fun series of comments in the British Medical Journal on the proper amount of hydration for humans.  You’d think given the high scientific status and education of the people involved in this that there would be a Correct Answer of widespread agreement, but you’d be wrong.  Note that the comments, essentially eLetters To The Editor, contain references (!!!) and writer declarations of competing interests. Thus, everyone is smart, everyone has references, no one has a vested interest in nothing but the truth, yet no one agrees on how much water people should drink.  All that sincerity and still no Truth.

Persuasion mavens in the bottled water industry figured this out a long time ago which is why Americans spend over $60 billion a year on something they can get from any tap.  All that insincerity and all that profit!

All Bad Science Is Persuasive.


Selling Exercise or Blondes and Champagne

Persuasion is always about Something and nowadays a common Something is Health. You can make a lot of money and still do no good (or harm) for anyone by persuading on health. The killer app in this domain, of course, is life rather than death. You can sell a bucket of steam as long as it offers more life or at least less death. And, as I’ve documented on this Blog, most of the More Life Nostrums probably fall in that No Harm category which is another way of saying No Benefit either. Several recent metas have shown exceedingly Small effects, say 48/52, for exercise on More Life and given that most of these are observational designs even those effects must be considered with a wary glance and a quick check of your wallet. If you want to sell health behaviors, may I suggest that you focus on Benefits that are more likely to be More Likely and Actually Real: Cognition! I’d argue that exercise, for example, is more beneficial not for longer life, but more thoughtful, controlled, and decisive life.

Consider this meta.

Colcombe and Kramer found 18 intervention studies published in the peer review literature between 1966 and 2001 that specifically looked at the experimental (!!!) impact of aerobic activity in older adults (55+) on cognitive function. I want to highlight the studies they excluded here.

Articles were excluded if (a) the design was cross-sectional, (b) participants were not randomly assigned, (c) the exercise program was unsupervised, (d) the exercise program did not include an aerobic fitness component, or (e) participants’ average age was below 55.

The presence of random assignment is absolutely, positively critical for the quality of the science. None of that huge database of insurance or physician records followed by all that adjusting and transforming and then humongous statistical significance but piddling Effect Sizes. These are field interventions with random assignment to controlled conditions.  Older adults were randomly assigned to Treatment (aerobics) or Control (vigorous TV watching, maybe), then tracked over time for the impact on cognition.

Colcombe and Kramer looked at four particular kinds of cognition: Speed, Visuospatial, Controlled, and Executive. Here’s how they defined each category.

Effect sizes were coded as belonging to the speed category if the task represented a measure of low-level neurological functioning, such as simple reaction time (e.g., the speed with which one can make a manual response to a flash of light) or finger-tapping speed. Likewise, effect sizes were coded under the visuospatial category if the task tapped the participants’ ability to transform or remember visual and spatial information (e.g., viewing three line drawings and then replicating them from memory as in the Benton Visual Retention Task). Effects coded to represent controlled processes were from tasks that require, at least initially, some cognitive control (e.g., pressing one key when presented with the letter C, but pressing a different key for the letter M, as in a choice reaction time task), but do not have those characteristics exemplified by tasks in the executive category. Tasks in the executive category were those related to the planning, inhibition, and scheduling of mental procedures. For example, this category would include the Erickson flanker task, which requires participants to respond to a central cue (much as in a choice reaction time task), but simultaneously suppress a set of conflicting or irrelevant cues presented next to the target stimulus item.

Okay. So within these 18 higher quality peer review studies, what’s the impact of aerobic exercise on cognition in older adults?

An examination of the estimates for each task-process category revealed that exercise had the greatest effect on executive processes (g= 0.68, SE= 0.052, n= 37, p < .05); this effect was significantly greater than the effect of exercise on any other type of cognitive process. However, as is evident in Figure 1, exercisers also improved reliably more than control subjects on controlled (g= 0.461, SE= 0.035, n= 74, p < .05), spatial (g= 0.426, SE= 0.062, n= 23, p < .05), and speed tasks (g= 0.274, SE= 0.050, n= 32, p < .05).

Those g effects can be interpreted exactly as d effects (mean differences between groups divided by the standard deviation). The Executive effect at .68 is a Medium Plus or a Windowpane of 30/70. The effects for Controlled and Spatial are Small Plus or a Windowpane of about 40/60, and for Speed the effect is Small, about 45/55. Here’s a nice graphic display that compares the change over time in the Treatment Group versus the Control Group.

