The Wrong Guys

The social sciences have developed in ways that are not necessarily to our advantage. How to fix?

Jonathan Haidt thinks that social psychology has problems (I might put it a bit more strongly) that might be ameliorated by adding ideological diversity – Republicans or Jacobites or whatever.

Personally, I doubt it.  I don’t think that the guys who produce all those unreplicable results were driven by the search for that Eureka moment when you finally figure it out, finally see it clearly.  I don’t think it’s really a product of people with the right motives whose statistics are sloppy, which we do see in some other fields and can sometimes fix.

I think they’re just no damn good.

 

 

 

Advertisements
This entry was posted in Uncategorized. Bookmark the permalink.

73 Responses to The Wrong Guys

  1. JayMan says:

    That’s my thought on the matter as well.

  2. JayMan says:

    Good to have you back!

  3. Donnie says:

    Who’s gives a shit about social science…Welcome back!!

  4. Anonymous says:

    I don’t think Jonathan Haidt thinks the replicability crisis in soc psy is driven by the ideological monoculture. Those are two different issues.

    • gcochran9 says:

      I think otherwise. I think that the problem isn’t just being a monoculture: it’s a bad monoculture.

      • magusjanus says:

        I think Haidt’s point is that it easier to illustrate the nature (and errors) of the “Bad” monoculture by having a rival culture with a vested interest in attacking it doing so. the idea being they’d keep each other “honest” as it were.

        (glad you’re back btw)

  5. My impression from my days in anthro was that the entire field had become dominated by the desire to tear down the work of those who’d come before, rather than build upon it. So you were always searching for things others had missed or ways to claim they were wrong. Anything that held steady over time, like easily replicated results, was therefore useless. Much better to come up with some new way the old stuff was wrong than to replicate the old.

    I know that falsifiability is fundamental to science, but this was something else, falsification solely for falsification’s sake.

  6. Steve Johnson says:

    The problem isn’t that they lie for ideological ends. The problem is that they’re picked for ideological reasons.Yes, they’re no damned good but why? Because the part of the selection process that’s supposed to weed out the no damned good isn’t there.

    P.S. Welcome back.

  7. M. M. says:

    Welcome back. Else, you’re spot, again, quelle surprise. As for Haidt, I wonder what he’d say when drunk. Maybe he’s just being diplomatic on tactical grounds.

    • rzg says:

      When drunk. Or after a spot of heart surgery.

      @Cochran, I wonder if you’ve any thoughts on the degree to which a tendency to no-damn-goodness might have a genetic marker or two associated with it? And whether this might prove to be a useful tool for… well, helping to catalyze prejudicial reactions, against humans who can be expected be less damn good than others?

  8. ursiform says:

    Welcome back, Greg.

  9. No damn good because…. those who are very bright, and can do the Maths, understandably go elsewhere, in the familiar STEM direction, and leave social psychology to novelists, daydreamers, and professional rebels. Real social psychology is done by Amazon and Google, who harvest more real data in one day than a century of psychologists.
    Long may your health persist and your irascibility flourish.

  10. Bla says:

    Welcome back.

  11. silberstreak says:

    Social science is to science as mud pies are to pies.

  12. MawBTS says:

    I see they have rebuilt Greg.

    Personally, I doubt it. I don’t think that the guys who produce all those unreplicable results were driven by the search for that Eureka moment when you finally figure it out, finally see it clearly

    Tumblr user su3su2u1 had a great rant against “Eureka moments”, and how untrustworthy they are. In his career as a physicist, he’d so often feel like he’d blown a problem wide open, had resolved a critical issue, had neatly wrapped a bow on everything…

    …and then he’d go to bed, wake up, and realise he’d missed something, or made the problem too simple, or done something else that invalidated his Eureka moment.

    He was of the opinion that actual scientific progress is often people who feel like they’re way over their heads, groping in the dark. Darwin is a good example: he was baffled by a lot of his findings and just had to hope that later scientists would find a way to make his theory work (Mendel etc).

  13. Spotted Toad says:

    A good scientific theory starts with regularly observable phenomena; then you get an accurate description of the patterns underlying the phenomena ; then ideally you get the lasting rules that drive those patterns.

    First you get Babylonian astronomers up through Tycho Brahe, accurately recording what they see in the sky, then you get Kepler describing the structure of the cosmos that would produce those patterns, then Newton explaining why that structure would work the way it did. It took a while.

