June 12th, 2010 by Ben Goldacre in bad science, irrationality research | 61 Comments »

Ben Goldacre
The Guardian
Saturday 12 June 2010

As someone who strives – sanctimoniously – to be right, I’m a masochistic fan of research showing that people who are wrong have better lives than I do. This is why I particularly enjoyed a study from the current edition of Psychological Science showing that being superstitious improves performance on a whole string of different tasks.

Now to be honest, I’m always a bit conflicted about this kind of psychology research. On my left shoulder, there is an angel who points out that it’s risky to extrapolate from rarefied laboratory conditions to the real world; that publication bias in this field (the phenomenon where uninteresting findings get left in a desk drawer unpublished forever) is probably considerable; and that it’s uncommon to see a genuinely systematic review of the literature on these kinds of topics, bringing together all the conflicting research in one place. I am not Malcolm Gladwell, if that helps to frame the issue more clearly, and I think his books are a bit silly and overstated.

On my right shoulder, there is a devil who thinks this stuff is all just really cool and fun. He is typing right now.

The researchers did 4 miniature experiments. In the first, they took 28 students, over 80% of whom said they believed in good luck, and randomly assigned them to either a superstition-activated or a control condition. Then they put put them on a putting green. To activate a superstition, for half of them, when handing over the ball the experimenter said: “here is your ball. So far it has turned out to be a lucky ball”. For the other half, the experimenter just said “this is the ball everyone has used so far”. Each participant had 10 goes at putting on the green, trying to get a hole-in-one from a distance of 100 cm: and lo, the students playing with a “lucky ball” did significantly better than the others, with a mean score of 6.42, against 4.75 for the others.

Then they moved on to a second experiment. 51 students were asked to perform a motor-dexterity task, an irritating fiddly perspex game where they had to get 36 little balls into 36 little holes by tilting the box. Beforehand, they were randomly assigned to one of three groups, each hearing a different phrase just before starting. The superstition activator was “I press the thumbs for you”, a German equivalent to the English expression “I keep my fingers crossed”. The two “control” or comparison groups were interesting. One group were told “I press the watch for you”, with the idea that this implied a similar level of encouragement (I’m not so sure about that) and the other were told: “on ‘go’ you go”. As predicted, the participants who were told someone was keeping their fingers crossed for them finished the task significantly faster.

Then things got more interesting, as the researchers tried to unpick why this was happening. They took 41 students who had a lucky charm, and asked them to bring it to the session. It was either kept in the room, or taken out to be “photographed”. Then they were told about the memory task they were due to perform, and asked a whole bunch of questions about confident they felt. The ones with their lucky charm in the room performed better on the memory game than those without, but more than that, they reported higher levels of “self-efficacy”, which was correlated with performance.

Finally, they probed these mechanisms even further. 31 students were asked to bring their lucky charm, it was either taken away or not, and they were given an anagram task. Before starting, they were asked set a goal: what percentage of all the hidden words did they think they could find? Then they began: as expected, participants who had their lucky charm in the room performed better, and reported a higher degree of “self-efficacy” as before. But more than that, people who had their lucky charm in the room set higher goals, and also persisted longer in working on the anagram task.

So there you go. Almost everyone has some kind of superstition (mine is that I should mention I noticed this study through my friends Vaughan Bell and Ed Yongon Twitter). What’s interesting is that superstition works, because it improves confidence, let’s you set higher goals, and encourages you to work harder. In a lab. You now know everything you need to decide if this applies to your life.

If you like what I do, and you want me to do more, you can: buy my books Bad Science and Bad Pharma, give them to your friends, put them on your reading list, employ me to do a talk, or tweet this article to your friends. Thanks! ++++++++++++++++++++++++++++++++++++++++++

61 Responses

  1. Mark P said,

    June 12, 2010 at 1:12 am

    An experiment carefully leaving out all the bad bits of superstition is, unsurprisingly, going to miss all the bad bits.

    What about telling the putters that “We have always found Tuesdays to be unlucky” before they start? Or telling them that the guy next door has put a curse on them?

    Really superstitious people, in my experience, suffer from more concern over bad luck than gain confidence in good luck. Therefore I doubt they do better in real life.

  2. Flat Eric said,

    June 12, 2010 at 1:23 am

    This does at least help explain why I’ve spent 22 years trying, without luck, to get better at golf.

  3. reprehensible said,

    June 12, 2010 at 2:33 am

    I got 40% in a module this week but luckily for me the admin staff believe in making up results if they don’t really know them. Fortunately my tutor broke blinded marking protocol to tell me it was really about 72 but don’t hold my breath for confirmation. I also wish I could get close to that score in golf!

  4. JimLee said,

    June 12, 2010 at 3:15 am

  5. Mark Wainwright said,

    June 12, 2010 at 3:42 am

    Ben, you introduce this study as ‘showing that being superstitious improves performance’, but as described it doesn’t do so, as there is no control group of non-superstitious people.