The bad and repetitive news here is that exercise doesn’t help that much on Speed of cognition. The practical upshot of this is apparent: Don’t play Jeopardy against kids. You can’t respond quickly enough after 55 even if you know the Answer in the form of a Question. Age kills speed in all forms. Yeah, Sweating to the Oldies with Richard Simmons will help on Speed, but just enough to lose at Jeopardy a little less. If you’re over 60 and running quarter mile intervals, you will still get your tight butt kicked on speed tasks.

The good news here is that exercise makes an obvious, apparent, Oh, Really? impact on Executive cognition. When you get those annoying cold calls selling insurance or time shares or whatever, the more you exercise the easier it will be to laugh in the receiver rather than getting confused and giving out your credit card or life savings to some Dilbert cube dweller in the Boiler Room. You can still figure out whether that large and colorful GREAT SAVINGS sign actually means you keep more of your money. In other words, the exercise effect on Executive cognition means you can be pretty much as smart (or as stupid) as you were at 35. It will be easier for you to wave off those friendly invitations to play Jeopardy. Instead you can make your way over to that sweet young blonde sipping champagne and looking bored. You’ll have everything you need: Executive, Controlled, and Visuospatial cognition.  And, your slower Speed will actually make you seem more thoughtful, wise, and dangerous.

So, here we are, sitting at the end of the bar, enjoying a splash of Taittinger with our new friend, content in the benefits of exercise, realizing advantages to aerobics we’d not considered. We may not live any longer, but we’ll sure have a lot to think with and about.

And, if you cannot sell that, you are no persuasion maven.

Stanley Colcombe and Arthur Kramer. (2003). Fitness Effects on the Cognitive Function of Older Adults: A Meta-Analytic Study. Psychological Science (14), 125-130.

doi: 10.1111/1467-9280.t01-1-01430

P.S.  That’s Grace Kelly in the Taittinger ad and neither the Princess nor the champagne appeared in the Colcombe and Kramer paper.  Just a persuasive thought to make the point.

P.P.S.  Your blonde may look different with no loss of meaning, implication, or possibility.

P.P.P.S  The effect sizes for the Control Groups are in the range of most of those Run or Die effects from the zealots trying to sell More Life Through More Sweat.  Interesting values comparison, isn’t it?  Run or Die versus Run and Try.


It’s about bin Laden, Stupid

This is based on a news story with hard hitting investigative reporting, so shakers of salt and shots of tequila.

But . . .

You’re part of the team trying to find Osama bin Laden.  You would like to get compelling evidence of his location, say something 2.0 and high tech, say something like DNA.  Yeah, baby.  Get a sample that matches a known sample from a family member and you’ve got a great lead on the guy.  But how do you get that DNA?

Persuasion.  Here’s the Box.

The CIA organised a fake vaccination programme in the town where it believed Osama bin Laden was hiding in an elaborate attempt to obtain DNA from the fugitive al-Qaida leader’s family, a Guardian investigation has found.

What a fabulous persuasion play.  A Peitho Award Nomination to the CIA team that figured this out.  Vaccinations have become cultural truisms, apparently now global and universal, that are so widely ingrained that no one thinks a second thought about any program that offers a shot.

To sink the needle even deeper the CIA elaborated the basic scheme.  Agents convinced a government Pakistani physician in a neighboring province to offer a free vaccine program in the area where bin Laden might be hiding.  It is not clear whether the physician was a willing participant in the plan or was misled.  He’s been arrested by Pakistani authorities which suggests they think he knowingly cooperated with the CIA, but Pakistan is in an uproar over this incident and could be simply keeping up appearances with this arrest.  Persuasion Is Strategic so it is crucial to hide your Strategy from everyone else.  We simply don’t know about the physician’s motivations, just his behavior.

The physician first began putting up posters in the poorest sections of the region and started the program there.  He then moved the operation to an office closer to the presumed bin Laden compound in hopes of attracting bin Laden’s children to the clinic.  Furthermore, the vaccine itself was advertised as being manufactured by a well known and trusted Pakistani pharmaceutical company.

The doctor went to Abbottabad in March, saying he had procured funds to give free vaccinations for hepatitis B.  Bypassing the management of the Abbottabad health services, he paid generous sums to low-ranking local government health workers, who took part in the operation without knowing about the connection to Bin Laden.  Health visitors in the area were among the few people who had gained access to the Bin Laden compound in the past, administering polio drops to some of the children.

