A review of the literature indicates that linear models are frequently used in situations in which decisions are made on the basis of multiple codable inputs. These models are sometimes used normatively to aid the decision maker, as a contrast with the decision maker in the clinical vs statistical controversy, to represent the decision maker "paramorphically" and to "bootstrap" the decision maker by replacing him with his representation. Examination of the contexts in which linear models have been successfully employed indicates (...) that the contexts have the following structural characteristics in common: each input variable has a conditionally monotone relationship with the output; there is error of measurement; and deviations from optimal weighting do not make much practical difference. These characteristics ensure the success of linear models, which are so appropriate in such contexts that random linear models may perform quite well. 4 examples involving the prediction of such codable output variables as GPA and psychiatric diagnosis are analyzed in detail. In all 4 examples, random linear models yield predictions that are superior to those of human judges. (shrink)

Dawes points out the fallacy in many commonly held beliefs in therapy and takes issue with many current treatment methods.

Proper linear models are those in which predictor variables are given weights such that the resulting linear composite optimally predicts some criterion of interest; examples of proper linear models are standard regression analysis, discriminant function analysis, and ridge regression analysis. Research summarized in P. Meehl's book on clinical vs statistical prediction and research stimulated in part by that book indicate that when a numerical criterion variable is to be predicted from numerical predictor variables, proper linear models outperform clinical intuition. Improper (...) linear models are those in which the weights of the predictor variables are obtained by some nonoptimal method. The present article presents evidence that even such improper linear models are superior to clinical intuition when predicting a numerical criterion from numerical predictors. In fact, unit weighting is quite robust for making such predictions. The application of unit weights to decide what bullet the Denver Police Department should use is described; some technical, psychological, and ethical resistances to using linear models in making social decisions are considered; and arguments that could weaken these resistances are presented. (shrink)

Social dilemmas occur when the pursuit of self-interest by individuals in a group leads to less than optimal collective outcomes for everyone in the group. A critical assumption in the human sciences is that people's choices in such dilemmas are individualistic, selfish, and rational. Hence, cooperation in the support of group welfare will only occur if there are selfish incentives that convert the social dilemma into a nondilemma. In recent years, inclusive fitness theories have lent weight to such traditional views (...) of rational selfishness on Darwinian grounds. To show that cooperation is based on selfish incentives, however, one must provide evidence that people donotcooperatewithoutsuch incentives. In a series of experimental social dilemmas, subjects were instructed to make single, anonymous choices about whether or not to contribute money for a shared “bonus” that would be provided only if enough other people in the group also contributed their money. Noncontributors cited selfish reasons for their choices; contributors did not. If people are allowed to engage in discussion, they will contribute resources at high rates, frequently on irrational grounds, to promote group welfare. These findings are consistent with previous research on ingroup biasing effects that cannot be explained by “economic man” or “selfish gene” theories. An alternative explanation is that sociality was a primary factor shaping the evolution ofHomo sapiens. The cognitive and affective mechanisms underlying such choices evolved under selection pressures on small groups for developing and maintaining group membership and for predicting and controlling the behavior of other group members. This sociality hypothesis organizes previously inexplicable and disparate phenomena in a Darwinian framework and makes novel predictions about human choice. (shrink)

Suppose the principles explaining how the human mind (brain) reaches logical conclusions and judgments were different from – and independent of – thoseinvolved innormatively valid reasoning. Then such principles should affect both conclusion generation and recognition that particular conclusions are or are not justified. People, however, demonstrate a discrepancy between impaired performance in generating logical conclusions as opposed to rather impressive competence in recognizing rational (versus irrational) ones. This discrepancy is hypothesized to arise from often generating an incomplete specification of (...) a logical or judgmental problem when attempting to solve it – versus a recognition of such incompleteness when it is pointed out. The basic argument is developed, with common examples, in the context of specifying or failing to specify all possible combinations in simple logical arguments and is then extended to probabilistic reasoning, where complete versus incomplete specification corresponds to attending to all or to only some components of Bayes theorem-based reasoning. (shrink)

Suppose the principles explaining how the human mind (brain) reaches logical conclusions and judgments were different from – and independent of – thoseinvolved innormatively valid reasoning. Then such principles should affect both conclusion generation and recognition that particular conclusions are or are not justified. People, however, demonstrate a discrepancy between impaired performance in generating logical conclusions as opposed to rather impressive competence in recognizing rational (versus irrational) ones. This discrepancy is hypothesized to arise from often generating an incomplete specification of (...) a logical or judgmental problem when attempting to solve it – versus a recognition of such incompleteness when it is pointed out. The basic argument is developed, with common examples, in the context of specifying or failing to specify all possible combinations in simple logical arguments and is then extended to probabilistic reasoning, where complete versus incomplete specification corresponds to attending to all or to only some components of Bayes theorem-based reasoning. (shrink)

Professionals often believe that they must “exercise judgment” in making decisions critical to other people’s lives. The relative superiority of statistical prediction rules to intuitive judgment for combining incomparable sources of information to predict important human outcomes leads us to question this personal input belief. Some professionals hence use SPR’s to “educate” intuitive judgment, rather than replace it. In psychology in particular, such amalgamation is not justified. If a well‐validated SPR that is superior to professional judgment exists in a relevant (...) decision making context, professionals should use it, totally absenting themselves from the prediction. (shrink)

Professionals often believe that they must “exercise judgment” in making decisions critical to other people's lives. The relative superiority of statistical prediction rules to intuitive judgment for combining incomparable sources of information to predict important human outcomes leads us to question this personal input belief. Some professionals hence use SPR's to “educate” intuitive judgment, rather than replace it. In psychology in particular, such amalgamation is not justified. If a well-validated SPR that is superior to professional judgment exists in a relevant (...) decision making context, professionals should use it, totally absenting themselves from the prediction. (shrink)

As illustrated by research Koehler himself cites (Dawes et al. 1993), the purpose of experiments is to choose between contrasting explanations of past observations – rather than to seek statistical generalizations about the prevalence of effects. True external validity results not from sampling various problems that are representative of “real world” decision making, but from reproducing an effect in the laboratory with minimal contamination (including from real world factors).