  6. Jut said,

    June 12, 2010 at 7:32 am

    another poor quality psychology study. Who would have though…
    I’m still with Feynman in that it’s a Cargo Cult science

  7. Stitcheroo said,

    June 12, 2010 at 8:58 am

    Er … isn’t it equally possible that lucky charms and superstition actually work rather than some “improves confidence” explanation? Aren’t you (or the researchers) trying to impose some pre-conceived idea about why this is happening (based on a very limited, non controlled study from what I can see)? There may be some interesting stuff here but to draw any conclusions from what is presented is ludicrous.

  8. Karellen said,

    June 12, 2010 at 10:12 am

    @Stitcheroo: How do you account for the results where golfers were told about a ball being lucky, or not, when a) no-one ever claimed that the ball was lucky, and b) the people who were not told it was lucky got the same ball, but did not do as well?

    Similarly for the “fingers crossed” experiment, I’m assuming that the researchers did not actually “press the thumbs” for the subjects, so the act of performing the superstitious rite is not the influence here. Rather, it is the subjects belief that the rite is being performed that is important.

  9. Stitcheroo said,

    June 12, 2010 at 10:38 am

    @Kerellen. My point is that you cannot measure performance in such a simplistic way. You may be right about the the subjects’ “belief” but there may be many more factors that are at play. The effect of belief and mental attitude in affecting performance is an interesting one but performance is affected by many factors (state of mind on the day, fatigue for example). If it were that simple, I would give my daughter a lucky charm for tennis and her mental attitude would be all sorted.

    Of course the other factor may be that “superstitious” people are more likely to give up more easily if their lucky charm is not with them. But, again, there is not enough information to show that this is actually what happens. There may be something happening here but we’re all guessing about why.

  10. Karellen said,

    June 12, 2010 at 11:19 am

    @Stitcheroo: “The effect of belief and mental attitude in affecting performance is an interesting one but performance is affected by many factors”

    Yes, but that is why you take multiple individuals and randomly assign them to the experimental or control group. That has the effect of evening out all the other factors (what are the chances that most of the fatigued people happen to get the “control” group?). So, as far as it is possible to measure, being told about whether the golf ball is lucky or not *is* the only significant difference between the members of the two groups. That’s what makes it a scientific study – you try very hard to account for and eliminate all the factors you’re not interested in, and measure results just on the one that you are.

    Of course, science is as much about attacking and disproving theories than it is about forming them (anyone can make up a random theory) – so if you’d like to attack theirs and suggest *some* actual possible way that there could be a significant difference between the two groups, given that their members were randomly assigned to reduce the possibility that, for example, all the keen golf players happened to get into the experimental group (rather than just saying, “oh, there could be many factors”, which is, if you’ll excuse me, a bit of a cop-out) then, honestly, please do.

    (Oh, and would any statisticians care to chime in with the likelyhood of these results being due to chance given the smallish sample size?)

  11. Dudeistan said,

    June 12, 2010 at 11:22 am

    Even as late as the early 20th century, people in my neck of the woods (rural Wales) could go to a local shrine and place a curse on anyone who had crossed them.

    Unfortunate recipients of a curse would suffer genuine distress simply because so many people were superstitious. Accounts of unfortunate events happening to the cursed were common place.

    Of course it was important for the cursed to be told they had been cursed. Nothing unfortunate appeared to happen to them if they were oblivious to the curse placed on them.

  12. Bishop Gillian Wakefield said,

    June 12, 2010 at 2:24 pm


    That’s 3 years old – although it would be interesting to see a follow-up revealing exactly how many kittens were drowned (although without a control group I expect it won’t reveal very much)!

    Anyway, thanks BG for reminding me of a great song. I can’t believe it’s not on my itunes; soon to be rectified…

  13. skyesteve said,

    June 12, 2010 at 2:31 pm

    Being on a life-supporting planet – what are the chances of that?! Thanks f**k for luck! Hey, that could be a bumper sticker for me – below the one that says “my other bumper sticker is funny”
    In all seriousness, I don’t think this study is really a study of superstition as much as, say, the value of being more confident and more positive in life. Such people tend to do better regardless of whether or not they walk under a ladder, smash a mirror, have a black cat cross their path, think England will win the World Cup…

  14. Stitcheroo said,

    June 12, 2010 at 3:02 pm

    @Karellen. My point wasn’t that the study didn’t show that there was an effect from being given the placebo good luck charm. It was that there was a conclusion to be made that “being superstitious improves performance”. This is simplistic; it is possible that you could get the same effect on performance by giving one group so-called “expert” advice versus another without it (while keeping the actual advice the same). The difference in results would not be explained by belief in superstition but be due to the same effect.

