June 9, 2016

The massive error in the much publicized "Correlation not causation: the relationship between personality traits and political ideologies."

That article in American Journal of Political Science that everyone was talking about — because it made conservatives seem mentally disturbed — had to run a correction admitting they'd botched the data and what they'd presented as correlating with conservatism actually correlated to liberalism. Power Line has the details and ends with this paragraph:
In other words, if this study hadn’t come out conforming to the liberal narrative and sliming conservatives, it wouldn’t have attracted much notice. By the way, your tax dollars paid for this essential social science research. A note at the end says, “The data for this article were collected with the financial support of the National Institute of Health.” And people wonder why Republicans in Congress want to cut off federal funding for social science research. As an alternative, I suggest redirecting federal social science funds to Retraction Watch.

52 comments:

PB said...

Maybe this whole thing was intentional. They knew it would be vilified if the correct conclusions were published, so they flipped the results and liberals grinned while they read the article and reviewed the data. Now with the corrections made, liberals can't go negative on the story without acknowledging they the data supports the correct conclusions.

jaydub said...

Hmm. Wonder what the odds are that antropogenic global warming is also a math error?

Patrick said...

An effing joke, either way.

John said...

Was it peer reviewed?

Too many people fetishize peer review when it seems about asas likly to result in bs and retraction as non peer reviewed research papers

traditionalguy said...

Glad you posted this. That was worth the reading, twice.

Where category confusion reigns it is seldom possible to get such a straight retraction from the "Scientists" that published it. They usually resort to the Hillary Gang's Modified Limited Hang Out.

Psycho this and Psycho that, or Narcissist, are the most misused categories of instant ersatz Social analysis that are profoundly hurled at all Republican Presidents.

tim in vermont said...

One of my favorites in this genre, that purported to show that conservatives were stupid, used for the conservatives, self-identified conservatives who came from a group of recent immigrants attending community college. I am not sure that recent immigrants would have all that strong of a grasp of American conservatism, and community college is open admission. I think community colleges are great, I mean where would Keith Olberman be without his Kow Kollege? But still...

CWJ said...

Three authors of the original paper. Three!

n.n said...

The value of any philosophy is correlated with its principles, or lack thereof. Ideally, the principles will be internally, externally, and mutually consistent as we process a reconciliation of moral and natural imperatives.

tim in vermont said...

You can't imagine how shocked I am... Seriously... This is my shocked face: :^O

n.n said...

The ball of yarns is unwinding, one thread at a time.

rhhardin said...

It's probably a math error. The result is nonsense either way. Statstics doesn't work that way.

Earnest Prole said...

Oopsie-Poopsie

tim in vermont said...

Kind of like Mann's Hokey Stick:

For critics visiting this site [Climateaudit.org], there isn’t a shred of doubt that Mann et al 2008 used these proxies upside down from the Tiljander interpretation. See the original post here for details. My interpretation was triple checked by two Finnish statistics professionals (Jean S and UC) who are intimately familiar with Mannian methods and who confirmed using Mann’s Matlab code that the Tiljander series are used upside down in both the CPS and EIV versions of Mann et al 2008.

When somebody desperately wants a certain conclusion, they probably don't even know they are doing it. You know, like O.J. believes he is innocent.

mtrobertslaw said...

The recent riots by the progressive left provide empirical evidence for the corrected findings.

Michael K said...

This reminds me of a scene in a comedy where the professor had written an enormous formula on the blackboard.

Then, he stood back and looked at it for a minute.

Then he went back to the blackboard and changed all the pluses to minuses and the minuses to pluses.

HoodlumDoodlum said...

"The data for this article were collected with the financial support of the National Institute of Health.”

Tell me again what a scandal it is there isn't enough money to protect Americans from the zika virus.

Fernandinande said...

This sounds like one of those papers that wouldn't have gotten published if it had had the corrected results to begin with.

This guy writes fairly often on authoritarian lefties.

Rick said...

PB said...
Maybe this whole thing was intentional.


I doubt they set out to commit fraud. But they investigated only connections which were interesting to them, which effectively meant only connections which supported the left wing narrative. And the control parameters as low as possible because only interesting results make your name. When you add they were incompetent and the system isn't interested in catching errors which confirm the narrative you have these results.

