Psychology Today journalist Matthew Hutson covers some fascinating experiments just published in this week’s Science that found that reducing participants’ control increase the tendency for magical thinking and the perception of illusory meaning in random or patternless visual scenes.
Hutson covers all six experiments, but here’s a sample from his article which should give you the general idea:
In the fourth study, people who recalled a situation where they lacked control were more likely to see nonexistent images in snowy pictures and were also more likely to suspect conspiracies in ambiguous vignettes. (In one story, three local construction companies raise their prices after their owners all spend the same weekend at one bed and breakfast. In another, the protagonist was denied a promotion right after his boss and a workmate exchanged a flurry of emails.)
The fifth experiment showed that describing the stock market as volatile (versus stable) renders people more likely to spot false correlations in reports on company financials—and then make stock investments based on their unfounded conclusions.
Finally, the sixth study showed that feeling good about yourself reduces the frantic grasping for straws. There were three groups. One group recalled not having control, another recalled not having control and then performed a self-affirmation task, and a third group did neither. The first group saw more figures in snowy pictures and perceived more conspiracies than the other groups did. Apparently, increasing self-esteem fosters a sense of control over one’s life and reduces the need to seek additional stability in random noise.
Two of the ‘snowy pictures’ are shown on the right. The one on the top is completely random, the other has an embedded picture.
This is particularly interesting to me, because one of my own studies I completed with some colleagues in Cardiff also involved getting participants to perceive images in random visual patterns.
We did something a little different though, in that we didn’t have any hidden images, so every time someone saw something we knew it was illusory.
However, we also managed to alter how often people saw the images, but we used electromagnets (a technique called TMS) to alter the function of the temporal lobes which have been previously thought to be involved in the magical thinking spectrum – from everyday examples to diagnosable psychosis.
This study was inspired by an earlier study by neuroscientist Peter Brugger, who found that people who professed a belief in ESP (‘telepathy’) were more likely to see meaningful patterns in visual noise than those that didn’t.
Both the new study and our study are interesting because they show how this type of magical thinking can be manipulated.
However, this new study takes it to a whole new level because it involves a whole range of magical thinking tests (not just the ‘snowy patterns’) and shows how a number they are subject to the tides of emotion and feelings of being in control.
Link to Hutson’s excellent write-up.
Link to study in Science.
Link to DOI entry for same.
Mind Hacks 
Wow, so glad I caught this post. I spent quite a while last fall (in the midst of applying to grad schools, actually) trying to find any studies of this phenomenon in non-pathological populations. I had rotten luck with “apophenia” and “pareidolia” (though I suppose the former should’ve turned up your paper) and I guess I didn’t come up with “illusory patterns” or “magical thinking”. I’m totally going to print these up and go read them in bed instead of studying for my midterm.
Thanks for posting this work, and keep up the great blogging.
I’m a bit disturbed by the claim by these researchers that randomly generated patterns do not contain real images. Certainly such randomly generated images do not contain specific content created by the researchers, but they must realize that a randomly generated image of the type they are using of arbitrary size will actually contain the precise images that they are adding in. In other words, any image – even one that is arbitrarily complicated such as a page of text – can be created by a random image generator if it is run for a long enough period.
The “purely random” images that the researchers used actually contain a great deal of information, so it is not a surprise that a large number of people perceived partially obscured images in them.
Indeed, the images they were interpreting were not purposely added by the researchers and were not as clearly displayed as those that were, but to suggest that to see the chair in the above image is “real” and that the “monster” or whatever it appears to be in the second image is illusion is stretching it. Both images are illusion, as there is certainly no real chair there. Both are only images.
It would seem that what the researchers have discovered is that people make a greater effort to interpret difficult data when they are in a particular state of mind. It is unscientific of them to be making inferences about the reality of these people’s conclusions. This is stepping into the realm of philosophy.