  15. kim said,

    June 12, 2010 at 3:10 pm

    Ben, it’s a shame you don’t like Malcolm Gladwell because I think there’s a lot of overlap in what you do (perhaps that’s *why* you don’t like him?) For example, in his latest collection, he has one essay where he explains why mammogram screening isn’t nearly as effective as people think it is, and another where he demolishes the art of criminal profiling, pointing out that criminal profilers use a lot of the cold reading techniques used by psychics.

    Someone once told me that the world’s most common superstition is a fear of tempting fate. Am sure this is true – even perfectly rational, non-superstitious people (hey, like me!) are frightened of tempting fate.

  16. Paulski said,

    June 12, 2010 at 3:20 pm

    As said above, I think that superstition works both ways – negatively and positively. My Mum is superstitious and she seems to be held back by her beliefs as much as she is spurred on, probably more so. I also think in general superstitious people are more switched on to negative superstitions than the positive alternatives. For example, people may cross their fingers for luck, but this won’t manifest itself as strongly as another negative superstition, such as being on plane number 13 or something.

    Anyway, I understand the crux of the article wasn’t whether they are on the whole good or bad, rather if they have any effect at all. I think that some suggestion of being in a positive situation has some positive effect is quite obvious. Surely it’s just a placebo effect? And also I think there are further problems with the test scenario. Saying that people have had some good fortune with a certain golf ball may not entirely be down to ‘luck,’ as the ball is directly involved with the game. A better ball will probably produce a better shot. Therefore, if I believe I am using a better ball (although the ball is in reality no different) I am likely to be more confident in my abilities, and we know how a positive attitude will often produce a positive outcome. And this is a placebo, isn’t it? They should have used an element outside of the game instead as the point of superstition.

    In the same way, being told that someone is crossing their fingers for me may produce a similar confidence in my ability. As an example, compare how football teams almost always perform better in front of a home crowd than an away crowd.

  17. Mr Skills said,

    June 12, 2010 at 3:29 pm


    “another poor quality psychology study. Who would have though… I’m still with Feynman in that it’s a Cargo Cult science”

    I agree that this is a bit of a ropey-looking study, but to damn the whole of psychology on that basis is like damning the whole physics on the basis of the miriad “toast landing butter side down” experiments that are so beloved of the press, or to damn all dieticians because you’ve watched “you are what you eat”.

    I’ve just been off to read what Richard Fenyman said on the matter, and it’s highly entertaining, but some of his cherry-picked examples are poor. For example, he suggests that lack of improvement in education and criminality show that research in these areas has failed. But he is making the huge assumption that scientific knowledge  in these politically charged spheres is ever acted upon without massive political interference – or even that the research itself is
    allowed to be conducted properly by the powers of the day.  He then gives us a
    couple of anecdotes – one of which is about a single tutor refusing a piece of research back in 1947!

    To be fair, there is plenty of bad science that goes on in the name of “psychology”, and it’s a shame that it tends to be the more zany stuff that is more likely to get into the national press. But to extrapolate that to the whole of such a huge field is doing a disservice to the dedicated scientists who work day in and day out doing the more important (and less sexy) research, and doing it with exactly the sort of genuine scientific integrity that Fenyman advocated.

    I discovered Ben Goldacre’s work precisely because I am passionate about that sort of integrity and honesty in science, and I learned that integrity in the first place because it was drilled into me while doing a psychology degree in the late 90s. He laments that this sort of integrity is something that you have to ‘pick up’ because it is not specifically taught – well it was specifically taught to us. And the idea that we we would not have been allowed to repeat an experiment is laughable – the importance of this was drilled into us at every opportunity.

    To be fair to Fenyman, back in the period he was talking about, psychology was such a new field that there was some pretty dreadful work that went on. We spent a long time being taught about research in the early 20th century and how it tended to be influenced by the dogmas of the day. These things are a particular danger in psychology, but they are a danger that scientists in the field rare well aware of, and I doubt you will find a more self-critical bunch than those conducting proper psychological research. Just remember that they are not the ones who get on TV.

  18. H. E. Pennypacker said,

    June 12, 2010 at 4:09 pm

    I’m very pleased that Ben has highlighted interesting work from experimental psychology, you don’t see that often in the media (unless pretty but often meaningless pictures of the brain are involved).

    Just to address a few of the comments here:

    Stitcheroo thinks the effects are due to increased confidence. Sure, that’s very much possible. But the point is that here confidence was increased by superstitious thinking. You may be right that expert advice would have a similar effect. But what’s remarkable is that the mere presence of a lucky charm would have the same effect as expert advice. Relying on expert advice is rational, relying on a lucky charm is irrational. We see here that people are surprisingly irrational.