The system works as propaganda even if each link is less so.

mockturtle said...

There are too many instances nowadays where correlative results are inferred by the media and the public to be cause/affect and too many 'researchers' who don't make it clear that a correlation is associative and what that implies.

Static Ping said...

I think this explains why my opinion of science and scientists keeps dropping. Pretty much the only thing I pay attention to is things with practical physical application, as in the scientists can show me it does something. How to destroy a field's reputation: too much sloppy work, too much trying to get attention and grants, too many liars.

Bruce Hayden said...

Tell me again what a scandal it is there isn't enough money to protect Americans from the zika virus

Surprise, surprise, the Administraion cannot document where the money went that had previously been allocated for this, but are asking for more. It does appear that some of it was spent on fighting anthropogenic global warming/climate change/etc, apparently because the Republicans in Congress had somewhat shut down that gravy train. So, no surprise that we hear constant screams about Zika, without mentioning why there is no money for it.

Char Char Binks said...

Is it settled science yet?

eddie willers said...

Maybe this whole thing was intentional.

You bet your bippy.

Nonapod said...

Confirmation bias, data massaging, misrepresentations of facts, and outright fraud seem to be rampant in certain fields of science. It turns out humans are pretty poor at being truly objective and empirical in certain situations. Shocking I know. It's one of my personal frustrations as someone who frequents many science and tech sites where there are often painfully obvious biases at work in articles on or related to things like climate science, social sciences, and anything related to quantifying intelligence in humans.

Big Mike said...

When I first read about the study my intuition was that it seemed to be the opposite of what I observed in my daily life. Glad to see my intuition confirmed.

When Richard Feynman coined the term "cargo cult science" in his 1974 Cal Tech commencement address he applied it to the social sciences (so called sciences) as well as things like ESP and UFOs. If you read the speech, you'll see why. And this sorry episode tends to confirm Feynman's position.

Paul Snively said...

Yet another reminder that people who are oh-so-concerned about religious dogma need to include scientism in their cross-hairs. And can we finally admit that people respond to incentives, and being publicly funded is a huge incentive? Like other posters, I'm not saying "fraud;" I'm saying people receiving a paycheck have a very human, even subconscious, tendency to see what their paymasters want them to see.

And yes, if it's actual climate statistics, from professional statisticians (as opposed to "climate scientists," who are not statisticians) you want, ClimateAudit is indeed where you need to go.

MadisonMan said...

We apologize for any inconvenience caused
by our error.

Last line of the Correction.

I wonder why the paper isn't withdrawn by the Journal.

Widmerpool said...

This Andrew Ferguson article is well worth reading on this:

Making it all up

chuck said...

You need a 'science bullshit' tag.

mockturtle said...

@ Widmerpool

This Andrew Ferguson article is well worth reading on this:

Making it all up


Good article but it's not only behavioral science that has fallen prey to this but physical sciences, as well.

BTW, I still have an old copy of The Journal of Irreproducible Results, a very funny satire magazine. [But not as funny as it used to be....]

Ann Althouse said...

I have "bad science." Will add that tag.

Drago said...

mtrobertslaw: "The recent riots by the progressive left provide empirical evidence for the corrected findings."

Your comment is not precise enough, given the riot-swath the left cuts across the globe.

How about - The most recent riots by the progressive left in San Jose where leftists are visiting violence against Trump rally supporters, approvingly by lefty California state governmental officials and most lefty pundits/commentators, provide empirical evidence for the corrected findings.

That way we could distinguish it from all the other lefty "mostly non-violenty lefty riots", like those in Venezuela where the enlightened lefties are simply shooting up opposition rallies. Probably with guns supplied by Obama.

Drago said...

It won't matter in the least that the conclusions of this study have been reversed.

Utterly irrelevant. The left got their original headlines and the rest goes down the memory hole.

Years from now leftists will be referencing the original announcement as "truth" and anyone who doesn't believe that original headline will be a "science-y denier".

Guaranteed.

"I can see Russia from my house!"

Peter said...

