Skip

Maybe there isn't a "positivity ratio" after all.
October 17, 2013 8:38 AM   Subscribe

Positive psychology superstars Barbara Fredrickson and Marcial Losada had put forward a theory, seemingly with experimental confirmation, that was bolder than bold: that mankind, whether working alone or in groups, is governed by a mathematical tipping point, one specified by a ratio of 2.9013 positive to 1 negative emotions. When the tipping point is crested, a kind of positive emotional chaos ensues—“that flapping of the butterfly’s wing,” as Fredrickson puts it—resulting in human “flourishing.” When it is not met (or if a limit of 11.6346 positive emotions is exceeded, as there is a limit to positivity), everything comes grinding to a halt, or locks into stereotyped patterns like water freezing into ice. Nick Brown smelled bull.
posted by shivohum (68 comments total) 19 users marked this as a favorite

 
Not technically a double, but we have talked about Brown and Fredrickson before.
posted by googly at 8:47 AM on October 17, 2013


Senator Franken's take on daily affirmation...
posted by The 10th Regiment of Foot at 8:51 AM on October 17, 2013 [1 favorite]


I'm still reading the linked article, but here's the journal article the team produced.
The complex dynamics of wishful thinking: The critical positivity ratio
Nicholas J.L. Brown, Alan D. Sokal, Harris L. Friedman
(Submitted on 26 Jul 2013)

We examine critically the claims made by Fredrickson and Losada (2005) concerning the construct known as the "positivity ratio". We find no theoretical or empirical justification for the use of differential equations drawn from fluid dynamics, a subfield of physics, to describe changes in human emotions over time; furthermore, we demonstrate that the purported application of these equations contains numerous fundamental conceptual and mathematical errors. The lack of relevance of these equations and their incorrect application lead us to conclude that Fredrickson and Losada's claim to have demonstrated the existence of a critical minimum positivity ratio of 2.9013 is entirely unfounded. More generally, we urge future researchers to exercise caution in the use of advanced mathematical tools such as nonlinear dynamics and in particular to verify that the elementary conditions for their valid application have been met.
(Direct link to PDF.)
posted by maudlin at 8:55 AM on October 17, 2013 [3 favorites]


Oh, that's gonna leave a mark:
In critiquing a section of Losada’s earlier research that characterized high performance teams as “buoyant” and low performance teams as “stuck in a viscous atmosphere,” to take an example, Brown wrote mockingly:

"One could describe a team’s interactions as 'sparky' and confidently predict that their emotions would be subject to the same laws that govern the dielectric breakdown of air under the influence of an electric field. Alternatively, the interactions of a team of researchers whose journal articles are characterized by 'smoke and mirrors' could be modeled using the physics of airborne particulate combustion residues, combined in some way with classical optics."
posted by maudlin at 8:56 AM on October 17, 2013 [20 favorites]


I'm torn about this because clearly this is absolutely 100% BS and that seems pretty obvious. On the other hand, I am deeply, deeply sick of the idea of the "plucky amateur" who beats the professionals at their own game. Professionals are often professionals for a reason, because they have experience and training, and I feel like at least in the US we fetishize the Bold Outsider who comes in to Shake Things Up in the stodgy world of whatever; business, education, science, government, all kinds of things, and it's really, really frustrating for people with a serious stake and experience and knowledge.

Also, the fact that something completely contradicts common sense but has been experimentally proven means we maybe need more science, not more common sense.

So as I say I'm torn; charlatans are the worst and preying on people's frustrations or fear or unhappiness is really really awful and I'm glad someone looked into this, but I just don't like the knee-jerk support of the pro-common sense guy against science.
posted by Mrs. Pterodactyl at 9:03 AM on October 17, 2013 [9 favorites]


“It may feel like we’re taking this dead horse, flogging it, quartering it, baking it out in the sun,” Brown says. “A lot of people are saying ‘Stop already!’ And we’re going, ‘We can’t!’ When you’re dealing with people like this—particularly with people like Losada, although I must say I was disappointed by Fredrickson’s reaction too—you have to kill every point stone dead.”
Couldn't agree more with this. It only takes a couple of charlatans spouting bullshit being backed up by otherwise honest people who are afraid to admit they don't understand to set an entire field back decades.

I'd want the both of them drummed out of the profession.
posted by Mooski at 9:04 AM on October 17, 2013 [16 favorites]


FTFA

In a January 2011 special edition of the American Psychologist that was dedicated to the program, Seligman, who guest-edited the edition, wrote along with two military personnel that CSF's goal is to, “increase the number of soldiers who derive meaning and personal growth from their combat experience,” and "to decrease the number of soldiers who develop stress pathologies."