    Paulski: we can call it a placebo effect if we want to, in fact that’s the whole point of the experiments. A manipulation which should not have an effect (since we’re changing someone’s superstitious beliefs rather than physically making a change in the odds of success or failure) has an effect due to psychological reasons alone. It doesn’t make it any less interesting, quite the opposite. Your point about the golf ball is well made, but the fourth experiment really shows that there’s more going on than that. Lastly, your point about boosting confidence is ok, but again the interesting issue is what causes that boost to take place: the authors showed it can happen purely as a consequence of superstitious thinking.

    Skyesteve: the confidence thing again, see above.

    Karellen: Sure, the sample sizes in all four experiments are smallish, like you say. However, the fact the the effect was repeated four times in different groups of participants in different experimental settings is powerful evidence for the effect being statistically robust. Replicating a finding several times in different groups is always more impressive than just throwing more participants at a single experiment. The latter strategy can make even small clinically insignificant effects look statistically significant. Hence I’m not worried about the sample size here.

  19. Sqk said,

    June 12, 2010 at 5:01 pm

    Shame I can’t get a copy of the article… If we leave aside the sheer number of holes that could be picked in it and the study I have another question: What are the students actually students OF?

    – Unless each group was equally divided by subject that would have skewed things nicely anyway. Someone from a subject requiring physical agility may do better at tasks requiring physical co-ordination anyway.
    The most obvious example of subject specific study (actually a research assignment to drum this in to future pharmacists) I can think of off the top of my head is the exclusion of pharmacy students from a study about the retention of pharmacy information. It had to be information that the subject was new to, to test the retention. However they discovered the hard way that inclusion of PhD students from one particular completely unrelated subject had exactly the same effect as they too could remember facts and figures in the same way.

    – I just ask as a matter of interest as I’d like to think in the future when I’m, say, walking across a bridge, that the person who designed it believed in getting the maths right above having a lucky streak.

  20. Jut said,

    June 12, 2010 at 9:32 pm

    @Mr Skills
    Yes I know. I was being grouchy, although Psychology still needs to do a whole lot of growing up.
    Maybe the journals need to be a little more selective about what exactly they publish.
    Let’s face it, an experiment this poor is the kind of thing you would expect from undergrads, and to have it published with such a definitive title is asking for flack.

  21. H. E. Pennypacker said,

    June 12, 2010 at 10:52 pm

    Jut: can you elaborate on what you view as the crucial weaknesses of these experiments? The title of the paper by the way (Keep Your Fingers Crossed!: How Superstition Improves Performance) wasn’t particularly offensive to me…

  22. Timmy said,

    June 13, 2010 at 2:10 am

    As I have always said,,,,Voodoo only works on those who believe in it

  23. Rogue Medic said,

    June 13, 2010 at 2:15 am

    H. E. Pennypacker,

    “But what’s remarkable is that the mere presence of a lucky charm would have the same effect as expert advice. Relying on expert advice is rational, relying on a lucky charm is irrational. We see here that people are surprisingly irrational.”

    Many do have an irrational faith in the ability of an “expert” to do something, but we have very little understanding of what expertise is, or how to evaluate it.

    Nassim Nicholas Taleb wrote about this in “Fooled By Randomness.” He focuses on economics, but behavioral economics combines economics and psychology.

    I think that the study of “expert” advice is a good idea.

  24. Rogue Medic said,

    June 13, 2010 at 2:28 am

    I think that the biggest failure of this study, based only on what was written in this post, is the lack of an evaluation of the negative effects of superstition. This has been mentioned in other comments, but it would be relatively easy to include an arm evaluating the negative effects.

    “This has been a lucky ball,” vs. “Here is your ball,” vs. some comment like “This ball has been giving everyone problems,” or “I have a bad feeling about this,” or some other negative superstitious stimulus.

    Viewing only the positives of superstition is not really much different from only reporting the benefits from a drug trial. There may be significant benefits, but if the harm is even more significant, it would be irresponsible to praise this as something good. For example, Thalidomide was good at relieving “morning sickness,” but the side effects were not noticed until later. The side effects made any benefit seem insignificant. This is an extreme example, but there are plenty of others.

    All drug prescribing is supposed to be a balancing of the benefits and harms of a particular drug, or drug combination (since few seem to take only one drug at a time). The idea that there can be a treatment that has not possibility of harm (First, do no harm) is an extremely dangerous delusion. Avoiding evaluating the negatives of any intervention is not good science. It is appropriate for this study to be covered here.

    There is good psychology research. As long as the scientific method is applied in an unbiased fashion, and all variables that can be controlled for are controlled for, and the limitations are thoroughly and accurately discussed, I do not see a problem with this research. The problem is with people inappropriately reading too much into the limited information that is appropriate to draw from the research.

    We look for explanations. Even though there may be dozens of explanations that fit the results of a study, we seem to feel the need to choose just one as “the explanation.” Perhaps it is the one that most closely matches our previous beliefs. This is no less a superstition, than having a lucky charm.