If the results are just flipped, wouldn't it make more sense to correct the study than to retract it? Or is it just that the flipped conclusion is so self-evidently absurd as to invalidate itself?

mockturtle said...

Bad environmental 'science' is some of the most damaging. Perhaps some of you remember the spotted owl uproar. Many loggers in the Pacific Northwest were put out of work to protect the spotted owl, whose decreased numbers were attributed to logging [Of course! Blame humans!], only to find out that it was the barred owl, another species, that was displacing--and killing--the spotted owls.

tim in vermont said...

Then he went back to the blackboard and changed all the pluses to minuses and the minuses to pluses.

Well, a single sign change could do that, but your story reminded me of a story of Mozart's pet starling. Supposedly he played a little tune for it on his piano, and the bird sang it back, but wherever Mozart sharped a note from the natural scale, the bird flatted it.

tim in vermont said...

only to find out that it was the barred owl, another species, that was displacing--and killing--the spotted owls.

So it was GLOBAL WARMING!

Drago said...

Peter: "If the results are just flipped, wouldn't it make more sense to correct the study than to retract it?"

The larger point is that all these "studies" are BS on stilts, and always have been.

It's just that the lefts long march thru the institutions and subsequent large scale blacklisting of conservatives/libertarians (the "Alger Hiss is innocent!" crowd will be along shortly to decry that claim) has put the lefties in a position to systematically abuse the funding/findings processes for all this politically driven "science".

Which is hardly surprising given the 100% occurrence rate of abuse whenever the left captures an institution.

The soviets "psychology" elites told us that anyone who desired freedom in the west was psychologically impaired.

Our "elites" tell us that anyone who falls into the conservative camp is psychologically impaired.

The evidence is manifest and thus, it will have to be denied even more vigorously!

Henry said...

The true finding of this study is found in this paragraph by its authors: In line with our expectations, P [for “Psychoticism”] (positively related to tough-mindedness and authoritarianism) is associated with social conservatism and conservative military attitudes.

Headline: Study Reveals that Social Scientists have Biased Expectations.

Hopefully Jonathan Haidt will pick up on this.

Harold said...

Speaking of birds, we always hear about birds needing those old old growth forest to thrive in; modern woodsy areas and clear cut areas just don't cut it for bird habitat.

We've recently had some really major power pole replacements done in our area, with newer, much wider and taller power poles installed. And the wires all moved from the old poles to the new. On top of one of the brand new not 8 week since installed poles just down the street from us sits an eagle's nest. High above the surrounding newly cleared utility right of way. Where the adult eagles have a birds eye view of prey moving on the brush free ground, enabling them to swoop down and feed. Feral cat population has been going down...

Those stupid birds just don't seem to understand that environmentalists know the birds are better off in old growth trees, and they keep building their nests on power poles with a clear view of the surrounding area. Stupid birds. They need to get an education so they know what's good for them.

wildswan said...

Behavioral genetics has had a terrible reputation since it was the source the false assertion that blacks have a genetic IQ deficit - as set forth in The Bell Curve by Charles Murray. So one would hope that our first black president would not make assertions from this tainted field a source of government policy. But he has done that. In September 2015 Obama isssued an executive order that departments and agencies were to identify policies and programs in which behavioral science insights could improve program outcomes - including, for example, "accelerating the transition to a low carbon economy." https://www.whitehouse.gov/the-press-office/2015/09/15/executive-order-using-behavioral-science-insights-better-serve-american
Where might the government get these "behavioral science insights" if not from its own NIH sponsored behavioral science programs. And we see just how good this government sponsored research is when we read about this piece of NIH sponsored behavioral genetic research. But the research is popular with the new eugenicists. This research was featured in Biodemography and Social Biology, volume 57, 1 in 2011. Biodemography and Social Biology is the journal of the American eugenic society, now known as The Society of Biodemography and Social Biology. It's been a while since eugenics was government policy but it didn't work before and I don't think we should try it again.

Oso Negro said...

Umm, Wildswan, it is NOT a false assertion. There is a gap of about a standard deviation that you can't wave away with jazz hands, safe spaces, or participation trophies. But let's suppose it is false. I propose that the rules of basketball be changed so that instead of shooting free throws, the player does calculus problems for points. Let's see how that works.

