Models of rational thinking assume that individuals who need to add information and then subtract it will return to their original starting point. Empirical studies, however, show that individuals exposed to such addition-subtraction processes systematically tend to overcorrect. One study, for example, compared evaluations of two job candidates. The first candidate had two positive recommendation letters. The second candidate had a third, negative letter that was eventually discovered to be irrelevant. Participants perceived the second candidate as better, even though, in the end, both candidates presented the same information. The introduction and later rejection of the negative information led to an over-favorable evaluation. Moving to the legal context--where evidence is often presented and later refuted--overcorrection suggests unavoidable distortion in the evaluation of the parties' remaining evidence. The following analysis presents the wide body of studies concerning overcorrection and demonstrates the effect that overcorrection may have in the legal context. Rather than arguing for the deficiency of the legal process, the analysis presents the surprising and effective mechanisms that the law employs to tackle overcorrection or mitigate its results. The risk of overcorrection can explain several of the characteristics of factual determination in the legal process.
Available at: http://works.bepress.com/guttel/4/