    Maybe the problem for psychologists is they want it all at once. They want a novel, unusual phenomenon (that’s what the Diedrick Stappel fakery was all about), that thanks to the supposed magic of an RCT reveals the causal agent producing that phenomenon. Even more than that, they want something actionable, a causal agent that Somebody can Do Something About.

    With psychometrics, we have years upon years of data, that fits into regularizable patterns, and that we are beginning to identify causal agents behind. But anything actionable in anything more than a folk wisdom sense is probably a ways off.

  14. Soc-student says:

    OT I’m a student in the social sciences and thinking about changing to another program where I would be able to do genetics research, specifically on substance abuse and disease. My issue is that I don’t know if I’m smart enough to do it, and I’m wondering what would be hardest thing I’d have to learn? The reason I’m asking is that I’m planning to self-study beforehand to see if I can hack it.

    • Just go for it and adjust later on if you can’t hack it. Being scared of your possible lack of potential is the biggest mistake you can make. It is far a bigger mistake than trying and failing. And you might very well succeed. I’ve been through this before and finally learned my lesson. Just try it.

    • James Miller says:

      Look at the textbooks assigned in the program you are thinking of switching to. You can probably buy cheap used copies of the old editions on Amazon.

  15. another fred says:

    I think you are dealing with people who feel threatened by the real world and are desperate to deny it.

    The threat comes from two things mixed in varying degrees in different people.
    1. A sense that the world is horribly unfair, usually rooted in an unhappy childhood,
    2. A perception that if man continues on the usual path we are headed for catastrophe.

    My reply is:
    1. Unfairness is how nature operates, get over it.
    2. I think we are headed for a “catastrophe”, but it will mostly come in its usual form of war which will cut back the population and there’s nothing we can do about it except struggle to survive (see 1, above).

  16. jb says:

    Welcome back! Are you feeling 100%, or is there still a way to go?

  17. In fact, Philip Tetlock is a social psychologist, and he is doing excellent work. He bothers to see whether predictions are accurate

  18. pyrrhus says:

    Contrary to Heinlein, corruption is always more likely than incompetence in modern society. Best wishes for your recovery!

  19. Andrew says:

    I have no doubt that a few of these social scientists are brilliant. But, there is truth in the saying that the problem with (many) smart people is that they just fool themselves more brilliantly. The social sciences unfortunately have attracted intelligent people burdened with biases and/or lack self-awareness and on the next lower level an army of students that mostly lack the ability to handle the rigor required to do quantitative science follow.

    Social scientists have assumed that too much behavior is learned when all behavior is heritable. Statistics is required to disambiguate the learned encoding from the heritable layer. Social scientists need to understand about heritable behavior and statistics to do this. They don’t or they do it poorly and so the social sciences mostly thrive in universities that cater to students who want to take a soft subject in order to graduate.

    Anthropology was supposed to be interdisciplinary study of man, instead it has fragmented into cultural, physical, forensic, etc. Why has an interdisciplinary field fragment?

    Welcome back!

  20. Good to have you back and in fine form.

  21. Asher says:

    Social scientists don’t seem very interested in cause and effect. Just keep asking them “what causes/caused that?”. You won’t get an answer.

  22. Philip Neal says:

    You are back! Hip hip hooray!

  23. To speak in familiar numbers, social scientists tended to do much better on SATV than SATM, sometimes to the tune of 200 points. There are also people who score +200 on SATM over SATV, but they are rarer.* Each goes wrong in different way. The former tend to be better at social cues and knowing what should be believed. Politicians, educators, and journalists/writers/artists tend to fall into that group, which tends to self-reinforce so that everyone there eats. It works great for them. It doesn’t work so well for increasing the sum of human knowledge, however.

    *I believe those few are loaded to spatial abilities.

  24. Abraham Lincoln says:

    The social sciences are no damn good because they are the beating heart of the whole tolerant diverse liberal progressive political correctness thing. Their sole reason for existence is to provide “scientific” legitimacy to politics like feminism and immigration and welfare and whatnot, which means they merely need to appear to be science. In other words, to “fix” social science, you’d have to first kill the beast.

    But I suspect you know this already.

  25. Mt Isa Miner says:

    Glad you’re back. Take care of yourself.

    • Fourth doorman of the apocalypse says:

      I was a Mt Isa minor a long time ago for a year or so. However, I found Darwin more interesting, eventually.