The Guardian report discloses that it cannot prove whether any bin Laden family members participated in this program, but if any did, it would have been possible to obtain DNA from the family member and compare it to a known DNA sample from the sister of Osama bin Laden who died in 2010 in a Boston hospital.  Such DNA evidence, of course, would argue strongly for the likely presence of bin Laden himself.

If you’re a serious persuasion scholar you are already smiling at a huge irony in this play.  One of the more famous and productive persuasion researchers of the past 50 years, William McGuire at Yale University, developed Inoculation Theory based in part on that deeper understanding of “cultural truisms” or beliefs that no one questions.  McGuire found that it was fairly easy to attack those beliefs precisely because no one had thought about them.  When challenged, people have great difficulty defending a cultural truism and can be turned quite easily.  McGuire developed and tested Inoculation Theory as a means of strengthening these defenseless attitudes and beliefs giving rise to my Headline – the best defense is a weak offense, meaning that people will develop a better defense when they face weak attacks.

The CIA turned the inoculation metaphor a different way and used the thoughtlessness these truisms live in to mount a different kind of persuasion attack.  They discovered that while bin Laden was extremely careful in his movements, he apparently gave little or no thought about the risks of vaccines.  Hey, it’s a Pakistani government physician running a free clinic for poor kids.  Hey, the vaccine is manufactured by a trusted Pakistani pharma.  Hey, the clinic is just down the road from my house.  Hey, go ahead; send the kids to get vaccinated.

Hey, maybe I’ll get a shot . . .

P.S.  If a CIA robot program catches this, please pass it along.  I appreciate your service to America.




Sanitizers versus Sanity

Here’s the scene. Imagine a university campus with buildings and hallways and thousands of people passing through them. In the hallways are affixed 65 hand sanitizer units near doors and entrances for convenient public use. Now, to persuade more usage of these sanitizers during flu season, we’re going to design signs based on different persuasion theories, then compare theory-based signs to see which do better at motivating more usage. And, since we’re pretty smart, we will employ a Latin Squares design which means that each theory-sign will appear on every sanitizer unit for the same amount of time in a planned rotation. We’ll weigh each sanitizer on a regular schedule to count how much sanitizer glop is used.

Now. Here are the Sign Headlines for the 4 different theories the researchers employed.

1. Germs Are Out To Get You. Get Them First!

2. Everybody Is Doing It. Are You?

3. Stay Healthy This Season. Sanitize Your Hands.

4. H1N1. Getting It Is As Easy As Passing Me By.

Okay. From reading those headlines, tell me the theories . . . What’s taking you so long? This is theory based science, you knuckleheads. It’s easy. It’s obvious.

But not to me.

When I read those headlines I thought Door Number Two was a Norms-based appeal and pretty obviously so. But the others? 1 seems a kind of fear appeal. Yeah, 2 is the norm. 3 is, what? Confused, but positive. Perceived benefits? 4 is humor? Ironic? Another weak fear appeal. What?

According to the study authors, Headline 1 is Perceived Susceptibility from the Health Beliefs Model, 2 is indeed Norms from the Theory of Reasoned Action/Planned Behavior, 3 is a gain framed message from Prospect Theory, and 4 is a loss framed message from Prospect Theory.

This is a great illustration of the Tower of Babel problem that exists when people operationalize persuasion plays. I see Headlines 1, 3, and 4 as hopelessly confounded with the risk of a consequence whether positive or negative. All three address a common element in slightly different ways yet each is supposed to operationalize a unique and different persuasion concept. And, I suspect that you probably had trouble determining the “true” theory behind each Headline and also see disconcerting similarities between messages that are supposed to indicate different concepts.

The authors provide a kind of manipulation check on the Headlines. They asked about 40 people to read each Headline then rate it on a series of items on various unreported dimensions, but the actual items are not listed. Worse still, the authors cherry pick the results, highlighting items that “prove” the gain or loss or norm or whatever. There is no open, complete, and systematic manipulation check here. Thus, the authors own data indicts the study.

I’d also raise one huge concern about the absence of a true message control Headline, something like,

Here’s A Hand Sanitizer For Your Use!

Such a sign would control for pure information and notification effects that simply alert people to the presence of an object as they move through a busy and crowded hallway. Any sign, whether informative, persuasive, or merely graffiti, would attract attention and greater usage than a thingy hanging off the wall in a noisy field. To compare Any sign to No sign is not a good test of the effects of different theory-based signs.

Finally, the authors offer no rationale for testing the different theories against each other. They select HBM, norms, and framing apparently based on popularity or frequency of use, then set to compare them. Why? I can think of no good reason to assume that a smart persuasion agent could not effectively employ a wide variety of different theories and produce roughly equivalent outcomes, assuming you ran the right Persuasion Play with the right Persuasion Box. Yet, this crew picks a peck of popular plays and pits them against each other.