    “Science is a way of trying not to fool yourself. The first principle is that you must not fool yourself, and you are the easiest person to fool.” – Richard Feynman.

  25. sciencerocks said,

    June 13, 2010 at 11:25 am

    As someone who works in psychophysics (one of the earliest, if not the earliest, form of experimental psychology), I get a bit cross when general statements are made about psychology’s lack of scientificness. Psychophysics is particular beautiful because it is founded on the idea of developing *quantitative* models to explain observations.

    Feynman is one of my science heroes and I’ve read loads of his stuff over the years. I am confident his cargo cult science comments have been over-generalised. After all, he did some psychology research himself!

  26. anon3455 said,

    June 13, 2010 at 4:11 pm

    I’d say the trust in superstition partially activates the sort of state that is associated with excelling in sports, video games, etc. You know, the sort of (more or less) trance-like state of confidence, concentration, determination and relaxation (“looseness”) that probably quenches some second-guessing from motor functions and improves muscle memory recall etc.

    I think something similar happens in certain modes of concentrated thinking (eg. programming “in the zone”).

  27. Michael Grayer said,

    June 13, 2010 at 6:15 pm

    Sorry to drag the football into this column, but I couldn’t resist adding to the “negative superstition” comments…

    Although it’s based on a sample size of 1, I wonder if “negative superstitions” among goalkeepers regarding the ball that’s used for the world cup ( had any influence on Robert Green’s mishap last night?

  28. Michael Grayer said,

    June 13, 2010 at 6:21 pm

    Sorry to drag the football into the comments but I couldn’t resist. And it is relevant to the “negative superstition” comments.

    I wonder if negative superstitions surrounding the ball used for the World Cup ( played any role in Robert Green’s mishap last night. Sample size of 1, I realise, so not very scientific, but still…

  29. Michael Grayer said,

    June 13, 2010 at 6:23 pm

    Sorry. Browser failing to post properly and then failing to refresh the page made me think that my first attempt to post that point hadn’t worked so I retyped it. Gah.

  30. pauldwaite said,

    June 13, 2010 at 10:25 pm

    Wait. You’re not Malcolm Gladwell?

  31. BigEoinO said,

    June 14, 2010 at 9:24 am

    @pauldwaite, Ben did say “I am not Malcolm Gladwell”
    But, with their matching wild hair, I can understand your confusion…. so maybe they are brothers separated at birth?

  32. smu95rp said,

    June 14, 2010 at 12:57 pm

    Think you may have got your shoulders mixed up, Ben.

  33. themonk13 said,

    June 14, 2010 at 3:19 pm

    Me thinks that Doctor Ben’s tongue was so far in his cheek, while writing about the psych article in question, that the dear Doctor will be… lucky… if it does not require the attention of a surgeon to extract the tongue. That being said, superstition is a great song. Whether done by Stevie W., or covered by Stevie Ray V., the song is far more memorable than the experiments. Now it is back to doing my…Drunkard’s Walk 😉

  34. judefawley said,

    June 14, 2010 at 4:23 pm

    Not on topic in anoything more than a more general sense but… make a bid against pseudoscience by bidding that Bristol City Council do not fund a naturopathy project:

  35. Sqk said,

    June 14, 2010 at 4:27 pm

    Well it’s the “…he may one day go nuclear and take out three rows of medical correspondents with a single lungful of biting sarcasm.” that I’m waiting for. I get described as ‘angry’ periodically and look forward to learning from the Master. As someone who isn’t a medical correspondent I wouldn’t be there to see it and instead would have to rely on the noise, cloud and write-up for how it went, but I’m sure I’d hear about it soon enough.

    Though, given the context of the comment, I suspect that there’s a school of thought that feels that a little thinning of the numbers may be good thing and would teach the fourth row to pay better attention.

  36. fontwell said,

    June 14, 2010 at 6:04 pm

    These superstition studies are obviously restricting themselves to things that are matters of judgement and therefore inherently dependant on state of mind, so this result is hardly surprising.

    Superstition won’t help you pick the winning lottery numbers or substitute for proper study before an exam.

  37. H. E. Pennypacker said,

    June 14, 2010 at 10:00 pm

    fontwell: How is putting a golf ball, or carrying out a manual dexterity task, or solving anagrams “things that are matters of judgement”? The participant is not required to make judgements in these tasks, and their performance can be objectively measured. A lot of people in these comments are quick to offer vague hand-wavy criticisms (not the way to proceed in science by the way, we value precision), but few appear to be able to elaborate on them.

