Psychology seems from the outside like such a weird mix of extremely good and super interesting science and totally unbelievably stupid junk science. Maybe it's mostly the former in university departments and the latter is for press consumption only, but it sure seems like a lot of the totally horrible junk scientists have university appointments (I'm thinking of e.g. that crazy attachment parenting obsessed professor we talked about a while back who I think had an appointment at Notre Dame).
Hey, this could be my new research program. I'll just start applying the Lorenz equations to random phenomena that have come up in philosophy!
R1 here I come!
People Unfogged commenters sure are intimidated byinundated with math.
I guess we may be dealing with just one junktastic scientist (who is also a tarheel, sucks for you Apo). See also this piece, linked in the link.
In the 2005 report, Fredrickson and Losada failed to show how individuals' self-reported feelings could be mathematically described as quantities that vary smoothly over time, as the Lorenz equations require,
Apparently there were other problems as well, but this issue -- imposing continuity assumptions on non-continuous phenomena -- is absolutely endemic in the social 'sciences' including more mathemtically sophisticated ones.
The problem discussed in the piece linked in 6 -- taking residuals from a kitchen-sink regression, or some other purely statistical / not theoretically informed mechanism for finding independent variation -- is also endemic. Good to see someone call that out.
This book from 2003 uses nonlinear dynamic systems (like the Lorentz attractor) to model marriages:
http://www.amazon.com/gp/aw/d/0262572303
I am pretty sure they got the math "right" because the opening chapters explain what a nonlinear dynamic system model is pretty clearly. They didn't get any useful results though.
SEE! I TOLD YOU—SCIENCE IS BOGUS!!! GLOBAL WARMING IS A HOAX!!!!!
A bit related, Wonkblog seems to be going farther into get-a-snappy-headline-from-the-press-release territory.
1: nah, not really. An unfortunately high percentage of social psychology (especially on squishy topics like emotion and, like, general positive thinkinging) is nonsense. But there is an awful lot of psychology that isn't social psychology!
In general, the higher level and more abstract the subject of the research (life positivity, for example) the more cultural continence and lack of rigor and PR-hunting one finds. At least, seems that way to me.
I will offer a shorthand translation for you, Sifu:
7.1: the assumption of continuously differentiable production functions in economics always drove me a bit nuts, although I suspect my bigger beef is with aggregation.
7.2: the underlying point is that local average treatment effects can be very different than the broader treatment effect you may naively think you are getting. As this article points out, that's a problem that exists for regression as well as natural experiments. So you need a theory for why the subgroup you are picking out with your independent variation has external validity, or why they aren't some eccentric subpopulation that happens to be picked out by your data mining.
Oh man this is the best:
The researchers determined that creativity, helpfulness to others and other elements of "flourishing" characterized people who displayed a ratio of positive to negative emotions above 2.9013:1 and below 11.6346:1.
Yes, please quantify my ratio of positive and negative emotions to six significant figures with your fancypants emotimeter.
I want to read the article with the Lorenz equation but when I follow the links in the article, even going through my institution's library, I get something telling me I have to pay $12 to read the article. Really?
There's a preprint of Brown, Sokal & Friedman (2013), "The Complex Dynamics of Wishful Thinking: The Critical Positivity Ratio" on arXiv.org. It's nice to see that Alan Sokal is still fighting bogus applications of mathematics.
I've known someone who had marriage counseling from Gottman (author of book in 8). She thought the counseling was helpful. You're right that the math is clearly explained, too.
The summary of Gottman on marriages, as I remember it, is that (a) contempt is unsurvivable, so make do without sarcasm and belittling; (b) many marriages work better if the woman is slightly subservient, so why not go along to get along?
(b) still makes me absolutely furious, possibly because I totally believe it came honestly out of observation and is the best local result most women can get. (He doesn't phrase it quite as bluntly, either.)
13: the more cultural continence
Cultural continence is a bad thing? You'd rather the person peed on themselves at the opera?
19: I'm pretty sure he also found that marriage work best if the man accepts the woman's influence (essentially, deals with the problems she brings up and does what she suggests much of the time). At least, that's what my partner tells me. But google also confirms it (try keyword Gottman wife influence). That doesn't conflict with the wife accomodating the husband too, but it complicates whatever definition of 'subservient' you are thinking about.
21: The kingdom works best if the crown accepts the subjects' influence. Not egalitarian.
Actual kings usually do not think that way.
I don't know enough kings to say. Bosses and bureaucrats will use similar language, which bodes ill. If it's set up with one who Decides and one who Influences, sure, it will work better if the Decider is kind and attentive. But the Influencer is still subservient.
Actual kings may yet behave that way, though - a working knowledge of public opinion could certainly be useful even if the king weren't thinking about the relationship in terms of accepting influence.
In any case 22 and 23 are consistent as long as actual kings' kingdoms don't usually work in the "best" way possible.
(This may all be separated from the original question by enough degrees to constitute a questionable analogy.)
... and may also have missed the point of 22.
16: If your library is supposed to have a subscription and it's not working, you should contact a librarian - there's probably a feedback link or "ask a librarian" link or something where you can report problems. In my experience as a library user, they're usually pretty good about fixing problems with electronic journals because 1) having a library that works is important, 2) you do not want angry faculty, and 3) they're probably paying a shitload of money for the journal so it damn well better work.
Halfordismo enacts the people's will, but only I can discern their true will from their purported surface "opinions." Deep down you WANT me to have that crocodile tank and gold plated submarine.
Under Halfordismo, enacting the people's will merely means that they will be forced to eat meat.
This thread is done, but I just went back to my Gottman book and reread the 'sharing influence' part. It's quite clearly aimed at actual sharing of power and decisionmaking, he uses that phrase several times -- unless men genuinely share decisionmaking authority the marriage will fail. It will vary by couple whether that ends up completely egalitarian or not but I think it's unfair to portray it as Clew does, as just advocating more benevolence for the monarch.
Okay. I remember his finding an equilibrium with men who would not share authority, but could be pacified by their wives. But if he doesn't describe that as a good thing, my rage was misplaced.