Let’s soldier on to the Results. Here’s the key table.



The percentages in the last column indicate how much more sanitizer glop people used when the signs were present. Signs on average produced about 50% more use of glop compared to the no sign baseline. Each Headline produced a statistically significant increase (those asterisks). I’m actually confused about what the effect size is. If the percentage is a kind of Risk Ratio, then the effects are Small Windowpanes. If the percentage is a mean difference, then the effects are Medium to Large Windowpane. The authors only provide the ethereal statistics and no sensible quantities like ounces or doses or bar presses or USDA Approved Minimum Daily Average. I simply do not know what the data mean.

The authors then make statistical comparisons between the 4 Headlines. Please avoid the Brooks Effect while reading this.

The gain-framed sign was predicted to be particularly effective in promoting hand sanitizer usage. Consistent with this prediction, dispensers with gain-framed signs received 12.5% more usage than dispensers in the other sign conditions combined, a significant difference (p < .029). Although the gain-framed sign performed somewhat better than the loss-framed sign, this difference was not statistically significant (p < .40). It is interesting to note that the worst-performing sign was the perceived susceptibility sign, which had 9.7% less usage than dispensers in the other sign conditions combined (p < .059).

Let me mix my metaphors: This paragraph illustrates the classic statistical fishing expedition that reads the tea leaves. There is absolutely no good a priori reason to predict that any one theory-based sign should produce better effects than any other. No way. Thus, to look for differences after you’ve collected the data is fishing. Furthermore, if you carefully think about those “p < .0thingys” you realize that Headline 3 Gain Frame appears to be different from the other three which are not different from each other. However, the authors read the tea leaves to claim a variable pattern of effectiveness. Given that there is no a priori reasoning to prefer one theory over the other and that the team clearly took a fishing charter after collecting the data, these comparative differences are meaningless.

Let’s get out of here. Signs are a great and simple technology for persuasion. Properly done, as in great advertising, they work. Further, this kind of data collection is a fabulous real world and practical design that is also incredibly easy to conduct, not like dying in the lab, case by case. Sure, the generalizability, the external validity, is always questionable, but the natural setting observes people in their daily lives rather than in a controlled and artificial lab context. But that does not guarantee good research.

This study employs theories that are proven losers in lab and field settings – HBM and framing do not work as advertised. I appreciate that you may disagree, but you then have the burden of explaining the terrible meta results from large and active research literatures. This paper makes no attempt to explain their results against the larger literature and meta-analytic findings. It convenience samples theories with no rhyme or reason.

Further, the method and results are clearly spotty. The manipulation check on the Headlines is weak and poorly reported with sketchy and confused results. The main analysis only proves that an object in a hallway gets more use with a sign than without. Any quantitative differences between signs are merely chance differences and cannot be predicted or explained by any theory.

Updegraff, J. A., Emanuel, A. S., Gallagher, K. M., & Steinman, C. T. (2011). Framing flu prevention—An experimental field test of signs promoting hand hygiene during the 2009–2010 H1N1 pandemic. Health Psychology, 30(3), 295-299.


Retirement Dimmer

I’ll just cite one example, this WSJ story, about an emerging fantasy in the public imagination.  Its lede reveals the nightmare.

For all the benefits of retirement – less stress, fewer obligations, finally the time and money to travel the world – new research paints a somewhat bleaker picture.  Without careful attention and some preventative steps, retirement, it turns out, may take quite the toll on our faculties.

If you spend anytime viewing the pop press, you’ve seen this same idea expressed with different words.  Retirement makes you stupid.

Except if you read the scientific literature you discover that there’s no good science to support this assertion.  Only glib, facile, and young epidemiologists reading the tea leaves.  Like this one.

Here’s the abstract.

Background Occupational work involves many factors capable of protecting cognition. The ‘disuse’ hypothesis suggests that removal of such factors at retirement may increase the risk of cognitive decline.

Objective To examine whether retirement is significantly associated with cognitive change after adjusting for preretirement cognitive function, personal, social, health and lifestyle factors, work characteristics and leisure activity.

Methods Participants were from the Whitehall II study, a prospective study of London-based Civil Servants. Short-term memory, the AH4 Part 1 (a test of inductive reasoning), verbal fluency and the Mill Hill Vocabulary Scale were collected at ages 38–60 years, and again, on average 5 years later, at 42–67 years, providing pre- and postretirement cognitive functioning assessments for 2031 participants (470 retired and 1561 working). Linear regression was used to test the association between retirement and cognitive performance adjusted for preretirement cognition.