Bruce Hayden said...

@wildswan -,are you claiming that Blacks don't test, on average, maybe one std deviation below whites and orientals in IQ tests, as was claimed in The Bell Curve? If so, I would appreciate a link to your sources.

ceowens said...

This is fun.

http://tylervigen.com/spurious-correlations

Saint Croix said...

are you claiming that Blacks don't test, on average, maybe one std deviation below whites and orientals in IQ tests, as was claimed in The Bell Curve?

My claim is that pigment is irrelevant. Ergo we should not be dividing our society up in this way.

If groups of people are testing lower on IQ tests, there is very likely a cultural reason for that. For instance, parents reading to their children, versus parents who let their kids watch TV instead.

The dumb thing is to think this stuff is locked in. And the really dumb thing is the idea that you can reason from statistics to judge an individual you do not know. As Mark Twain said, "There are lies, damn lies, and statistics."

Saint Croix said...

As more and more black people start voting Republican, we will finally stop with this racialism. I hope!

Here is Ice T, defending the 2nd Amendment. I wonder where he got that idea?

tim in vermont said...

Those stupid birds just don't seem to understand that environmentalists know the birds are better off in old growth trees,

Cowbirds are the American version of the Cuckoo, and they lay their eggs in the nests of other songbirds, putting great pressure on these birds. Cowbirds like to be near open fields and it takes a pretty large expanse of woodlands to allow songbirds to nest out their reach. Some songbirds have evolved an ability to kick the cowbirds eggs out of their nests, but many have not. It has only been a few hundred years since the seemingly endless expanse of woodlands in the US has been chopped into much smaller lots.

damikesc said...

By the way, your tax dollars paid for this essential social science research.

I love paying for terrible "research" that insults me with no scientific basis. Really. I do.

Too many people fetishize peer review when it seems about asas likly to result in bs and retraction as non peer reviewed research papers

Indeed. We saw, with the climate email scandal, that they "game" the peer review system. We then learn that peer reviewed research cannot be replicated.

Makes one wonder what the benefit is, since "scientists" are no better at bucking popular trends than politicians.

Surprise, surprise, the Administraion cannot document where the money went that had previously been allocated for this

And the press, the VITAL part of maintaining our freedom, seems uninterested in learning. We need to give the media special freedoms...why?

It does appear that some of it was spent on fighting anthropogenic global warming/climate change/etc, apparently because the Republicans in Congress had somewhat shut down that gravy train. So, no surprise that we hear constant screams about Zika, without mentioning why there is no money for it.

This makes me wonder why, if a Republican President dislikes what an agency is doing, they simply don't give them the money allocated for their actions.

damikesc said...

Anybody want to place odds that the delta smelt issue that has basically killed off tons of California farms was utter and unmitigated BS? All they have to do is ruin people's lives for a few years, making them leave (before they starve to death) then they've won, no matter what the "research" says.

Fernandinande said...

wildswan said...
Behavioral genetics has had a terrible reputation since it was the source the false assertion that blacks have a genetic IQ deficit - as set forth in The Bell Curve by Charles Murray.


They merely speculated about genetics. Later research confirmed the overwhelming importance of genetics.

So one would hope that our first black president would not make assertions from this tainted field a source of government policy

"Tainted" meaning "verified over and over again."

Fernandinande said...

Big Mike said...
If you read the speech, you'll see why. And this sorry episode tends to confirm Feynman's position.


Not really. These guys just made a rather spectacular and amusing error: now it is corrected. (Though there may be other problems with this paper).

But there is a lot of terrible psych and sociology research, mostly because they assume a "blank slate" and ignore genetics, but they're not nearly as bad as the psych crap from the Freud-to-1960's era, e.g. "cold mothers cause autism", which was based on the theory of "sounds good, must be true" rather than any real research. And even that was better than the previous "possessed by demons" theory.

cubanbob said...


"If groups of people are testing lower on IQ tests, there is very likely a cultural reason for that. For instance, parents reading to their children, versus parents who let their kids watch TV instead."

Or culture follows IQ.