  26. Karl says:

    You think they are no damn good? That’s because you think their job is science. I don’t think they are paid to do science. They are paid to produce propaganda. Do you think they aren’t any good at that?

    • another fred says:

      Very good point.

    • Spotted Toad says:

      Yeah, social science reminds me of Milton’s statement that he would “justify the ways of God to man.” The fear evinced by mainstream social scientists is not just of any specific hatefact becoming known, but of knowledge in general.

  27. Greying Wanderer says:

    wb

  28. Jim says:

    The social sciences are not really sciences in the sense of having a goal of discovering true propositions. They play a role in modern society much more akin to that of an ideology or religion.

  29. j says:

    I’d not say social science practitioners are “no good”. Most are not bad nor evil. The facts show that they tend to be measurably less intelligent and less math orientated (see Jayman above). But science’s progress is based on the work of a few geniuses (like Newton, Leibnitz, etc.) that drag after them the masses. In social sciences, it appears to me, the geniuses (that exist) are suffocated by the reigning mediocrity. The furious peasants kill off the over-smarts. The miracle is that in physical sciences that doesn’t happen. At least not so often.

    • Andrew says:

      I agree but many of the geniuses in social sciences are warped and they have gotten away with it because social sciences in the past have been more difficult to falsify whereas the physical science has been tightly tethered to math and thus falsifiable.

  30. Cognitive psychology seems to have its act pretty much together, judging by Stanislas Dehaene’s books, or Pinker’s How the Mind Works. Lots of replicable findings that aren’t just restating the obvious. What’s the difference? Less controversial subject matter? Sharper, more motivated researchers?

  31. sinij says:

    Many areas of modern science have serious issues with reproducing the results. This can be only fixed by addressing core academic issues – pressure to publish regardless of merit of findings, almost universal opposition to publishing negative results, and last-century scientific publishing journals.

    The start would be to require publishing of all data along with the article, and acceptance of replication studies in top-tier journals. Once we clean up hard sciences, we can start demanding better from soft sciences.

    • gcochran9 says:

      So, once we solve the Kepler problem, we can then demand that the softies clean up their act? Or is after we figure out the electroweak theory? Or continental drift?

      • sinij says:

        You think it will be THAT difficult to clean up scientific publishing?

        • gcochran9 says:

          Your comment was nonsensical. There is no need or reason to make all the hard sciences perfect (they are already pretty good) before making any effort to clean up absolute swamps of iniquity like sociology.

          • sinij says:

            I guess my attempt at a joke didn’t turned out to be funny.

            Yes, sociology is festering. No argument here.

            I disagree with you on “already pretty good”. Theoretical sciences aside, my personal experience is that it is quite swampy even on this side of the divide. I also believe the reasons for this malady are shared, if we clean the house here it will also largely address soft science problems.

            Key problems are: a. imperfect mutual anonymity of a peer review process allowing cliches, popularity, and group think to corrupt the process, b. discrimination against publishing replication or negative result work, and c. no mandatory publishing of your research data.

  32. May you mock the damned fools for decades to come.

    • Chandra Chisala who wrote this article was bright enough to get a degree in biochemistry form the University of Zaire and now has a fellowship at Stanford. He states that “I do not believe that the current US or UK immigration system mostly selects for the best people from places like Africa.” He is a liar. he knows he is lying, his life is proof of the lie. Anyway Jayman and others have decent responses to his bullshit.

    • TWS says:

      Unz has some strange ideas about race and uses his own models rather than actual data because he simply cannot massage reality to fit his preconceptions. His models do that just fine though.

      He appears wedded to the open borders free movement of peoples he likes. It’s good for him so it must be good for everyone.

  33. ohwilleke says:

    Forget social psychology. Hurray for medical science!

  34. Asher says:

    Okay, so let’s say psychology as a discipline, today, is just no good. Is it possible that there is some discipline we might psychology that would be any good?

    What would it’s starting axioms look like?

  35. Robert King says:

    Thinking of psychology in terms of axioms is unlikely to prove fruitful. As a matter of history, trying to think of psychology that way for about a hundred years hasn’t proven fruitful. The physics envy that underlies this is effectively skewered by the currently smartest person on the planet, Marvin Minsky here (and following) https://www.youtube.com/watch?v=-pb3z2w9gDg
    There’s a good reason for this–the brain is made of hundreds of little robots all evolved for doing different jobs and lots of them getting in one another’s way. So don’t expect any algorithms any time soon. What we are going to do (and, indeed, what we are doing) is painstakingly identifying each of these little functional elements in terms of computational neuroscience. So if you want your big theory there is going to be a lot of smearing of detail.

