![]() The DK effect also occurs in different environments, for example the problem of face recognition. A more precise formulation of the regression better-than-average approach was provided in the noise-plus-bias model ( Burson et al., 2006). Combining these two facts leads to the regression better-than-average approach, and it explains the asymmetry of the DK effect: overestimation in the bottom quartile and underestimation in the upper quartile. This phenomenon is called “regression to the mean” and has been known since Galton (1886) studied the relationship between the height of sons and fathers. Second, the slope in the linear regression of estimated performance on actual performance is not equal but less than one. In a survey of engineers, 42% thought their work ranked in the top 5% among their peers ( Zenger, 1992) and in a survey of college professors, 94% thought they performed “above average” ( Cross, 1977). ![]() Most people think they drive better than average ( Svenson, 1981). First, it is well-known that people tend to overestimate their performance. The attack on Dunning and Kruger was initiated by Krueger and Mueller (2002), who suggested a regression better-than-average approach which is parsimonious and “does not require mediation by third variables, such as metacognitive insights into one's own problem-solving abilities 1.” Their approach is based on two empirical facts. By developing a rational model of self-assessment, they showed that the DK effect can be produced by two psychological mechanisms.īut there has also been much criticism and this criticism typically relies on a statistical rather than a psychological explanation of the DK effect. Their model for the probability of a correct answer is an extension of the one-parameter item response theory (IRT) model, known as the Rasch model ( Embretson and Reise, 2013). (2021) replicated two of Dunning and Kruger's studies using a sample of 4,000 participants. (2016) explained the DK effect based on the anchoring and adjustment heuristic, while Jansen et al. (2021) and partial responses in Kruger and Dunning (2002), Dunning et al. (2018), West and Eaton (2019), Gabbard and Romanelli (2021), and Mariana et al. Most studies recognize that there is a DK effect and provide a psychological explanation, sometimes agreeing, sometimes disagreeing with Kruger and Dunning's metacognitive explanations see Ehrlinger et al. There has been both criticism and support. First, is there a DK effect? And second, is the explanation provided by Dunning and Kruger correct? The Kruger–Dunning article raises two questions. The accuracy of the prediction was high in the top group and low in the bottom group, and the prediction in the bottom group was highly overestimated. Calculating the average perceived ability in each group, Dunning and Kruger obtained Figure 1. The students were then split in four groups according to their actual test scores. After the test they asked the students to assess their performance in the test. In their original article, Kruger and Dunning (1999) tested undergraduate students enrolled in various psychology courses at Cornell University for their ability in humor, logical reasoning, and English grammar. The DK effect and Dunning and Kruger's explanation of it has been discussed and challenged extensively. This second effect, although not discussed in Kruger and Dunning (1999), is often also associated with their names. A closely related effect, also important but arguably less dangerous, is that people of high ability tend to underestimate their ability. Dunning and Kruger claim that the reason for this bias is that people of low ability are not good in seeing and judging themselves (a deficit in metacognitive skills). ![]() This hypothetical cognitive bias was first described in Kruger and Dunning (1999) and, if true, it is potentially important and dangerous, because it means that people of low ability not only perform tasks poorly but (even worse) that they think that they perform these tasks well. The Dunning–Kruger (DK) effect states that people with low ability tend to overestimate their ability.
0 Comments
Leave a Reply. |