Recently, it has been reported that CSF has done little to reduce PTSD. Nevertheless, the government is expanding the $50-million-per-year program.


Um, the goal of the program wasn't to "reduce PTSD" iit was to “increase the number of soldiers who derive meaning and personal growth from their combat experience,”. I wish the author would comment on whether the program's stated goal was effective.
posted by MisantropicPainforest at 9:08 AM on October 17, 2013 [3 favorites]


I am deeply, deeply sick of the idea of the "plucky amateur" who beats the professionals at their own game.

Mrs. P, I hate that narrative, too, but this plucky amateur didn't fall victim to the Dunning-Kruger effect. He knew his limitations and he was smart enough to get two people with more expertise to team up with him. He didn't have the deep knowledge of math required to analyze the original claims, but then again, neither did the original researchers.

(I have a B.Sc. in Psychology from the University of Toronto, but the only math I was required to take was a single semester on basic stats. It really should have been a B.A.)
posted by maudlin at 9:11 AM on October 17, 2013 [14 favorites]


The theory was well credentialed. Now cited in academic journals over 350 times, it was first put forth in a 2005 paper by Barbara Fredrickson, a luminary of the positive psychology movement, and Marcial Losada, a Chilean management consultant, and published in the American Psychologist, the flagship peer-reviewed journal of the largest organization of psychologists in the U.S. - from the narratively link.

The American Psychologist is the flagship journal of the APA but don't mistake it for being a top research publication. The APA's orientation toward therapy and clinicians and weak support of research is the reason why there are so many other splinter organizations composed of research psychologists. They have distanced themselves from the clinical side of psychology.

A lot of stuff in psychology that gets into the popular press is treated as click-bait puff pieces by many serious researchers. They roll their eyes and get back to doing their work. The problem is that the failure to adequately police the fringes of the field often discredits the center.

I'd also like to point out that this is actually pretty much science in action.
posted by srboisvert at 9:13 AM on October 17, 2013


Never mind being sophisticated enough to understand differential equations. Claiming that 2.9013 is the critical number, as opposed to 2.9012 or 2.9014, should have set off every bullshit detector in the inner solar system.
posted by skyscraper at 9:13 AM on October 17, 2013 [27 favorites]


The idea of applying mathematics to psychology is so ridiculous in and of itself that I can't understand why anyone ever took it seriously.

Um, the goal of the program wasn't to "reduce PTSD" iit was to “increase the number of soldiers who derive meaning and personal growth from their combat experience,”. I wish the author would comment on whether the program's stated goal was effective.

The whole program is based on absolute bullshit. Might as well audit the thetans out of them.
posted by Sys Rq at 9:15 AM on October 17, 2013 [8 favorites]


Um, the goal of the program wasn't to "reduce PTSD" iit was to “increase the number of soldiers who derive meaning and personal growth from their combat experience,”. I wish the author would comment on whether the program's stated goal was effective.

Yeah, but what does that mean and how is it measured? To be honest, I don't really want to increase the number of soldiers who derive meaning and personal growth from their combat experience, but I do want people not to have PTSD. Even beyond that, "reduce PTSD" is a pretty definable goal; there are diagnostic criteria and statistics. The stated goal is handwavy BS; what, are you going to pass out a questionnaire asking "do you derive meaning and personal growth from your combat experience?" before and after and then compare results? The stress pathologies thing is legit but the personal growth is nonsense unless you make it measurable in some way.

Mrs. P, I hate that narrative, too, but this plucky amateur didn't fall victim to the Dunning-Kruger effect. He knew his limitations and he was smart enough to get two people with more expertise to team up with him.

On preview, yeah, this is actually a great demonstration of the RIGHT way to do this; he took the time to understand the field and gain expertise and get in touch with people who could help and it's actually pretty awesome and I'm super impressed with him. The actual story is great, it just drives me nuts that instead of celebrating this guy's attempts to become a professional and make things better that way the little epigram thing that says "A plucky amateur dared to question a celebrated psychological finding. He wound up blowing the whole theory wide open." Why does it have to be that? Why can't it be "A thoughtful and intelligent guy questioned a scientific finding and, through effort, smartness, and learning was able to team up with other thoughtful, intelligent people and prevent charlatans from being sketch"? I mean I know that's long and not punchy but I think it's actually way awesomer.
posted by Mrs. Pterodactyl at 9:17 AM on October 17, 2013 [11 favorites]


skyscraper: Never mind being sophisticated enough to understand differential equations. Claiming that 2.9013 is the critical number, as opposed to 2.9012 or 2.9014, should have set off every bullshit detector in the inner solar system.

