Overcoming Bias: Some Empirics
By Bryan Caplan
Wilson and Brekke’s justly famous article also contains an eye-opening survey of the empirics of “mental correction,” better known at GMU as overcoming bias. While I’m sure the sub-field has advanced since 1994, it’s amazing how much was already known at the time:
A number of studies have attempted to reduce biases in information processing and judgment by forewarning people about or drawing their attention to potentially biasing information and examining the extent to which they are able to avoid the unwanted effects of this information… In general, these studies have revealed a wide range of seemingly contradictory effects. Some studies have shown that an increase of people’s awareness eliminates mental contamination; some have found that awareness causes people to adjust insufficiently, leading to undercorrection; some have indicated that awareness causes people to adjust their responses too much, leading to overcorrection; and some have shown that awareness does not cause people to adjust their responses.
One case they discuss in detail is efforts to correct for priming effects.
Consider, for example, the classic priming effect, whereby people’s judgments shift in the direction of the primed category (e.g., if the category of “kindness” is accessible, people typically rate another person as more kind than they normally would; Higgins et al., 1977; Srull & Wyer, 1989). Recent studies have shown that making people aware that the category has been primed by an arbitrary event causes them to adjust their responses (Lombardi, Higgins, & Bargh, 1987; Martin, 1986; Martin, Seta, & Crelia, 1990). Interestingly, however, increasing awareness does not make the priming effect disappear; it often reverses, resulting in a contrast effect (Lombardi et al., 1987; Martin, 1986; Martin et al., 1990). For example, if people realize that kind thoughts are accessible for arbitrary reasons, they end up rating the target person as less kind than they normally would.
If Wilson and Brekke were economists, they’d probably be more inclined to treat a mixture of undercorrection and overcorrection as evidence in favor of human rationality. Either way, it’s fascinating to discover that this research not only exists, but has been sitting on the shelf for decades. Why didn’t I hear about this in grad school?