  38. Stitcheroo said,

    June 15, 2010 at 7:27 am

    Hold on. The first trial was of 28 students, 23 of whom believed in luck. They only did 10 putts of a golf ball and achieved, on average, less than 6 and a half when told they had a lucky ball against 4.75 for those without. I wouldn’t say that was significant. Certainly not to conclude that there was an effect caused by a belief in superstition. It is possible that the 5 who did not believe in luck brought the average up and what would that tell us? Let’s forget that test.
    On the other tests there is clearly something going on in terms of confidence but not necessarily linked only to superstitious belief as such. It is possible that the ones told that they could keep their lucky charms felt more in control and self confident because they had something familiar with them. I would feel more confident in exams if I had a family picture in front of me and that is nothing to do with my belief in superstition.
    Similarly, the ones who had researchers say they crossed fingers could be taking comfort from the fact that someone cared how they did.
    I agree that these are not matters of judgement (fontwell has missed the point of the experiment I think when they talk about lottery and exam performance) and there could be some interesting things that come out of this type of research. However, to draw conclusions about superstition and performance we need to see all of the factors that might actually be affecting that performance.

  39. Rogue Medic said,

    June 15, 2010 at 9:00 am

    H.E. Pennypacker,

    “fontwell: How is putting a golf ball, or carrying out a manual dexterity task, or solving anagrams ‘things that are matters of judgement’?”

    I think if you change “judgment” to “state of mind,” you might agree. Although, I think that taking a test is affected by state of mind. For some people, their state of mind may be more important than the information they have learned.

    The suggestion of picking a lottery number appears to be a good way of eliminating the effect of state of mind. There should be no demonstrable difference in the results, but this would require a much larger sample size.

    Something with far fewer variables, such as picking the color (or number) in roulette, would be one way of removing the influence of the subject. The use of a lucky charm vs. the use of nothing vs. the use of nothing and being told that they are being affected by bad luck. There should not be any effect of luck.

  40. kneecap said,

    June 15, 2010 at 11:32 am

    It’s always good to hear a positive review of superstition, especially when carried in a (rightly) sceptical column.

    I am, personally, a “supersitious” person. Although I don’t practice any of the more outrageous supersitions such as “lucky charm”. I do, however, practice the science of Numerlogical Focusing. So you can imagine my joy when, with my Personal Key Numerlogical Focus Number of “6”, I see in June (6th Month) the 12th (6 x 2) and column with exactly 12 (6 x 2, again) paragraphs which supports supersition. I can’t wait to tell more people how this proves that Numerlogical Focusing works.

    Thank you Ben for a great article.

    Background: Numerlogical Focusing is a science based around finding “Key Focus Numbers” (sometimes called “Lucky” by lay-people) based upon date, time and location of birth. “Key Focus Numbers” (KFNs) are calculated as “natural” numbers in the range 1 to 13 inclusive*. Focusing on this number, or multiples thereof, will allow you to concentrate your energies at times when it will be of particular effecacy. e.g. a person with a KFN of 6 will be better concentrating their efforts for success in June (the 6th Month) or at 6am or 6pm (the effect of KFNs are not am/pm specific)

    ( *although there are 13 star signs, including Ophiuchyus, KFNs are not connected with outlandish superstitions such as astrology [which has never been scientifically proven]. The 13s in this case are just a co-incidence)

  41. Thomas King said,

    June 15, 2010 at 12:13 pm

    as any good general/rugby coach will say, morale is everything. if my rugby coach knew there was something he could do to make us believe we would win every week (even if it had no “real” effect) he would do it. when we go into matches believeing we are going to lose, we lose.

  42. Guy said,

    June 15, 2010 at 2:48 pm

    I do assume that “I do, however, practice the science of Numerlogical Focusing.” is a joke??

  43. kneecap said,

    June 15, 2010 at 10:16 pm


    Ah, KFN-7, “the doubter”. I’ll have you know you Numerological Focusing is a true science based on ancient principles. I have learnt of its … oh, you’re right, I just made it up. I wanted to start a new pseudo science and make a load of money without having to work hard. I had tried the Doctrine of Signatures but didn’t get taken seriously when I told people I could cure them of leaves. I still think Numerological Focusing has legs (KFN-11)

  44. Mark Lapierre said,

    June 16, 2010 at 12:55 am

    @Sticheroo: A difference between 6.42 and 4.75 is a significant difference (according to a t test with a sample size of 26, at p < .05, Cohen's d = 0.83) when the standard deviations (1.88 and 2.15 respectively) are taken into account.

    There were not 23 students who believed in luck. The paper mentions that 80% of the participant *population* believed in good luck (as revealed by a pretest), not 80% of the *sample*. But supposing you were right and 23 believed in luck while 5 didn't. And suppose further that the 5 did perform better, and suppose again that they all somehow ended up in the group which performed better, and let's go the whole way and also suppose that if those 5 participants' scores were removed the significant difference would disappear. That's a hell of a lot of supposing. You can speculate that that's the case, but unless you show it to be the case you are not justified in dismissing the results here. Particularly in light of all the other significant results.