No kidding, that's exactly what I noticed. You have five significant figures? Really? So your sample size, was it in the millions or tens of millions?
posted by Mitrovarr at 9:18 AM on October 17, 2013 [5 favorites]


The idea of applying mathematics to psychology is so ridiculous in and of itself that I can't understand why anyone ever took it seriously.

You probably can't understand because your ability went out the window with the baby and the bathwater.
posted by srboisvert at 9:19 AM on October 17, 2013 [3 favorites]


You probably can't understand because your ability went out the window with the baby and the bathwater.

Well, by all means, explain to me how it's the least bit relevant.
posted by Sys Rq at 9:20 AM on October 17, 2013 [1 favorite]


I think what srboisvert is saying is that you shouldn't dismiss using math to study pschyology because of this singular flawed study.
posted by MisantropicPainforest at 9:22 AM on October 17, 2013 [2 favorites]


Statistics are math, are they not? I mean, psych should be rigorous. It should not be based on "this sounds right", because people get hurt if it turns out the thing that sounded right wasn't, just as much as they do in medicine or, I don't know, engineering space shuttles. Many mathematical models for things are approximate. There is, however, a pretty massive difference between approximate and pulled wholly out of a completely different discipline just because you can draw some metaphorical connection between the two situations. That's not applying math to psychology; that's just straight BS.
posted by Sequence at 9:24 AM on October 17, 2013


One weird trick from a mom for disproving psychology. Click here to find out her secret. Psychologists hate her!
posted by idiopath at 9:24 AM on October 17, 2013 [5 favorites]


I mean I know that's long and not punchy but I think it's actually way awesomer.

I agree: there's the seed of a much better, pithier blurb there. Too bad that the site didn't try to build one.

On preview: Mathematical Psychology is a real field. Some pretty robust theories and models applicable to memory, learning and perception are listed here. But I think that the appropriation of mathematical models from other disicplines to explain personality, group and organizational behaviour is very risky, to say the least.
posted by maudlin at 9:26 AM on October 17, 2013 [1 favorite]


Yeah, statistics are math. And, yeah, obviously statistics are relevant to the measuring of effectiveness of treatment and whatnot.

But this is applying math (or, as it turns out, just a bunch of meaningless numbers) directly to psychology, as if math itself is a cure-all.
posted by Sys Rq at 9:28 AM on October 17, 2013


Yeah, and since they did a poor job that means that we should never ever ever apply math directly to psychology?
posted by MisantropicPainforest at 9:31 AM on October 17, 2013


A bit off-topic, but here's how the level of accuracy of 2.9013 to 1 (that is, five significant digits) would compare to other measurements. It's roughly equivalent to accurately finding the weight of the average person to within the weight of a paper clip, or to accurately measuring the height of the average person to within the width of a single human hair.

Such accuracy in measurements is, of course, meaningless when applied to the physical aspects of humans; just combing your hair would significantly change your height in hair-widths, and exhaling a full breath would reduce your weight by about half a paper-clip (according to my back-of-the-envelope calculations).

So why would anyone think that we could realistically find that kind of accuracy in measuring the social or emotional aspects of human beings?
posted by math at 9:33 AM on October 17, 2013 [13 favorites]


the goal of the program wasn't to "reduce PTSD"

What? According to your own quote, one of the goals was "to decrease the number of soldiers who develop stress pathologies."
posted by mbrubeck at 9:33 AM on October 17, 2013


Yeah, and since they did a poor job that means that we should never ever ever apply math directly to psychology?

Is there a way to do it that isn't utterly meaningless?
posted by Sys Rq at 9:33 AM on October 17, 2013


The 2.9013 ratio is correct, but it takes exactly 10,000 hours to achieve it.
posted by uosuaq at 9:36 AM on October 17, 2013 [8 favorites]


Call me crazy, but I thought psychology was supposed to be medicine and medicine was supposed to be based on "SCIENCE!".
posted by The 10th Regiment of Foot at 9:37 AM on October 17, 2013 [1 favorite]


Is there a way to do it that isn't utterly meaningless?

Sys Rq, look into fields like mathematical and computational cognitive science / computational brain modeling.
posted by maudlin at 9:39 AM on October 17, 2013 [3 favorites]


Is there a way to do it that isn't utterly meaningless?

I don't know, I'm not a psychologist.
posted by MisantropicPainforest at 9:39 AM on October 17, 2013


Sys Rq: But this is applying math (or, as it turns out, just a bunch of meaningless numbers) directly to psychology, as if math itself is a cure-all.

