Here’s James Hansen, NASA climate scientist and loud voice on human-induced Climate Change. He got arrested at a protest in front of the White House.

Hansen also took to the microphone before getting handcuffed.

Before his latest arrest Monday outside of the White House, Hansen used a microphone to implore President Obama to act “for the sake of your children and grandchildren.”

We do it for the children!

Darryl Hannah also attended and wore tight blue jeans with a peasant blouse and tweed riding jacket. Her hair looked great. But, a picture is worth a thousand words.

Hannah also spoke.

She told Canada’s CTV network, “I want to add my body and my voice to the thousands of others who are laying themselves on the line and saying, ‘No, we do not want to be party to this incredibly destructive path.’”

No dancing bears were observed at the scene.

Some environmentalists apparently read the Persuasion Blog.

David Roberts, an writer with environmental blog Grist, questioned whether it was smart of Hansen to protest in this fashion. “I know I’m not supposed to say this, but James Hansen managed his transition from scientist to activist *terribly*. All influence lost,” the green blogger tweeted out prior to the arrest.

To quote again that underappreciated persuasion theorist, Donald Rumsfeld, you go to persuasion war with the army you’ve got, not the one you’d like!

I’ll take the Jag and you get the ‘Vette.  We’ll test them and us on Dead Man’s Curve (YouTube).

The headline:  Watching NASCAR racing causes more aggressive driving and accidents in West Virginia!

Scientific science proves it.

Now, of course, to be truly scientifically scientific, we must acknowledge nuance.  It’s not exactly Watch NASCAR and Crash.  But almost.  It’s pretty much the same thing.  Just some nuance.  But in the main, watch NASCAR and Die! Let’s go with that. Here’s the science.

Get a database from the State Police of all the aggressive driving accidents in West Virginia from 2003 to 2006.  Add in the dates on which a televised NASCAR event ran.  Now, correlate the number of aggressive driving accidents with days of the week, looking to find more accidents following NASCAR events.  And what do you discover?

Well, you don’t find any relationship between accidents and days after NASCAR events.  None.  The headline relationship (Watch NASCAR then Drive Aggressively) is not there.  That simple correlation does not exist although the headline says it does.  And the author carefully leaves that out.  Instead, he goes for nuance.

See, you need to adjust yourself and your data.  In public, of course, the way polite people do.  Then you can say with a straight face that watching NASCAR causes accidents.  Here’s how that goes.

First, covary out the effect of many variables already proven to causes accidents.  Bad weather.  Night time.  Season.  This has one great benefit.  All that explained variance attributed to factors that have nothing to do with NASCAR, seriously reduces the error term.  You begin with 100% error variance, then you pull out known variance (weather, lighting, season), and now you have, say, 70% error variance.  When you start testing the NASCAR effect, it’s against that much smaller error term which makes even little effects seem much bigger.  Without this adjustment, NASCAR viewing by its little, loud self apparently has no effect.

Second, look for the NASCAR effect on the day of the event and each of the five days following the event.  Now, you don’t have to use any science to explain why you look on six days and not, say ten or thirty or just one.  And, you don’t even have to guess whether the effect will happen on Day 1 or 2 or 3 or 4 or 5 (and if it doesn’t happen on any one of those Days, then we can always create different combinations of those 6 Days).  What looks like precision and detail and nuance is actually ignorance.  We have no idea when or how viewing NASCAR on TV causes accidents, we just know it does.  Sometime.

Now, what does this get you?

At Step 1, the control variables significantly predicted aggressive-driving accidents (R = .60, R2 = .34), F(14, 1432) = 51.62, p < .001. The addition of the NASCAR events at Step 2 resulted in a significant increase in the predicted variance of accidents (R2change = .06), Fchange(7, 1425) = 20.72, p < .001.  The entire equation of both control and NASCAR variables predicting aggressive-driving accidents was significant, F(21, 1425) = 44.63, p < .001.

What this means is that all those nonNASCAR variables have a very Large Windowpane effect (20/80) with that R2 value of 34%.  That Large effect massively reduces the error term in the statistical testing so that now at Step 2 when we test all our sniper shots (Day 1, 2, 3, etc.), and obtain a Small Windowpane effect (40/60) on highly partialled correlations, it is statistically significant.

The researcher also inspects each sniper shot and finds statistically significant effects for . . . what do you think?  Hey, you’re smart enough to read this, so you’re smart enough to think about it.  We code 6 days (day of, plus 5 following) and investigate each day’s accident rate.  What’s the outcome?

All of them?  Yeah, but in a descending linear decline with stronger effects Day Of and weakest on the last day.  That’d be pretty good theory, at least with that epi dose-response mantra.