    You say we need to see all the factors that might be affecting performance. You ask the impossible, particularly where individuals are concerned, with our many and varied motivations and capacities. But rather than attempting the impossible, what we do, and what the researchers here have shown, is to attack a problem from multiple angles, with multiple, random groups of participants. H. E. Pennypacker has already made this point. When the same effect is seen many times, in different but related ways, with many groups of people randomly selected, it is reasonable to conclude that all those factors you want to be controlled are not having a significant effect. Except if one or more of those factors could be shown to be conflated with superstition in *each* of the 4 experiments.

    So, is there anything in particular which may account for each of the results which the researchers have allegedly confused with superstition causing increased self-efficacy?

  45. Mark Lapierre said,

    June 16, 2010 at 1:25 am

    @Rogue Medic: The researchers do briefly discuss negative effects of superstition in their conclusion. They acknowledge that negative superstitions could have adverse effects, and that it would be interesting to study. However that was not their chosen focus – performance improvement was (they justify all of that, of course). But I do agree that an evaluation of the negative effects of superstition would be worthwhile. For example, it’s possible that if someone attributes too much of their performance to luck, their intrinsic motivation will suffer, let alone their performance when their lucky charm goes missing. Likewise the belief in the contribution of luck might limit an individual’s potential – if they believe a lucky charm helps them perform better, they might not strive to reach the same level of performance without the lucky charm.

  46. H. E. Pennypacker said,

    June 16, 2010 at 10:52 am

    Thanks to Sticheroo and Rogue Medic for thoughtful posts. A couple of quick comments:

    Stitcheroo: I like in principle your idea of controlling for possesion of familiar objects. However, if it turned out that being in possesion of a family photo was equally helpful to being in possesion of a lucky charm, this would suggest that a lucky charm (associated with superstitious belief) has influence equal to a family photo (associated with emotional value?), and would extend the conclusions put forward by the authors, not undermine them.

    Rogue Medic: I’m not sure why you’d want to eliminate the effects of state of mind in a study which is specifically interested in manipulating state of mind. Recall that this study shows that superstitious belief influences state of mind to such a degree as to affect various types of cognitive and motor performance.

  47. merrick said,

    June 16, 2010 at 1:32 pm

    That’s all well and good, but the version of Superstition you want is live on Sesame Street, it way outfunks the high levels of funkaciousness the studio version achieves.

  48. merrick said,

    June 16, 2010 at 1:33 pm

  49. Rogue Medic said,

    June 18, 2010 at 5:24 am

    Mark Lapierre,

    “The researchers do briefly discuss negative effects of superstition in their conclusion. They acknowledge that negative superstitions could have adverse effects, and that it would be interesting to study. However that was not their chosen focus”

    I have not read the paper. I caddied in high school, so I think more of the negative performance effect of any psychological stimulus in golf, than the positive. It takes so little for a good game to fall apart. Think of Greg Norman as described by Malcolm Gladwell.

    “The Art of Failure
    Why some people choke and others panic.”

  50. Rogue Medic said,

    June 18, 2010 at 5:34 am

    H. E. Pennypacker,

    “Rogue Medic: I’m not sure why you’d want to eliminate the effects of state of mind in a study which is specifically interested in manipulating state of mind. Recall that this study shows that superstitious belief influences state of mind to such a degree as to affect various types of cognitive and motor performance.”

    I like to control for variables as much as possible. As with homeopathy, full moon effects, and other superstitions, I think it is important to make clear that this is purely a placebo effect.

    Too many people already believe in these things. I wouldn’t want a study that could be easily misinterpreted as suggesting there is any kind of non-placebo effect of superstitions.

  51. uggi said,

    June 18, 2010 at 1:01 pm

    This doesn’t actually sound like superstition – more like the effect of encouragement, already documented in other studies. The idea that someone else wants you to do well is inspiring, and makes you try harder.

    As for good luck charms… well, that’s just a different shaped teddy bear, really, isn’t it? ;p

  52. Stitcheroo said,

    June 19, 2010 at 8:53 am

    Thank you uggi.

  53. irishaxeman said,

    June 21, 2010 at 12:59 pm

    We have to assume none of the tiny sample were golfers – as a reasonable golfer would expect to hole more than 8 out of 10 from a metre away on a flat plane. What was not controlled for as far as I can see (assuming putter is the same as is the ball throughout) is grip/face/target alignment, itself a product of skill and visuomotor ability. With such a small sample you only need a small bias in these to produce a significant result, not withstanding different levels of task anxiety and pre-existing history of self efficacy. Was there standardisation of these factors? Were there practice trials?
    However I will note that female golfers of my acquaintance are known to accuse golf balls of bias (as in ‘you naughty ball’).

