Superstition Works

I was very slow to get off the starting block on this one as even Ben Goldacre has already blogged this!  However, I have had something major to deal with this week (and no, its not you dear lady) but here is a short blog about a forthcoming study in Psychological Science about the tangible benefits of superstitions. A team from Köln in Germany has shown that university students of whom more than 80% believe in luck, perform significantly better on a putting task if they think they have been handed a “lucky” golf ball. They also did significantly better than controls on a second experiment if they were told that someone was crossing their fingers for them. In a third experiment, students who had brought their lucky talisman along to the testing session did better when it was in the testing room. The fourth experiment demonstrated that these lucky students attributed their better performance to improved self-efficacy. So there we have it. If you believe in lucky charms then you perform better because of perceived self-efficacy. It’s all in the Science of Superstition book, though I call it the perception of control.

I should point out however, that one error in the book I made was to endorse the Skinnerian account of where supersitions come from that has been re-iterated by other scholars on the subjects such as Andrew Vyse, in his 1997 book, Believing in Magic: The Psychology of Superstition. We have both reported Skinner’s famous 1948 paper on the origins of superstitious behaviour where he put several pigeons in Skinner boxes, set the food dispenser to deliver food every 15 seconds, and left the birds to their own devices. Later he returned and found the birds engaging in odd, idiosyncratic behaviours, such as pecking aimlessly in a corner or turning in circles. He referred to these behaviours as “superstitious” and offered a behaviourist analysis of their occurrence. The pigeons, he argued, were simply repeating behaviours that had been accidentally reinforced. Therefore superstitions were thought to simply emerge by the random reinforcements that we occasionally encounter such as having a great tennis match and wearing the same clothes again to try and repeat the success. However, this turns out not to be entirely robust and while some humans show this propensity, it is not inevitable.

Anyway back to our German psychology study. Maybe this research on superstition and self-efficacy explains Germany’s stellar performance in their game against Australia and why England fumbled the ball!

13 Comments

Filed under General Thoughts, supernatural

13 responses to “Superstition Works

  1. endlesspsych

    If supersititons “work” it would be my guess that they do so by reducing anxiety and stress via the ability for people to excercise some degree of control over situations in which they have none.

  2. brucehood

    Exactly endlesspsych, hence that’s why I call it the “perception of control” in the book which has been shown to reduce stress

  3. D’oh, thanks a lot, Bruce. I found that “superstitious pigeon” experiment to be one of the most fascinating accounts in your book, and, having swallowed it uncritically and failing to do any verification for myself, have repeated the anecdote to others on multiple occasions.

    Oh well, thanks at least for admitting the error. I do appreciate it🙂

    On the central topic of the post… I strongly suspect that many of the techniques I employ in putting my 15-month-old son to bed are not actually doing anything, and are basically just “superstitions” on my part. But 1) it would be too costly (in terms of lost sleep) to methodically vary my technique to figure out what really matters and what doesn’t, and 2) much like the students in this experiment, doing all these silly little things gives me a feeling of control over what is otherwise a very frustrating situation.

    To put it more directly, it’s much easier to patiently endure a fussy toddler if I believe I am doing something that will help get him to sleep faster — even if my belief is totally false.

  4. brucehood

    Hi James, yes like many classic psychology studies there turns out to be a lot of misinterpretation –
    Harlow’s monkeys
    Watson’s little Albert
    Kitty Genovese case
    Solomon Asch’s conformity studies
    Phineas Gage
    and so….

    Maybe the “Psychological Myths” should be another book one day?

  5. Thanks Dr. Hood.

    I was explaining the Supersense at a party yesterday to my brother who is a skeptic and a genius. It really helped both of us to frame the tendency to believe in ghosts and gods and killers’ cardigans as aspects of necessary neurological processes, and he and I were devising experiments to test levels of Supersense (I have tons, he has very little). Thanks for your work!

    About magic and ritual: have you read the excellent article on Baseball Magic by George Gmelch? He looked for, and found, differences in the use of magic and ritual based upon the level of uncertainty facing baseball players. Pitchers, for whom 4 hits out of 10 tries is considered marvelous, cover themselves with magic and talismans, while outfielders, who can expect to catch the ball 9 times out of 10 tries, use very little. Gmelch does refer to the contortionist pigeons at the end of this article, though!

    As I was leaving my new age career and returning to college, I happened upon this article and was really helped by it. I was able to look at all my magic and rituals and taboos and ask., “Hmmmm, is this all just about fear and uncertainty?” It was nice to have a new frame for it.

    Your book is also a great frame; many thanks to you!

    And yes, please write the book on Psychological Myths when you have the time (hah!).

