Let us stipulate that there is no magic. Sleight-of-hand, deception, illusion, and conjuring, yes, but no “real” magic. On this, most science-minded people agree. But when it comes to superstition, there has always been an additional, less obvious question. Of course, superstitions do not have a magical effect on the world, but do they have psychological benefits? Could superstitions make difficult situations easier to handle? Furthermore, if they have an emotional or psychological benefit, could they also produce better performance in situations where skill is involved? The psychological benefits of superstitions—if they exist—would not be expected to change your luck at the roulette wheel, but perhaps an actor’s pre-performance ritual could reduce anxiety, allowing for better acting.
Despite several decades of research on superstition, these questions remained unanswered for many years. Most researchers assumed superstitions were irrational and focused their attentions on discovering why people were superstitious. It was often assumed that there might be some direct psychological benefits of superstition, but these were rarely studied.
Then in 2010 there was a great advance—or so it seemed. Researchers at the University of Cologne in Germany conducted the now famous golf ball study (Damisch et al. 2010). Participants were given a putter and asked to hit a golf ball into a cup on the carpet of a laboratory. Half the participants were handed a ball and told, “This ball has been lucky today.” The other half were told, “This is your ball.” As it turned out, more than 80 percent of the German participants reported believing in the concept of good luck, and when the results were tallied, the researchers discovered that participants in the lucky ball group sank significantly more of their putts than the other group. Furthermore, Damisch et al. replicated this result with different tasks and several different luck-activating superstitions. Of course, there still was no magic, but these studies seemed to have demonstrated that believing in luck gave participants the confidence to perform better than they otherwise would. A phenomenon long speculated to be a possibility had finally been demonstrated in a laboratory setting.
Except there was a catch. As I reported in my January 2017 online column (“Your Unlearning Report: The Trouble with Empathy, Implicit Bias, and Believing in Luck,” available at https://tinyurl.com/yc9nm8vu), a group of researchers at Dominion University in Illinois conducted a replication of the Damisch et al. study in 2014 and found no luck-enhancing effect on putting (Calin-Jageman and Caldwell 2014). Furthermore, the 2014 study included over three times as many golfers and was a pre-registered study—meaning that the design and methods of the study were publicly posted prior the start of data collection. The Dominion study was much more thorough and scientifically sound, and it came up empty. So, at least with respect to the effect of luck on putting performance, the jury is still out.
What about Rituals?
Some topics go unstudied for many years until someone comes up with a clever way to conduct the right kind of experiment. Such was the case when it came to rituals. The use of rituals in moments of trouble seems to be both timeless and universal. Consider the Jewish practice of sitting shiva and the lengthy mourning ritual that follows. On the surface, this and other rituals of burial and mourning appear to have a comforting value for the bereaved, but how would anyone test that hypothesis? It would be unethical to create a similar loss in the laboratory, and even if you could overcome that obstacle, how would you choose the right ritual to use?
In 2014, two researchers at Harvard Business School—yes, Business School—found a way. Michael I. Norton and Francesca Gino (2014) did a series of experiments that looked at the role of rituals in coping with loss. To create a sense of loss, they invited people into the laboratory in groups of nine to fifteen and held a drawing for a $200 cash prize. The person who won the prize got to take the money and leave early, and the rest stayed for the coping-with-loss part of the experiment.
To look at the effect of rituals, Norton and Gino created an arbitrary ritual for people to perform in the lab. After losing the lottery, participants were each placed in a separate cubicle and told that “Previous research has found that people often engage in rituals after a loss.” Then they were asked to follow the procedure below.
Step 1: Please draw how you currently feel on the piece of paper on your desk for two minutes; Step 2. Please sprinkle a pinch of salt on the paper with your drawing; Step 3. Please tear up the piece of paper; Step 4. Now please count up to ten in your head five times; Step 5. You have now completed this task.
Other participants were given the same “Previous research has shown …” information about rituals but not asked to perform one. Still others were given a similar instruction: “Previous research has found that people often sit in silence after a loss” and were asked to sit in silence in their cubicles.
The primary findings of the research were that performing the ritual increased feelings of control and reduced the negative emotions of loss. Sitting quietly or simply being told that some people engage in rituals was not enough. Only actual performance of the ritual produced the effect.
Once Norton and Gino had paved the way on rituals and loss, it was not long before others adapted their procedures to other uses of rituals. Furthermore, Norton and Gino had pointed out that the ritual they devised involved tearing up a picture of “how you currently feel,” a feature that gave the actions a kind of symbolic meaning. The implications for superstition and other performance enhancing rituals were not hard to see. The results with lucky golf balls had been murky, but if symbolic rituals were effective in coping with loss, perhaps they could be useful in other situations, too.
In 2016, Allison Wood Brooks, also of the Harvard Business School, along with colleagues from four other institutions, published a study called “Don’t Stop Believing: Rituals Improve Performance by Decreasing Anxiety” (Brooks et al. 2016). Rather than create a situation of loss, Brooks and her colleagues brought people into the laboratory and confronted them with a number of potentially anxiety-provoking tasks: taking a difficult, timed math test or singing the song “Don’t Stop Believing” by Journey in front of other participants.