  54. Stitcheroo said,

    June 22, 2010 at 8:27 am

    @irishaxeman give some of the reasons to exclude this test. The possibility of a few out of the sample (of 14, not 28) being golfers would skew the results over such a small number of shots. To then say that the results are confirmed by “all the other significant results” is a leap of faith. In any case, just saying “you believe in luck” does not mean that you believe that the ball is lucky. I play sports and you may say you have a lucky ball or perform some ritual but this is short lived even when affecting performance. When it actually comes to it you forget about that and concentrate on technique or the opposition. Rafael Nadal is a good example, he has rituals at the start of the match that get him in to the right frame of mind (call it superstition if you will) and this ensures that he gets off to a good start. Even if he says he believes in “good luck”, he really knows that the rituals are there to get him focussed to enable him to concentrate on the game. As I said before, there are many factors that come into play.
    And irishaxeman, I’ve played golf with some male golfers who have some pretty weird rituals or were you just trying to be provocative?

  55. irishaxeman said,

    June 22, 2010 at 11:51 am

    Luck is merely a description of unplanned consequences, and as such ‘exists’. Superstition has been adequately explained by behavioural and evolutionary psychology.
    And bad design is bad design.
    No I have never heard a male golfer call a golf ball naughty. I have heard many males use other descriptions, often ending in ‘’.

  56. spotthelemon said,

    June 22, 2010 at 12:01 pm

    The most suprising thing about this research is that anyone would be remotely surprised by it. It’s essentially a variation on the placebo effect but with lucky charms (or procedures) instead of sugar pills.

  57. H. E. Pennypacker said,

    June 23, 2010 at 10:27 am


    Yes, we just have to assume that those people who did better in the golf task aren’t golfers. I think it’s a relatively safe assumption if we’re willing to assume that the people doing this research aren’t stupid and actually have some training in how to do research. I don’t know them so they might be stupid and have no training in research and accidentally assigned a lot of golfers to the experimental condition. But it’s quite a safe bet that they didn’t. On a serious note, one way they could have eliminated this issue would have been to use the same participants in the lucky ball and neutral conditions, but that solution would have had its own complications.


    You said: “To then say that the results are confirmed by ‘all the other significant results’ is a leap of faith”. I have to disagree with that quite strongly. Replication is a foundation of experimentation. The golf experiment on its own would not have been publishable, even the authors treat it as a preliminary finding. They then do variations on the experiment showing the same effect in each. This is the exact opposite of taking a leap of faith, it is to subject a hypothesis to systematic inquiry and is the gold standard in any field of science.

    As a disclaimer I should say that I’m not affiliated with this research in any way, and know nothing about research into superstition and have no strong opinions on that literature. I’m just trying to defend what seems on the surface like a relatively solid piece of research against some far fetched critique.

  58. thetallerpaul said,

    June 30, 2010 at 8:36 pm

    Methodological issues aside this seems like a lot of fuss over a nothing study about the placebo effect of encouragement. Wherever this comes from ( external, superstition, replicable statistical proof) it helps people perform. If you are superstitious keep doing what you need to do as long as you understand it’s not magic. An interesting test would be with a group of ‘rational’ people who were told that statistically you have a very small or high chance of success in the test exercise. My hypothesis is that the ‘rational’ respond in a similar way to encouragement or discouragement.

  59. Stella said,

    July 10, 2010 at 2:47 am

    Why does superstition have a deja vu quality about it? It seems that people talk about superstition from their predefined and possibly prejudiced point of view, pushing out opinions without taking in new information. It is almost a defense reaction.

    It is natural for humans to be superstitious. For survival it is better to believe something bad is coming down and try and escape or defend yourself. If you are wrong you may seem silly, but if the crackling twig is a predatory creature then you are right. Being wrong in interpreting situations is far worse than ignoring a suspicion that cannot be supported by facts.

    Recognising a superstitious response is helpful and should allow people to teach thme to understand and learn from their instinctive behaviours and not try to ignore them.

    In the test conditions provided the subjects were put into a competitive situation. The instinct would be to stay and compete where you saw that you had an advantage, or to hide or withdraw where you perceived your chances were reduced. I think the tests were valid and maybe need expanding or more interpretation.

  60. jjad said,

    July 12, 2010 at 1:39 pm

    So the lucky charm trick works? Yes, but the reason could be because people who have their precious charms taken from them to be ‘photographed’ may be worried about what is happening to it. Where has it gone? Will it be OK? What if it gets lost? This is a possible reason why these people felt less confident and did worse than those with their charms.

  61. cw37209 said,

    December 22, 2010 at 12:22 pm

    “…superstition works, because it improves confidence, let’s you set higher goals, and encourages you to work harder.”

    To be most credible, scientists should not only be right, scientifically, but should be well-written. That said, a gentle reminder: the verb “lets” should not include an apostrophe.