I think a better way to put what you are trying to say is "One should not apply inappropriate physics theorems to psychology". Applying statistics to psychology is absolutely appropriate and necessary if you want to approach it in a scientific fashion, and there are probably other fields of math which could contribute (equations from biology and chemistry when dealing with neurological interactions, game theory, genetic algorithms, etc.)

What you can't do is make an analogy to a physical principle, and then expect the physics equations that govern that principle to also govern your psychological process. I honestly can't believe that made it past review.
posted by Mitrovarr at 9:39 AM on October 17, 2013 [11 favorites]


That is an awful lot of significant digits.
posted by Bovine Love at 9:39 AM on October 17, 2013


What? According to your own quote, one of the goals was "to decrease the number of soldiers who develop stress pathologies."

Noted! I stand corrected.
posted by MisantropicPainforest at 9:39 AM on October 17, 2013


> A universal constant predicting success and fulfillment, failure and discontent? "In what world could this be true?" he wondered.

In a world where management manuals are the Bible and Jack Welch or Lee Iacocca are prophets.
posted by The Card Cheat at 9:40 AM on October 17, 2013 [1 favorite]


Just a friendly reminder that most published research findings are false.
posted by So You're Saying These Are Pants? at 9:43 AM on October 17, 2013 [2 favorites]


Man, this just reminds me of the work that was done to debunk all those management training seminars and all the material that keeps getting taught to MBA's on how to manage teams and how team building is essential to the success of a business and how if everyone is a proper team player, then the whole business will be a success. Oh, wait, no one bothered because it was obviously bullshit. Yet they still teach that shit today. They also use metrics that have less than zero to do with actual performance or success rates. Gee, I wonder why so many businesses seem so fucking demented in their corporate cultures.

Or, even more fun, when those same management seminars and training videos use the papers published by the APA that cite this now debunked paper, but no one bothers to look at those cites and whether they are linked to, you know, real science, and not some bullshit scam.

The fact that Losada is a fucking "management consultant" just makes me see red. I cannot stand these fucking clowns. The fact that someone can make a career around this kind of shit is a fucking travesty of our society, and I swear it is ruining more businesses and more lives than it helps. You know that funny saying from 'management consultants' about "happiness is about managing your expectations with reality" and other happy horseshit? Yeah, thanks, I have that daily affirmations calendar too. I prefer the original demotivationals though. At least then I can laugh instead of feel like someone is trying to pull one over on me.

Maybe I've just been exposed to way too many fucking horrible people who seem to make a living doing this kind of thing in the most cargo cultish manner you have ever seen. They don't even really understand the materials themselves, yet they are the experts on management. I have yet to see one that didn't have to deflect questions with "trust me". Maybe I'm just lucky or something.
posted by daq at 9:51 AM on October 17, 2013 [18 favorites]


I favorited math's comment because I couldn't come up with a concise way to point out that it is eponysterical.

And also very interesting.
posted by gauche at 9:52 AM on October 17, 2013 [1 favorite]


smelled bull.

Anything that measures happiness to four decimal places is bull.
posted by mark7570 at 10:05 AM on October 17, 2013


math: So why would anyone think that we could realistically find that kind of accuracy in measuring the social or emotional aspects of human beings?

Theoretically, it would be possible to produce a statistical measure of that precision on the population as a whole. You could do an average, for instance.

The problem is, to achieve that level of precision, you'd have to have two things: You'd have to have a property you could measure with extreme precision to cut down the measurement error, and you'd have to have a huge sample size (or extreme consistency between samples) to cut down on the statistical error. Which, for something like this, is impossible; you can't measure an effect like this with that kind of precision, you know it isn't going to be perfectly consistent between individuals, and if the sample size was large enough to reduce the error to five-significant-figures-low, we'd all know about it because the sample size would have had to have been so large we'd have all been part of the study group.

This is why you never really see numbers like this quoted outside of things like particle physics and chemistry. Particle physicists can do it because particle accelerators deliver extremely precise measurements with samples sizes in the billions (at least), and chemists can do it because the 10th time you measure the weight of a compound with your extremely precise scale, it's still going to be the same. But you'd never see this level of precision being claimed in biology, for instance (and if you do, you should look out!)
posted by Mitrovarr at 10:09 AM on October 17, 2013 [5 favorites]


Man, this just reminds me of the work that was done to debunk all those management training seminars and all the material that keeps getting taught to MBA's on how to manage teams and how team building is essential to the success of a business and how if everyone is a proper team player, then the whole business will be a success. Oh, wait, no one bothered because it was obviously bullshit. Yet they still teach that shit today. They also use metrics that have less than zero to do with actual performance or success rates. Gee, I wonder why so many businesses seem so fucking demented in their corporate cultures.