Results Mean cognitive test scores increased between the two assessments. However, after adjusting for age, sex, education, occupational social class, Mill Hill score, work characteristics, leisure activities, and indicators of physical and mental health, those retired showed a trend towards smaller test score increases over 5 years than those still working, although this only reached 5% significance in one test (AH4; β=−0.7, 95% CI −1.2 to −0.09) and did not show a dose–response effect with respect to length of time in retirement.

Conclusions This trend is consistent with the disuse hypothesis but requires independent replication before it can be accepted as supportive in this respect.

Editor’s Note:  Insert the Dancer’s Laugh About Here.

If you read the literature on Retirement Robs Your Mind, you know this study is hey diddle diddle right down the middle.  Convenience sampling.  Huge numbers of cases.  Lots of adjusting (and in public no less!).  And what do you get?

Despite analyzing multiple indicators of cognitive function, jimmied, shaken and stirred with adjustments, and using the most lenient and misleading marker of “significance,” the researchers find only one test that reaches the 0.05 alpha level.  Cue up the Small Pitiful Effects drum roll.  Worse, there is no dose-response finding, meaning that the effect does not change over time even though that what the Nightmare asserts.  Thus, researchers using a shotgun filled with double ought buckshot fired multiple rounds at the broadside of a barn from a distance of five feet and just nicked a plank with one pellet.

Yet, they write.

This trend is consistent with the disuse hypothesis but requires independent replication before it can be accepted as supportive in this respect.

Editor’s Note:  Insert Dancer’s Laugh Here Again.

If you want to understand declining cognitive effects and what you can do for yourself, for someone you love, or for callow epidemiologists, please consult the work of Tim Salthouse.  And, if you don’t want to read the whole thing, here’s the Headline:  Aging Beats Everything.  The rock of your body is rolling downhill and if you are doing NYT crossword puzzles at work, the effect is only psychological.

Okay, so science says we’re all getting older and we will get that narrator quote at the start of Citizen Kane (at 8:45, “. . . as it must to all men, death came to Charles Foster Kane.”).  What’s this got to do with persuasion?

Well, of course, All Bad Science Is Persuasive.  Yeah, that’s obvious.

And, now, consider the persuasion opportunities for selling Longer Minds.  Man, if you thought it was shooting fish in a barrel with Lifestyle and Mortality, think about the persuasion spaces for Puzzles and Cognition!  Even young people worry about losing their minds, so this isn’t just a Boomer or BoomerPlus generation thing.  In fact, I suspect the persuasion box here is easier with the under 60 crowd.  Old Boomers have already seen the dimming of the light even with all those NYT puzzles.  Kids don’t know it yet, but still fear it.  All you need to do is read the epi literature on aging and cognition to find The Cure.

This is too easy.

P.S.  The Dancer’s Laugh is heard at the end of this Paula Adbul YouTube video.  It’s the expert’s mark of derision.

Exploiting the Health Bubble

I could do this everyday – which would provide a steady stream of PB posts and also destroy my will to live – but here are two stories that illustrate the possibilities for persuasion mavens.  Lede with the Big picture first.

Healthcare spending will represent nearly one-fifth of the U.S. gross domestic product by 2019 . . .

Now, Telling Detail from a fighting MD who is also a journalist.

Smoking cigarettes is the cause of so much preventable, deadly disease.  But now new research shows sitting for long stretches of time may be just as dangerous.

In a country where people can and will spend 20% of their income on one area and where experts in that area can’t understand their own research, you know you are in a target rich environment.  As Butch Cassidy would say to the Sundance Kid, it’s easy, ripe, and luscious.

Begin as always with the expert in the news.  Kim Mulvhill, MD, has proof that smoking and sitting kill at pretty much the same rate and to her credit provides a link to a research study to support that assertion.

She provides that link certainly to build her credibility by citing other experts, but certainly not to build her argument because if you read the abstract you quickly see the research makes no comparison between smoking and sitting, but only looks at sitting.  And guess what?  The Observational Tooth Fairy finds a highly adjusted Small Effect Windowpane under the pillow you’re sitting on.  The effect probably does not exist given this data and it definitely is nowhere near the range of effects for smoking (which range from Medium to Large Windowpane increases depending upon the outcome variable).  Dr. Mulvhill can misread and misunderstand to her heart’s content and still write in public without fear of laughter, derision, or lawsuit.  She shouts Death in a crowded sitting room without consequence.