No, get clever here.  All those NASCAR crazies know about the risk of watching NASCAR then driving, so they are super careful the Day Of, and the effect actually starts the next day and we get that linear decline.  Yeah.  That’s it.

No, the effect is alternating with every other day showing the effect.  No, the effect alternates with whether the date is even or odd numbered.  No, the effect is . . .

You won’t guess the results because they are impossible.  They are the one iron shot in golf.  Not even God hits the one iron, as the legend goes, and these results are a one iron shot.  No one would predict the particular pattern of results and there’s no psychological theory of persuasion that would predict this finding.

You do get a NASCAR effect, but it varies with whether you look at accidents or injuries or whether alcohol was involved.  Sometimes it happens on the Day Of, sometimes at Day 5.  Nuance.

This is scientific science.  Shrink the error term at Step 1 by running everything that actually does influence accidents.  Then, with a smaller and more “sensitive” error term, run your 6 sniper shot as one large bet and voila, as a group they come back significant, and you can even find two that hold all the juice in this bad apple (Day Of and Last Day.  Of course.  You knew that.  Everyone knew that.  Hand me the one iron, willya?)

And, of course, there is more nuance.  You get a different pattern of results depending upon whether you look at just accidents or just injuries, for example.  And, there’s no NASCAR effect if the accident involved alcohol.  Exactly what theory would predict . . . if you could hit the one iron.  Stated more obviously, the results are all over the place, based on highly partialled correlations, and without any theoretical precision.  Yet another Inconvenient Truth, you might say.

Thus armed we can go on an extended reflection of the known and now expanded list of perils on the Media Violence Chorus.  Real media violence kills and really kills real people in real time.  Now, the fact that more people are more exposed to more mediated violence in more forms than any time in human history and especially compared to the notorious and dangerous 1960s, then how come the murder rate is at historically low levels?  All forms of violent crime have collapsed in the past 30 years despite all that mediated violence (and sex, too).  But you can easily find scientists who will tell you media violence kills with data like this.

Observational research can mislead one into thinking you know more than you do.  While I often criticize epidemiology and environmental research, we see the same bad science here with media effects.  The headline claim that X Causes Death is never quite true.  It is X (plus a lot of Y, Z, A, B, C, and D I don’t headline) Correlates with Death in this Database I Assembled.  The proof always requires a variable pattern of adjustment with no consistency across studies; hey, baby, whatever gets the p < .05 ssd and it’s all good.

All Bad Science Is Persuasive.

Guy D. Vitaglione.  (2011).  Driving Under the Influence (of Mass Media): A Four-Year Examination of NASCAR and West Virginia Aggressive-Driving Accidents and Injuries. Journal of Applied Social Psychology.

Article first published online: 18 JUL 2011

DOI: 10.1111/j.1559-1816.2011.00783.x

The esteemed British Medical Journal publishes a meta analysis on the effect of chocolate consumption on mortality with a particular eye to stroke, heart disease, and heart attacks.  This Table tells it all and rather nicely.  Click to enlarge.

As the authors put it,

On pooling the retrieved measures of association, we found that high chocolate consumption was associated with about a third decrease in the risk of cardiometabolic disorders—37% in the case of any cardiovascular disease (relative risk 0.63 (95% confidence interval 0.44 to 0.90)) and 29% in the case of stroke prevention (0.71 (0.52 to 0.98)) (see fig 2). No significant association was observed in relation to heart failure (relative risk 0.95 (0.61 to 1.48)).

How about that?  Chocolate is good for you, particularly High Consumption.  That means eat more chocolate rather than eat less chocolate.

Of course, you could also reduce your risk of death and gain more life if you:

Run:  RR = .75

Walk:  RR = .81

Pray:  RR = .82

Or you could engage in High Consumption of chocolate and earn,

Eat:  RR = .63

Of course, all of these findings are based upon Observational Research with Huge Samples and Highly Significant Statistical Significance with, albeit, Small Windowpanes.  And they are published in prestige peer review sources.

Pray and pass the bon-bons or walk and run.  They have the same effect.

What’s your preference?

Persuasion mavens, gather around the body and feast!  If you cannot make money on Eat, Pray, Walk, and Run you didn’t know Jack Kennedy.

One of the prime conceits of human nature is the belief that you are a rational person who thinks about your choices, especially large ones, and tries within the limits of your knowledge to make a smart decision.  But, this is a conceit.  More people more of the time rely upon what other people are doing than relying upon their own rational judgment.  Consider this big example involving how people invest their money.

Stock-picking funds, of course, are still big money makers for fund firms. Actively managed U.S. stock mutual funds held about $2.75 trillion at the end of last year and charge average fees of 1.38%, while their index-tracking counterparts held just over $700 billion and charge 0.88% on average.