Or, even more fun, when those same management seminars and training videos use the papers published by the APA that cite this now debunked paper, but no one bothers to look at those cites and whether they are linked to, you know, real science, and not some bullshit scam.

The fact that Losada is a fucking "management consultant" just makes me see red. I cannot stand these fucking clowns. The fact that someone can make a career around this kind of shit is a fucking travesty of our society, and I swear it is ruining more businesses and more lives than it helps. You know that funny saying from 'management consultants' about "happiness is about managing your expectations with reality" and other happy horseshit? Yeah, thanks, I have that daily affirmations calendar too. I prefer the original demotivationals though. At least then I can laugh instead of feel like someone is trying to pull one over on me.

Maybe I've just been exposed to way too many fucking horrible people who seem to make a living doing this kind of thing in the most cargo cultish manner you have ever seen. They don't even really understand the materials themselves, yet they are the experts on management. I have yet to see one that didn't have to deflect questions with "trust me". Maybe I'm just lucky or something.
posted by daq at 9:51 AM on October 17
[3 favorites −] Favorite added! [Flagged]


daq: I wish I could favorite this a thousand times, and flag it 'fantastic' a thousand times!

That affirmation stuff brings out the worst in me as a human being. It brought out the worst in me as a worker.

Tell me what to do, no contests, no micromanagement, no public humiliation. No dress codes no drug tests no personality tests no spying...GTF outta my way and let me do my job. Stand back, aweseness might occur... Or not .. Either Way...
The job will get done...

Now pay me my f...ing $€﷼¢£₩¥ and let me go home.

Thank you! That's all the positive affirmation I ever needed. Only two bosses ever got that about anyone. God bless both of them!
posted by Katjusa Roquette at 10:12 AM on October 17, 2013 [7 favorites]


I took Intro to Psych in college many years ago, and the only thing I remember about it is the amount of time the professor spent demonstrating that Psychology is a Real Science. That made me smell bull right away.
posted by Pararrayos at 10:42 AM on October 17, 2013


I'd also like to point out that this is actually pretty much science in action.

Except for how, you know, Losada and Fredrickson are continuing along their decidedly comfy, rewarding career paths in "science" — and Brown isn't. There may be a useful symmetry between validation and invalidation of an idea, but there's no such symmetry in terms of the publication and career consequences. Saluting this as "science in action" doesn't fix the fact that glib bullshit often sells better than careful research — and hence someone pursuing an academic (or research, or management-consultancy) career founded on bullshit is likely to be more successful (or, in today's economic terms, has a better chance to have a career at all) compared to an ethical researcher.
posted by RogerB at 10:44 AM on October 17, 2013 [2 favorites]


Why would demonstrating the scientific nature of a science be cause for bullshit?
posted by MisantropicPainforest at 10:52 AM on October 17, 2013 [1 favorite]


Others said they knew about the 2005 paper and had cited it, but with qualifications. “My opinion of the paper has always been that it was a metaphor, disguised as modeling,” said David Pincus, a psychologist at Chapman University who specializes in the application of chaos theory to psychology.

Seriously, try that excuse in physics. Psychology must move beyond this sort of thing if it is ever to be taken seriously as a hard science, and the importance of psychology to human behavior and policy means we desperately need psychology to be a field we can trust.
posted by Mitrovarr at 10:58 AM on October 17, 2013


Why would demonstrating the scientific nature of a science be cause for bullshit?

It reeks of insecurity, and one then wonders why they would feel so insecure if it was, in fact, a hard science.

Not that 'hard' and 'soft' are particularly good descriptors for different disciplines.
posted by Slackermagee at 11:06 AM on October 17, 2013


That's a pretty poor argument though, maybe the professor wasn't insecure, but wanted to instill in students who had very little exposure to academic psychology that its not the same as pop psychology.
posted by MisantropicPainforest at 11:15 AM on October 17, 2013 [1 favorite]


It reeks of insecurity, and one then wonders why they would feel so insecure if it was, in fact, a hard science.

Not that 'hard' and 'soft' are particularly good descriptors for different disciplines.


