In statistical inference, confirmation bias is a type of cognitive bias towards confirmation of the hypothesis under study. See also bias (statistics). To compensate for this observed human tendency, the scientific method is constructed so that we must try to disprove our hypotheses. See falsifiability.

In psychology, confirmation bias is a phenomenon whereby, in a variety of settings, decision makers have been shown to notice more, assign more weight to, and actively seek out evidence that confirms their claims, and tend to ignore and not seek out evidence which might discount their claims. As such, it can be thought of as a form of selection bias in collecting evidence.

In one classic experiment, subjects were shown the hypothesis that "If a card has a vowel on one side, it must have an even number on the other side" and four large cards, that looked something like:

E	4
7	K

Subjects were then asked which two cards they would turn over to test the rule.

Almost all experimental subjects did not choose the correct two cards, which would be E (via modus ponens) and 7 (via modus tollens). Most chose E and 4, committing the logical fallacy of affirming the consequent, and choosing a test that might confirm but which could never falsify the hypothesis.

It should be noted, however, that studies that used more concrete concepts showed very different results. In one study, students had no trouble with a version of the cards experiment in choosing "Beer" and "16" testing the hypothesis that "People who drink beer must be at least 21 years old."

Some have argued that confirmation bias may be the cause of self-perpetuating and self-fulfilling social beliefs.

This bias may occur at least partially because negatives are inherently more difficult to process mentally than positives.

More recent studies, however, have shown that while confirmation bias tends to be present as an initial condition, the repeated presentation of disconfirmatory data induces changes in theoretical thinking. In other words, the initial disconfirmatory data is regarded as the result of error, or some other externally attributed factor; it is only after similar results or data are repeatedly obtained that a change in causal reasoning strategies occurs.

See also

• disconfirmation bias
• expectancy effect
• list of cognitive biases.

References

• Wason, P. C. (1966). Reasoning. In B. M. Foss (Ed.), New horizons in psychology I, 135-151. Harmondsworth, UK: Penguin.
• Wason, P. C. (1968). Reasoning about a rule. Quarterly Journal of Experimental Psychology, 20, 273-281.
• Mynatt, C. R., Doherty, M. E., & Tweney, R. D. (1977). Confirmation bias in a simulated research environment: an experimental study of scientific inference. Quarterly Journal of Experimental Psychology, 29, 85-95.
• Griggs, R. A. & Cox, J. R. (1982). The elusive thematic materials effect in the Wason selection task. British Journal of Psychology, 73, 407-420.
• Fugelsang, J., Stein, C., Green, A., & Dunbar, K. (2004). Theory and data interactions of the scientific mind: Evidence from the molecular and the cognitive laboratory. Canadian Journal of Experimental Psychology, 58, 132-141.

External links

• Skeptic's Dictionary: confirmation bias (http://skepdic.com/confirmbias.html)
• Teaching about confirmation bias (http://www.devpsy.org/teaching/method/confirmation_bias.html)