  6. Hah! A just released TED talk by Mike Schermer makes reference to the superstitious pigeons:

  7. OK, I can’t let this one go past without a comment. I know that there are lots of problems with Skinner’s experiment and the conclusions he draws from it (it’s still fun to do, though). But, the way you put it, it sounds like you’re denying the whole idea that superstitious beliefs are formed when, under stress, people become too willing to see causal connections between events and, not having a natural explanation available, go for a supernatural explanation. Are you seriously denying that? Or, are you simply denying the behaviourist idea that what is central to this phenomenon is simply coincidental reinforcement? How, for example, do you feel about error management theory?

    And references. I want references!😉

    … if it’s not too much trouble, that is.

  8. brucehood

    Ok.. the main criticism of Skinner study is that the behaviours that the pigeons exhibited were not random but rather all examples of foraging behaviours (Timberlake, W., & Lucas, G.A. (1985). The basis of superstitious behavior: Chance contingency, stimulus substitution, or appetitive behavior? Journal of Experimental Analysis of Behavior, 44, 15-35. When the same operant conditioning was done with humans, only a proportion showed evidence of superstitious behaviour (Ono, K. (1987). Supersitious behavior in humans. Journal of Experimental Analysis of Behavior, 47, 261-271.)

  9. Right. I suspected you were referring to the Timberlake and the Ono studies. If I recall correctly, Vyse talks about the Ono study and it is useful in that it strongly suggests that people, at least, do tend to think of their superstitious behaviour as affecting the outcome – which would support a Skinner-like interpretation of the reinforcement studies, it seems to me. The fact that only some people show evidence of the behaviour does not strike me as particularly problematic. I am quite sure that had the participants been placed under stress the percentage exhibiting the behaviour would have gone up, as per the Keinan study on the influence of stress on ‘knocking on wood’ behaviour.

    I suspect that the core of the issue is whether ‘perception of control’ works as an explanation of the behaviour. The way I understand the thesis is that superstitious behaviour functions in people to create the feeling that they have control over situations they have no control over. In other words, that the way to explain superstitious behaviour is motivational. However, this explanation does not sound plausible to me once considered in an evolutionary context. Unease about loss of control is clearly adaptive and anything that could short-circuit it would be potentially maladaptive. It seems much more likely that the primary explanation has to be cognitive – superstitious behaviour is a by-product of efforts to control a situation that is, as a matter of fact, beyond our control. Once you have that situation it is, of course, possible that superstitious behaviour does lead to false feeling of control and lower tension.

    It strikes me that the story Nick Humphrey tells about the placebo effect is somewhat similar in that cutting out the cognitive side of the story would be like the body never holding back any resources – it might be good in individual cases but, overall, it would be maladaptive.

    Still, it seems to me that you have a somewhat different take on loss of control because you are talking about a study which suggests that superstitiously motivated belief in having control over a situation may increase performance. This is something different and more interesting than what much of the loss of control literature talks about since here the effect is not just emotional but actually in terms of performance. While a fascinating case, I would find it hard to believe that such an effect should be generally explanatory. Superstition goes way beyond what could be explained in terms of something like self-fulfilling prophesies, after all.

  10. Certainly, there is a tremendous gap between “superstition can increase performance” (which seems at least superficially plausible) vs. “superstition evolved as an adaptation to increase performance” (which is prime facie implausible, and therefore would require pretty extraordinary evidence).

    I’ll go one step further and say that not only do I doubt that the putative effect shown in this study is generally explanatory, and add that I doubt it’s even generally applicable. Intuitively we tend to feel that there are situations where superstition can improve performance (Dumbo, anyone?) so it would be interesting to emperically demonstrate that phenomenon, as these researchers believe they have done. But it’s not at all difficult to identify situations where superstition degrades performance.

    When I read this study, the takeaway for me wasn’t that superstition was adaptive or “good for you” — it was that superstition isn’t always “bad for you”.😀

  11. I am guessing that the “illusion of cause” as discussed in The Invisible Gorilla (Chabris & Simons, 2010) plays out in the formation of superstitious behavior. I’m curious as to how the “perception of control” differs and/or overlaps with the illusion of cause.

    On a very basic level are not all superstitious behaviors mediated by learning theory principles? Behaviors repeated based on the perception of control are “reinforced” by the NAcc via pleasure associated with the flood of dopamine (particularly when the pairing is confirmed, even if only by chance) . Departures from rituals are “punished” by arousal of the insula (and the associated negative emotions). As we know, intermittent reinforcement is the most powerful form of reinforcement (most resistant to extinction). Given our propensity for confirmation bias, expectancy effects, and the intermittent nature of ritual reinforcement, aren’t we just primed to be superstitious?

  12. Pingback: Superstition Improves Confidence and Performance - Science and Religion Today

What do you think?

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