Yes, the problem here isn't whether or not hard/soft is the right set of descrpitors, it's whether logical positivism is the right lens with which to view the totality of the world. There are a lot of things science can say about clinical psychology and how it works, but it's idiotic to pretend that human nature and interactions is best described by a digit with five (or any) significant digits. The insistence that it is (or should be), and the disdain for areas of life that are not quantifiable, is part of what leads to idiotic papers like this.
posted by OmieWise at 11:18 AM on October 17, 2013 [1 favorite]


Wouldn't that still be insecurity? The professor is afraid his/her, legitimate, field is being conflated with some currently popular theme in that case.

I don't think I've heard of an intro o-chem professor doing this, they just let the subject matter scare off anyone who was there after getting into a CSI (now Breaking Bad I suppose) kick.
posted by Slackermagee at 11:20 AM on October 17, 2013


No, that wouldn't be insecurity, that would be a desire to educate, because academic psychology is not the same as pop psychology.
posted by MisantropicPainforest at 11:37 AM on October 17, 2013


It's this is the "style" of research that gives sociology a bad name; the soft sciences are so important but there are incentives that cause supposed scientists to approach problems and questions in a medieval way. One unfortunate consequence is that the public, and especially some of those who are more inclined towards the hard sciences, grow unsupportive of and sometimes even look down upon these fields. It's not a very progressive dynamic, for any party involved.
posted by polymodus at 12:06 PM on October 17, 2013


I don't really understand how you can count emotions anyway. (Regret and frustration - two emotions or one? I'd say it depends what you're regretting and being frustrated about.) Or split them unambiguously into positive and negative. Positive in what respect? And are all positive emotions equally positive - wouldn't they be on a scale of positivity? Can't they also be felt with different intensity? Doesn't it matter which emotions? Aren't some emotions incompatible with each other? What would feeling twelve different emotions simultaneously feel like?

Naive questions, perhaps. I can only assume these were never meant to be emotions in anything like the everyday sense of the word.
posted by Segundus at 12:25 PM on October 17, 2013 [1 favorite]


Is there a way to do it that isn't utterly meaningless?

You have just overturned everything that the entire area of measurement in psychology has ever tried to do using your extremely valuable outsider perspective! You rascally outsider, you didn't even know the field you dismiss exists!
posted by srboisvert at 1:21 PM on October 17, 2013


Why would demonstrating the scientific nature of a science be cause for bullshit?

None of my physics professors ever felt the need to demonstrate the scientific nature of their subject.
posted by Pararrayos at 1:25 PM on October 17, 2013 [1 favorite]


“My opinion of the paper has always been that it was a metaphor, disguised as modeling”

Seriously, try that excuse in physics.


Does string theory count?
posted by ook at 1:39 PM on October 17, 2013


Yes, ook, it does. When was the last time you read a string theory paper? Ever count the equations?
posted by nat at 1:42 PM on October 17, 2013 [1 favorite]


You have just overturned everything that the entire area of measurement in psychology has ever tried to do using your extremely valuable outsider perspective! You rascally outsider, you didn't even know the field you dismiss exists!

Thank you for your detailed answer to my question.
posted by Sys Rq at 2:17 PM on October 17, 2013


None of my physics professors ever felt the need to demonstrate the scientific nature of their subject.

So? Is there a popular misconception that physics is not scientific?
posted by MisantropicPainforest at 2:35 PM on October 17, 2013


I take it this article has not been retracted?
posted by grobstein at 2:36 PM on October 17, 2013


I mean, this is not just a falsified result. This looks like a fraud, and a fraud that lead to a literature of hundreds of papers all built on sand.

Add this to the growing body of evidence that psychology needs to clean house.
posted by grobstein at 3:25 PM on October 17, 2013


So? Is there a popular misconception that physics is not scientific?

Is there a popular misconception that psychology is not scientific? If there is -- or was -- then as a 19 year old college sophomore I was entirely unaware of it. That is, until I took the Intro course and saw how much time they spent assuring me -- no really, we really mean it -- that this is a SCIENCE, with regressions, and p-values and everything.

That's about all I remember from Intro to Psych. Well that and the Milgram Experiment. In which, as I recently learned, Milgram's thumb was on the scale the whole time.

Science!
posted by Pararrayos at 3:28 PM on October 17, 2013


Science!

Psychology is a Soft Science kinda like Philosophy. Like Philosophy, Psychologists sometimes succumb to that odd Math Envy which sometimes leads them astray. Just don't say that: "Therefore, all Psychology is dumb", because human behavior is a complex thing which is always worth examining closely.
posted by ovvl at 3:48 PM on October 17, 2013


The problems in psychology are from the root of the fields inception. The purpose was to study the "abnormal" behaviors and to attempt to classify them in non-objective terms, while pretending that there was an objectively good or normal behavioral model. Instead of being able to form coherent thesis based upon empirically measurable control models versus applied variables, they attempt to form behavioral models based on individually subjective cultural norms, which may or may not be based in any direct reality.