Once the participants knew what they would have to do, Brooks and colleagues split people into ritual or non-ritual groups. Prior to the anxiety-provoking task, the ritual group performed a sequence of activities very similar to the one used by Norton and Gino in their 2014 study:
Please do the following ritual: Draw a picture of how you are feeling right now. Sprinkle salt on your drawing. Count up to five out loud. Crinkle up your paper. Throw your paper in the trash.
In a series of experiments, Brooks et al. showed that participants who performed this ritual did better on the task—high-pressure math problems or singing—than those who did not. Furthermore, they were able to show that the effect was mediated by a reduction in anxiety. So performing a symbolic ritual prior to a high-anxiety task reduced anxiety, which in turn produced better performance. The experimenters tested a number of controls to determine what was behind this anxiety-reducing effect. For example, in both the Norton and Gino study and some of the Brooks et al. experiments, the ritual involved the participants drawing a picture of their current feelings, after which they would rip or crinkle it up and throw away the drawing. To determine whether this was an important part of the formula, Brooks et al. created a new sequence that removed the artistic expression of emotion and the symbolic destroying/discarding of the emotions:
Please count out loud slowly up to 10 from 0, then count back down to 0. You should say each number out loud and write each number on the piece of paper in front of you as you say it. You may use the entire paper. Sprinkle salt on your paper. Crinkle up your paper. Throw your paper in the trash. (Brooks et al. 2016, p. 80)
For one group of participants, this new sequence was described as “random behaviors,” and for another it was described as a “ritual.” Finally, a third group did not perform the sequence of actions at all. The results showed that the ritual group had lower anxiety and performed significantly better on a timed math test than either the random behaviors group or the no ritual group. The authors suggested that merely calling the sequence a “ritual” was sufficient to give it the necessary symbolic function to reduce anxiety and increase performance. Brooks et al. did not find that participants had an increased sense of control, which was surprising because a desire for control has often been cited as a motivation for superstitious behavior (e.g., Hamerman and Johar 2013). Instead, the effect on performance was entirely due to reduced anxiety.
So What Does This All Mean?
Does this mean that superstitious rituals work? Yes and no. The Brooks et al. study suggests that superstitious rituals do work—not because they are superstitious but because they are rituals. Any old ritual will do, including writing numbers on a piece of paper, crinkling it up, and throwing it away. But it appears to be important that the sequence of actions be defined as a ritual and not as random behaviors. So there is no real magic, but there is a bit of calming magic in performing a ritualistic sequence before attempting a high-pressure activity.
As a prelude to conducting the experiments in their study, Brooks and colleagues surveyed 400 people with the following question: “Think about a time when you faced a difficult task and you felt anxious about it (e.g., a test, a sport competition, an interview). Did you engage in a ritual before performing the task? (YES/NO).”
Forty-six percent of participants answered “yes.” Interestingly, the rituals these people reported using were overwhelmingly nonreligious and absent of references to luck or superstition. When Brooks et al. went on to do their laboratory studies, they chose rituals that were not expressly superstitious, and yet the rituals “worked.” The minimum amount of symbolism required was the “ritual” label. Furthermore, there is some evidence that rituals work even if you don’t believe in them. In their study of rituals and loss, Norton and Gino (2014, Experiment 2) asked people if they had used rituals in the past and whether they believed “performing rituals influenced the way people feel (e.g., more calm, less sad).” Participants’ responses to these questions were completely unrelated to whether the rituals they performed in the study worked or not. You don’t have to believe in the efficacy of a ritual for it to help you feel better. All of these studies are preliminary, and it will be important to see whether they hold up when other researchers try to reproduce the results. Furthermore, there is much more we need to know about why and how rituals work. But these early findings are quite interesting.
Any Old Ritual Will Do
For skeptics who would like to discourage superstitious and irrational thinking, this line of research has both a downside and an upside. The downside is that the research by Brooks et al. suggests that superstitious rituals do work—not because they are magic but because they are rituals. As a result, the calming features of superstitious rituals and the improved performance they engender are likely to sustain superstitious thinking. The superstitious person’s beliefs will appear to be validated. The upside, however, is that skeptics now have a ready response to those who claim their superstitions work: Yes, your superstitions work, but it’s the ritual, not the superstition that’s making you feel better. Any old ritual will do.
- Brooks, Alison, Juliana Schroeder, Jane Risen, et al. 2016. Don’t stop believing: Rituals improve performance by decreasing anxiety. Organizational Behavior and Human Decision Processes 137: 71–85.
- Calin-Jageman, Robert J., and Tracy L. Caldwell. 2014. Replication of the superstition and performance study by Damisch, Stoberock, and Mussweiler (2010). Social Psychology 45(3): 239–45.
- Damisch, Lysann, Barbara Stoberock, and Thomas Mussweiler. 2010. Keep your fingers crossed! How superstition improves performance. Psychological Science 21(7): 1014–20. Available online at http://www.ncbi.nlm.nih.gov/pubmed/20511389.
- Hamerman, Eric J., and Gita V. Johar. 2013. Conditioned superstition: Desire for control and consumer brand preferences. Journal of Consumer Research 40(3): 428–43.
- Norton, Michael I., and Francesca Gino. 2014. Rituals alleviate grieving for loved ones, lovers, and lotteries. Journal of Experimental Psychology: General 143(1): 266–72.