Illusory correlation

Illusory correlation is the phenomenon of seeing the relationship one expects in a set of data even when no such relationship exists. When people form false associations between membership in a statistical minority group and rare (typically negative) behaviors, this would be a common example of illusory correlation. [cite book | last = Pelham | first = Brett | title = Conducting Research in Psychology | publisher = Wadsworth Publishing | location = Belmont | year = 2006 | isbn = 0534532942 ] Illusory correlation is when people tend to overestimate a link between two variables. However, the correlation is slight or not at all. This happens because the variables capture the attention simply because they are novel or deviant. This is one way stereotypes form and endure. David Hamilton and Terrence Rose (1980) found that stereotypes can lead people to expect certain groups and traits to fit together, and they overestimate the frequency of when these correlations actually occur [web cite|url=http://www.colorado.edu/conflict/peace/problem/stereoty.htm|title=Stereotypes] . People overestimate the core association between variables such as stereotyped groups and stereotypic behavior. [cite book | last = Kassin | first = S. | last2 = Fein | first2 = S. | last3 = Markus | first3 = H.R. | title = Social Psychology | publisher = Houghton Mifflin Co. | location = Boston | year = 2008 | isbn = 0618998586 ]

Chapman and Chapman (1971) studied the effect as it relates to psychodiagnostic signs. Their study showed that although projective testing is not helpful in the diagnosis of mental disorders, some psychologists continue to use such tests because of a perceived, illusory, correlation between test results and certain attributes. An example of a projective test is the "Draw a Person" test that asks patients to draw a person on a blank piece of paper. Some psychologists believe in a correlation between drawing a person with big eyes and paranoia. No such correlation exists, and when data that is deliberately uncorrelated is presented to college students they find the same illusory correlations that psychologists believe in. [Citation | last = Chapman | first = L. J. | last2 = Chapman | first2 = J. P. | title = Genesis of popular but erroneous psychodiagnostic observations | journal = Journal of Abnormal Psychology | volume = 72 | pages = 193-204 | year = 1967 ]

This bias can be caused by, among other things, an event that stands out as unique. For example, "The only time I forget my pencil is when we have a test". This is most likely an illusory correlation. It could be caused by only a few other pencil-less tests, which stand out particularly well in memory [cite web|url=http://changingminds.org/explanations/theories/illusory_correlation.htm|title=Illusory Correlation] .

See also

* Cognitive bias
* Observer bias
* Expectancy effect
* Post hoc ergo propter hoc
* Cum hoc ergo propter hoc
* Radical behaviorism
* Superstition

References