I am not saying that clinical psychology doesn't try to do these things. But unlike, say chemistry or physics, which can be theoretically modeled that then those models can be applied and the results of the theories actually match the results of the applied experiments, you have no base line of what a "normal" human being is. From observer projection (Freudien) to assumed cultural stereotypes (Jungian), the "grandfathers" of modern psychology have more in common with philosophy than with actual science. This is why neuro-science is a completely different field of study and has garnered more results in it's short existence than more than a century of psychology.

Now, I'm going to completely reverse myself. The actual field of psychology will never have a base human mind model to work from. And almost all of the theories about human behavior are based upon observation and very distorted experimentation. From trepanation, to lobotomy, to psycho-pharmaceutical testing, the field almost entirely formulates it's theories of mind from subjective situational observation and abnormal experimentation.

Now, I won't really fault the originators. They did not have any way of knowing of how neurons interact and the ways in which sensory input and memory worked. They could not. It is like trying to stare at the pupil of your eye in a mirror. Because your vision works by having your eye move rapidly around, you will never actually be able to hold your eye still enough to get it to line up. So you never get to see the light that bounces off the back of your eye reflected in the mirror bouncing back into your eye (at least not at a frequency that will register with your rods and cones and transmit the information through your optic nerve to your neurons to be recorded as an actual image). Go try it. Tell me what color your retina is using just a mirror. So the originating models have always been flawed from the start of the attempts to categorize and modify behavior.

Sociology is only mildly better, in that at least it admits that there is a huge difference between it's models and the realistic outcomes.

The one thing that psychology has learned, though, is a lot of insight into the tricks and traps of how to manipulate perception. It is not so much that they can cure mental illness or actually treat neural disorders (though through a lot of trial and error they did manage to stumble on a few drugs that are at least somewhat effective at treating schizoid disorders, though more study of the neuroscience is leading to much more advanced and focused treatments, due to learning about actual structural differences in individual brains), but they have at least done a lot of work in cataloging behaviors and perceptual motivations. They also learned a whole bunch of ways to direct behavior, though in a strangely round about manner. The use of psychology in advertising, for example, is fascinating, if not scary. The use of statistics and focus group responses has been instrumental in creating the megalithic marketing machine of modern capitalism. Though, much like a gun, I would swear there should be more regulation, most notably in political or pharmaceutical advertising (though there is a lot of regulation there already, if very convoluted and strange).

What psychology really should focus on is finding it's own first principals. What is the control model of human behavior. Is this a subjective model based upon the culture in which this human is interacting, or is this the base of "man in nature", meaning a blank slate human, created in isolation and only interacting in nature (or in a primitive society, without observer intervention, which, to date, has not actually been done to my knowledge). To a greater degree, many of the failures to seek first principals have stemmed from the strong cultural belief that humans are any different than any other animals that exist on this planet. We are all results of our ancestry and our environment. It is the belief that humans are apart from nature that has led to so many quite kooky beliefs, such as "positivism" and "The Secret" or further afield into the madness of "ancient aliens" a la Chariot of the Gods type tripe. Or then we get to the really weird like the anti-psychology backlash that has given us Thetans and the belief in Angels.

Have we covered this enough? I mean, yes, the field does need to exist, and it does need to be studied and we have so much to learn about ourselves and how we work as human beings. But it is the ignorance or elimination of the other major factors that make up human beings that I think makes many people feel that the existing practices and/or attempts at proving legitimacy have utterly failed to convince us that there is anything really rigorous in the existing literature or field of study. It's like the argument that economics isn't a real science because of spherical cows. I mean, even allowing for simplified models, it just feels absurd to try and base any real world policy or application upon something that ignores something as fundamental as that.

Now, does anyone need a hug?
posted by daq at 4:19 PM on October 17, 2013


Psychology is a Soft Science kinda like Philosophy.

Professional philosophers do not believe that they are doing science when they do philosophy. There are no experiments in philosophy. There are no control groups.
posted by MisantropicPainforest at 5:02 PM on October 17, 2013


I do not understand why you'd need complicated math beyond ordinary statistics to test this. For instance, there are studies of language use that simply observe families and count positive and negative utterances expressed to children— and one of the most famous, by Hart and Risley, finds that poor kids hear both fewer words and more "discouragements" than rich kids do. Both the number of discouragements v. encouragements and the number of new words heard correlate with the child's later vocabulary and school performance. Anyway, you don't need a complicated equation to compare how many encouragements v. discouragements people hear to one another and then look to see how this relates to other outcomes.