  36. James K says:

    We can learn something useful from the foundation of modern Sociology. Durkheim did not seek to infiltrate and subvert the field of Anthropology – even though the subject matter of Sociology and Anthropology is arguably the same. Instead he founded a new discipline.

    Sociologists will always see issues in terms of sub-Marxist ideas such as class struggle and oppression. It will be impossible to change their minds, and very hard to do useful work in a Department where such ideas dominate.

    It will be far easier to conduct rational enquiries into social science if you give it a different name, and organise it within the science faculty.

  37. teageegeepea says:

    This is a matter of medical rather than social science, but this account of how the cure for scurvy was forgotten in the 19th century reminded me of how antibiotic treatments for ulcers were forgotten more recently:
    http://www.idlewords.com/2010/03/scott_and_scurvy.htm
    Interestingly, the two accounts are opposite in how they relate to the germ theory of disease.

  38. Abelard Lindsey says:

    The soft fields attract soft people. That’s all you need to know about them.

    • AnonymousCoward says:

      Cochran is speaking to something beyond softness.

      ” I don’t think it’s really a product of people with the right motives … I think they’re just no damn good. ”

      Impure motives. Outright (self)deception. Statements consciously unaligned with the truth.

      People like Gould don’t simply get it wrong because they’re soft. There is something more, something worse.

      • Jim says:

        If they were just bumbling incompetents they would eventually fall over the truth. They’re ideologues.

        • albatross says:

          You need a system in place that rejects false statements and holds onto true ones–something that naturally tends toward getting a better picture of reality over time. In such a system, less able people will make less progress than more able people would have managed, but they can still make progress. My guess (as an outsider) is that a lot of the social sciences have a broken system–perhaps broken by political taboos that can’t be crossed, or by not having a good way to reject false statements, or too much ideological conformity meaning there’s nobody to ask inconvenient questions, or whatever else.

          In such a system, even really smart, hard-working people will fail to make much progress at getting closer to reality, because the correct statements aren’t any more likely to survive and be built upon than the incorrect ones.

          • albatross says:

            To build on that a bit, think of medevial medicine. There were some pretty smart and learned guys doing that. But they were overwhelmingly stuck in completely screwy models of disease, with almost no mechanisms for correcting their wrongness. In that system, boosting the average IQ of physicians from 110 to 120 probably wouldn’t have helped much. You needed a culture that allowed experiment, accumulation of evidence, and rejection of wrong ideas. (Probably you also needed some probability theory to do meaningful observations/experiments, and maybe microscopes to figure out the main cause of people getting sick and dying.)

          • In this comment, and your later medieval medicine example you say “you needed a culture that allowed experiment, accumulation of evidence, and rejection of wrong ideas.”. Correct. Now, where do such things come from, other than from a bright person who explains why they are necessary? Culture doesn’t just happen, someone has to invent the plough.

  39. John Hostetler says:

    “I think they’re just no damn good.”
    Meaning incompetent? Or up to no good?

    • Toddy Cat says:

      Of course, the two are not in any way mutually exclusive…

      • John Hostetler says:

        They have shown a great deal of competence at marching, up to no good, through every institution of Western civilization.

        The heritable factor Greg Clark is getting at in ‘The Son Also Rises’ has been approximated here as ‘moxie.’ In the spirit of traditional ethology, why not simply call it dominance?

        Nature is parsimonious and complex – it’s impossible to pack every desirable trait into 1400 cc’s of brain. The most intelligent people tend not to be the most dominant. When they become dominated by people with only slightly lower intelligence, their natural defense mechanism is to cry incompetence.

    • Andy says:

      Probably both.

  40. Paul N. says:

    „The social sciences have developed in ways that are not necessarily to our advantage.“
    Though they do have developed to the advantage of the money source, the government. That’s the fundamental rule of evolution: everything evolves to get the most of the resources. Admittedly that implies that government is not “necessarily to our advantage” which for most of you is difficult to digest.

    “How to fix?”
    Simply develop social science isolated from governmental support. BTW, if you keep your eyes open you can spot a lot of social science evolving outside governmental control. As a matter of fact that’s where social science becomes most exciting.

  41. Pingback: Social science dilemma | Fauceir Blog

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s