Free Dissertations - Psychology Dissertations

In Support For This Harvey (1997) Reports That This Bias Decreased When The ...

In support for this Harvey (1997) reports that this bias decreased when the participants were asked to give reasons for their answer being wrong. Other information-processing biases could include self-serving motivational biases, overreliance on strength of evidence rather than weight, and insufficient cognitive processing (reported by Juslin, 2000).
Theories of information-processing biases have been challenged by several models and findings. Gigerenzer et al (1991) came up with Probabilistic Mental Models (PMM) that argued that the question material was not representative of the natural environment (e.g. surprising questions) and that the overconfidence phenomenon is an artifact of design. However, Juslin et al (1995) reviewed 25 tasks that used random sampling and found that across these studies when accuracy was less than 65% there was overconfidence and greater than 78% there was under-confidence, indicating a small hard-easy effect. Therefore the difficulty of the question determines the overconfidence not the question selection. This result contradicts what the PMM would predict.
Soll (1996) modified the PMM by combining it with the Thurstone's (1927; reported in Harvey, 1997) idea that there is an inherently random component of judgments or learning relating to the regression effect. However, Budescu et al (reported by Harvey, 1997) demonstrated that the overconfidence phenomenon was still strong despite using randomly selected items and excluding the artifacts resulting from the regression effects of random error.In addition, Brenner et al (1996) generally argue against the idea that the slope of the regression line should be used as an index of overconfidence because the same data can be shown to be under- or overconfident depending on the analysis. He suggests that a simple comparison of means should be used as they are not subjected to regression effects.
Brenner et al (1996) investigated the difference between subjective probability of individual items and people's estimates of relative frequency of correct items. It is well established that people are under-confident when estimating relative frequency (e.g. May, 1988; reported in Brenner et al, 1996) but it is unknown whether this difference is due to the response scale (confidence versus frequency) or nature of the evidence for judgments. They found preliminary evidence suggesting that the difference is not because of the response scale.
In the follow study, the overconfidence phenomenon is investigated. Based on the classical findings relating to the overconfidence phenomenon, it was hypothesized that there would be a difference between the overconfidence effect for easy and difficult items with the overconfidence effect bigger for difficult items.

Order Now. It takes less than 2 minutes.

  1.  
  2.  
  1.  

Please note: The above dissertation snippet was written by a student and then submitted to us to display and help others. Thanks to all the students who have submitted their work to us.