So, why did they need one in the first place? I suspect that most of the people who cited the study carelessly figured that the researchers had done just what I've suggested above to get the ratio and didn't read the paper thoroughly before citing it. That's awful— but I imagine you could test this hypothesis quite easily and find out whether, outside the study I just cited, it still rings true. Hart and Risley's work, however, supports the idea, not exactly shocking, that children who are rewarded and spoken to more do better than those who are told no all the time and not given much attention.

The marriage counselor Gottman, who has had his own run ins with people saying his work doesn't hold up, also supposedly finds a similar phenomenon in marriages— again, with very simple math.

The answer here is replication and keeping it simple, I think.
posted by Maias at 5:10 PM on October 17, 2013 [2 favorites]


Yes, ook, it does. When was the last time you read a string theory paper? Ever count the equations?

You know the theory isn't literally about strings, right?
posted by ook at 5:47 PM on October 17, 2013


daq, I don't disagree with you completely. But the history of psychology as a field of study is marked by the question of 'what is psychology, what is it about?' The early psychotherapists gave way to the behaviourists, and from there, well, it all became a bit of a mess in terms of it being a strictly defined discipline. Depending on the particular focus of a particular psychologist they could easily be at home in academic departments from sociology, biology, neurology, philosophy, business/marketing, cognitive science, politics, and for the particularly psychometrically inclined maths. Possibly this is the problem the discipline faces - it is addressing an absurdly vast question, which breaks down into a number of also vast questions. The institution where I studied psych for a couple of years before changing streams paired its psych department with its aviation department, as they both had a research focus on human factors (which in aviation terms is a euphemism for 'pilot error and how to reduce it so they don't kill everyone on board').

In any case, research ethics basically precludes study of first principles, as you'd have to do some pretty fucked up shit to work such things out. I still feel sorry for the poor Rhesus monkeys that brought us Attachment Theory.
posted by Hello, I'm David McGahan at 6:45 PM on October 17, 2013


ook: "You know [string] theory isn't literally about strings, right?"

What? String theory is quite literally about strings. "String" isn't a metaphor anymore than "particle" or "wave" are metaphors. Just because a mathematical or scientific term has a different meaning outside the specialty doesn't make it metaphorical.
posted by Proofs and Refutations at 3:01 AM on October 18, 2013


There are no experiments in philosophy. There are no control groups.

Welcome to Intro to Philosophy Lab 231, today, class, we will be testing the being and existance of these cats, make sure you've all got your safety goggles and rubber gloves.
posted by The 10th Regiment of Foot at 4:56 AM on October 18, 2013


Hi,

I am the main subject of the article (Nick Brown). (I'm not sure how to prove that. You could look at the Twitter account with this username and see that it belongs to Nick Brown, but of course I could be hijacking that too.)

@Mrs. Pterodactyl: I don't like the "plucky amateur" meme too much either, but it wasn't my idea. Otherwise I quite like the way the back story has been written, after previous articles had discussed the science. It will save me telling the same war stories over and over again in tedious detail. Well, it probably won't, until I do something else that's as interesting, but it's the thought that counts.

@math and others who commented on the absurd precision of 2.9013: read our paper. We demonstrate that in fact the number is (exactly) 441/152, or in other words, a rational number (2.901315789473684210526, with the last 18 digits recurring). That's how absurd Fredrickson and Losada's claims are: they aren't even based on statistics, they're based on an equation that just divides one set of constants by another. (My bold proposition: To a first approximation, any time you think you've found a rational number in nature, you should start your method over.)

@daq: You can sign up for one of Losada's courses here. Check out the companies who, he claims, use his services, including Boeing, GM, and Apple. Sure. (Maybe one person from the HR department came to a public seminar one time.)

@RogerB: Tell me about it. I don't know how much it costs to have Fredrickson come and talk to your company about how positivity can help everyone more profitable, but a lecture from Seligman runs $30,001 and up. A "nice little earner" as we Brits say, especially if you can get away for years with making elementary mistakes when describing other people's research.

Finally, @everyone who has posted here: thanks for your support!
Nick
posted by sTeamTraen at 9:59 AM on October 18, 2013 [12 favorites]


heckuva job, Brownie!
posted by MisantropicPainforest at 10:05 AM on October 18, 2013


« Older And so my life began: Junior Lobby Boy-in-training   |   What if it decides to fire? Newer »


This thread has been archived and is closed to new comments



Post