Thus, for every 1 dollar people put in an index fund, other people put 3 dollars in an actively managed fund.  That’s a large and obvious difference.  People prefer to invest in a fund that is run by a human rather than in a fund that is run by a human using a math model.

Of course, as the quote details, you pay for that human manager.  Those active funds charge about a half percent more for their service.  That doesn’t sound like much, but if this is your retirement fund, then every year you are giving up that half percent and over 30 or 40 years, that’s some serious money (the joys of compounding interest, right?).

But, the human managed fund does a lot better than one of those mathed up index funds, right?  Well, not exactly, as another quote from this article demonstrates.

In the five years ending in December, 62% of U.S. large-cap funds trailed the Standard & Poor’s 500-stock index. Results were even worse for funds focused on smaller stocks, with 63% of small-cap and 78% of midcap funds lagging behind their benchmarks, according to S&P.

Now, if humans were indeed better at stock picking, then you’d expect these results to be reversed, but virtually all the time, an index fund does better than the majority of actively managed funds, especially as you ride time’s arrow.  Ever since the 1970s when John Bogle first implemented an index fund at Vanguard, the investing world has seen this outcome.  Yet remember that first quote.  People will put 3 dollars into a worse performing active fund for every 1 dollar in a better performing index fund.

Isn’t this interesting?  We like to believe we are rational, but we prefer the reassurance of Other People to the reality of Index Investing.  I’ll take a leap here and connect this idea to Web 2.0 and all that social connectivity.  You can make money on the Other Guy’s need for people even when that need is irrational and actually costs more money.

No wonder the Cues work so well.

Rock ‘n roll music, as a creative expression of social change, is dead.  The End of History killed it even though Francis Fukuyama didn’t predict this and since he has also renounced his Great Hypothesis, he would also resist the inference.  He spoke truly first and he errs lately, so I persist on his past.  The End of History killed rock ‘n roll.

I point you to Simon Reynolds and his book, Retromania, a keen analysis of contemporary music and its fascination with all things past.  Reynolds properly notes that since the third millennium, all pop music only looks back.  Re-tro.  Re-issue.  Re-visit.  Re-vival.  Re-make.  Re-join.  Re-trospection.  As a creative force of change and expression, pop music no longer leads, it re-gurgitates.  It can’t even re-bel anymore.  Reynolds who presents himself as a hipster also reads the lit crit and employs Professor Bloom’s Anxiety of Influence as one explanation for the re-living the past.  But, consider even larger forces than the poetic agon.

I would argue that the death of dynamic pop music (and pop culture more generally) is a predictable outcome from Fukuyama’s analysis of human progress in his 1989 work, The End of History and the Last Man.  Fukuyama argues that people have determined the most effective institutions for organizing large national groups:  liberal democracy, free markets and capitalism, and the SET of science, education, and technology.  Everyone wants this trilogy, if in their own accent, dialect, and tradition, and finds no appeal in other combinations of institutions – monarchy and genetic engineering, anyone?  We are arriving at the End of History, the Big History as in the conflict of dictatorships versus democracies.  We all want the same Big Things and we will find each other in the same place:  Overland Park, Kansas with good schools, safe streets, friendly neighbors, no crime, no pollution, and easy access to everything.

Fukuyama employed his analysis to predict the imminent collapse of the Soviet Union with the uncanny accuracy of the Great Carnack come to life.  And no ironic Greek oracles with mixed messages.  He called it with footnotes!  Since then Fukuyama has renounced his Great Hypothesis more on the threat of science than the lessons of history.  His expertise leads him to question his prior expertise which is pretty much the way of expertise.  Achieving the Correct Answer is the death of analysis and if you value your thinking more than your thought, you will always question everything, including your own Correct Answers.

I am content with some kind of Truth even with contingencies and find the End of History most useful for seeing the future and explaining the past.  Even with the End, life and death still rolls on, but the Great Clash of Progress has changed scale.  We no longer pursue Existential Answers on the battlefield involving millions of combatants; the World War series ends at II and will not return like the latest Roman Numeralled Super Bowl.  We fight over Tastes Great versus Less Filling, CenterLeft versus CenterRight,  Some Belief versus More Belief.  The center has held and Mr. Yeats was only right for his time.

But, the price you pay for peace and prosperity is bad pop music and worse pop culture.

P.S. The only hope for Real Change is from cycles.  Ah, Vico.  Argh, Nietzsche.

P.P.S.  Michael Azerrad offers a review of Retromania.  Even if you aren’t interested in the book, read the review.  Azerrad balances the objective (the book and author) and the subjective (Azerrad) on a razor’s edge.  He shows how to write a great book review.