“I am convinced there is a link between two events I have witnessed, but they are not related.”
A correlation is said to occur when an event (for example, a cloudy sky) is associated with another event (for example, a rain shower) . If it can be difficult to establish a relationship between two events, it is even more difficult to correctly predict such relationships. This can lead to a particular kind of error in our reasoning: the illusory correlation bias . It’s possible to manifest this bias either by overestimating the degree of correlation, or by perceiving a non-existent correlation. Expressed more simply, an illusory correlation is the false impression that there is a link between two elements, when in reality none exists, or it is much weaker than one believes . This bias is present in everyone in varied contexts such as in popular or folk beliefs, superstition, errors in clinical judgement or in stereotypes .
A harmless example that concretely illustrates the illusory correlation bias consists of perceiving a relationship between an individual’s personality traits and their astrological sign. For example, some people recognize personality traits typical of themselves or of others when reading a horoscope. You have surely heard someone say “it’s normal that you would be impulsive, you must be an Aries…” or “you Geminis are all curious” . These are certainly examples of correlation bias because astrology has no foundation and typical personality traits cannot be deduced from an astrological sign .
One of the dominant explanations of the bias implies the availability heuristic. It would seem that we find it much easier to remember information associated with rare or significant elements [1, 4]. Thus, we tend to believe such elements occur more often, because we remember them more easily, overestimating their frequency. When we overestimate the frequency of the correlation between two rare or significant elements, we are manifesting a correlation bias based on frequency [1, 4]. Our expectations can also lead to forming an illusory correlation: if we expect there to be a relationship between two elements, we may see one even though it is not there, or we may overestimate its importance . In this case, our pre-existing beliefs push us to make faulty judgements about our environment, which lead to the false impression that there is a correlation between the elements under consideration [1, 4]. In this specific case, we have an example of a correlation bias based on expectations .
The illusory correlation bias can make people look for ways to confirm their beliefs. If someone already believes that there is a correlation between two elements, they will notice and remember more events that confirm the belief, as opposed to those that refute it. Illusory correlations can also lead people to make irrational judgements or choices. For example, someone may refuse to drive under a full moon, because they mistakenly believe that there is a correlation between the full moon and a higher accident rate. Illusory correlations are also strongly implicated in the formation of stereotypes. An example of a stereotype might be “Accountants are all unhappy”. This is an illusory correlation because one is overestimating the correlation between a particular character trait and a group .
Thoughts on how to act in light of this bias
Don’t try to explain everything, sometimes it’s pure coincidence.
Try to use all of the information available.
Don’t necessarily seize on the most obvious hypothesis to explain a relationship.
How is this bias measured?
Certain experiments have presented participants with drawings produced by psychiatric patients. The drawings were then randomly assigned to a psychiatric diagnosis. The researchers noticed that participants made illusory correlations between characteristics in the drawings (for example, shape of the head), and the diagnosis (for example, low intelligence), even when no such association existed empirically .
This bias is discussed in the scientific literature:
This bias has social or individual repercussions:
This bias is empirically demonstrated:
 Chapman, Loren J. & Jean P. Chapman (1967). Genesis of popular but erroneous psychodiagnostic observations. Journal of Abnormal Psychology 72(3), 193–204.
 Chapman, Loren J. (1967). Illusory correlation in observational report. Journal of Verbal Learning & Verbal Behavior, 6(1), 151–155.
 Carlson, Shawn (1985). A double-blind test of astrology. Nature 318, 419–425.
 Berndsen,Mariette, Joop van der Pligt, Russell Spears & Craig McGarty (1996). Expectation-based and data-based illusory correlation: The effects of confirming versus disconfirming evidence. European Journal of Social Psychology 26(6), 899–913.
 Hamilton, David L. & Steven J. Sherman (1989) Illusory Correlations: Implications for Stereotype Theory and Research. In: Bar-Tal D., Graumann C.F., Kruglanski A.W., Stroebe W. (eds) Stereotyping and Prejudice. Springer Series in Social Psychology. Springer, New York, NY.
Individual level, Intergroup level, Availability heuristic, Need for cognitive closure, Need for security
Caroline Coupal, Bachelor’s student in cognitive neuroscience, Université de Montréal
Translated from French to English by Kathie McClintock.
How to cite this entry
Coupal, C. (2020). Illusory correlation, trans. K. McClintock. In E. Gagnon-St-Pierre, C. Gratton & E. Muszynski (Eds). Shortcuts: A handy guide to cognitive biases Vol. 2. Online: www.shortcogs.com