And, she does so because people have dialed down their WATTage to a flicker.  When you’re giving away 20% of your GDP on something you can control with common sense and a little self control, it’s obvious you are not a High WATT worker toiling on the Central Route, building long, dense, and deep webs of elaborated thoughts about your health.  You’re skipping polished stones over an ocean of thought, a Cue kid laughing on a Peripheral Route beach.

As a persuasion maven you cannot observe this scene without smiling a toothy grin.  Hello, Little Girl. Experts are falling over their own incompetence to help Other Guys and the Other Guys are making automatic payments in a subscription model called Forever Young.

Right now everyone is happy.  Everyone believes that their life and health is increasing every year and they are willing to pay more and more every year for that belief.  Their beliefs about life and health have become psychological assets that are worth more and more without regard to reason, reality, or reservation.  That’s a Bubble.

Persuasion in health is so easy it is almost hard.  In an orchard of nothing but low hanging fruit which one do you take first?  Wipe that grin off your face, chose the one you find personally easy, ripe, and luscious, and get busy.

This Bubble will burst someday.  Not soon.  Not if there is a conservative revolution in 2012.  Maybe 2020.  When my Medicare bill is half of your income.

Until then it is easy, ripe, and luscious.


Bill Gates, Persuasion Maven

Bill Gates continues his health oriented charitable work with a new initiative in China aimed reducing tobacco use.  He’s teaming with a Chinese Internet leader, Baidu Inc.’s Robin Li, to start a new campaign.

Focusing on the harmful effects of exposure to others’ smoke, the initiative will aim to help smokers quit and encourage nonsmokers to ask people not to smoke around them.  At a news briefing the two billionaires donned green T-shirts reading “Say no to forced smoking” in Chinese.

Please focus on the T-shirt slogan: “Say no to forced smoking.”  Gates clearly understands the persuasion literature and is not following the zealots’ course here.  The slogan demonstrate two smart persuasion plays and one smart strategy.  Plays first, then strategy.

First, note the Say No phrase provides an active TACT instead of the novice negative TACT (Don’t!  Stop!  Desist!).  To a novice the difference between Do This versus Don’t Do That seems unimportant, but mavens have read the research and realize active TACTs are easier to persuade.  Points to Mr. Gates on this subtle play.

Second, now realize the artful Reactance finesse from the Forced Smoking phrase.  Typically you find a finger shaking Marm scolding naughty kids to Stop! Quit! Don’t!  This move, while satisfying to the Marm in all of us – doncha just love bossing your little brother or sister – frequently backfires for reasons explained with Reactance Theory.  Perceived unfair restrictions on your behavior provoke that oppositional, You Can’t Make Me! response you get from your kid brother when you go into Boss Mode.  Even if the Marm is correct, that unfair restriction activates an automatic and natural Reactant response.

You see how Reactance kills persuasion.  The Other Guys stop listening to your persuasion plays and at best actively resist you not on the merits of the persuasion, but because of Reactance.  At worst Reactance can turn the Other Guys into your enemies for life, killing future persuasion plays.  Reactance is a bad thing, a predictable thing, and when it arises, it only proves:  You are no Persuasion Maven.

Realize that the Forced Smoking phrase shifts the source of the unfair restriction from the Marm to that rude smoker sitting next to you.  He’s the one who’s messing around with your freedom of action with his blue cloud.  Say No to That Guy!  Thus, this slogan captures all the Reactance inherent in this moment and redirects it to the smoker through a positive and active TACT from you.

Finally, consider the strategy inherent in this slogan.  Instead of making direct attempts to end tobacco use (Ban Farming!  Ban Manufacturing!  Ban Distribution!  Ban Sales!  Ban Smokers!), this slogan aims at nonsmokers and developing in them attitudes and beliefs that will then move that society over the long and winding road this change requires.  Gates and his partners realize that tobacco attitudes, beliefs, and behaviors in China are approximately like those in America in the 1960s.  While there is scientific evidence about the harms, that evidence is not widely distributed throughout Chinese society.  It’s going to take a very long term commitment in China to change basic ideas about tobacco.  Just think about the fact that in 1960 a little more than 50% of adults in America smoked; 50 years later we’re down to about 20%.  Gates and crew clearly are taking a long term view and building a strong base with this focus on nonsmokers.

I’ve admired Gates and his persuasion skill before and my respect continues with this report.  He knows what he’s doing.