Intuitive judgement forms the basis of decision making both by experts, in professional settings, and by people in everyday life. Psychologists have studied the rationalityof intuitive judgements. In this paper threeapproaches to decision making will be discussed : unqualified rationalism, qualified rationalism and irrationalism. The first approach holds that man is essentially rational, the second that serious cognitive biases exist, and the third that thinking is strongly influenced by noncognitive sources of distortion, i.e. emotions and motives. Evidence on judgement is reviewed and found to support the last two approaches. Various ways of improving judgements, as suggested by the three basic viewpoints, are then presented. KEY
WORDS Decision making Judgement Judgement bias Decision aidsIntuitive judgements form the basis of much decision making, both by experts and laymen. The amount of intuition may, of course, vary but its presence in even very elaborate and sophisticated judgement schemes is unavoidable. A striking example is afforded by the risk analysis of nuclear power plants as discussed by Fischhoff (1977) (cf. Sjoberg, 1979 who pointed out that subjective judgements had entered in a manner which could lead to substantial underestimation of the uncertainty of the estimated risk probabilities. Knowledge of the judgement process has great potential practical value. Psychologists have investigated judgement for a long time. It is the purpose of the present paper t o review some of that work. Studies of the quality of judgement constitute the first topic. It will be shown that biasing factors may enter into all sorts of judgements. How to minimize their influence is therefore an important problem-and that is the second topic of the paper.
THE QUALITY OF JUDGEMENTS
Low value of confidence as an indicator of qualityPeople are often quite convinced that they are right-and this may be so even if they are wrong in, e.g. rememberingcertain data (Fischhoff, Slovic and Lichtenstein, 1977). It has also been found that confidence increases as a function of the amount of information available for a decision, but that the actual quality of a decision starts to decrease (see Figure 1) after a modest amount of information (Oskamp, 1965). It seems that we become increasingly inconsistent as the amount of information increases. Thus, the subjective notion of confidence is of little help in establishing the quality of a decision or judgement. Confidence, in other words, cannot be trusted. Figure 1 is a schematic summary of research in which different ways of measuring the amount of information and reliability, or quality, ofjudgements have been used. E.g., amount of information can be measured by the number of cues, and reliability or inconsistency by the correlation between two sets of judgements of the same information given at two different occasions. The Figure does not pertain to the case where more information is actively searched in order to reply to specific questions, evaluate specific hypotheses, or e...