ABSTRACT
Amos Tversky and Daniel Kahneman used the term representativeness heuristic as a general label for these types of ‘errors’ or ‘biases’ in human judgment. The hypothesis underlying the representativeness heuristic is that most people have an image of how a roulette wheel behaves. Kahneman and Tversky have been consistent in pointing out that heuristics are often useful. For Tversky and Kahneman, heuristics are violations of good thinking that are necessary due to information processing limitations. For Gerd Gigerenzer, heuristics are intelligent adaptations that often make the information processing limits irrelevant to the ultimate performance levels that can be achieved. Sometimes the rule-based heuristics of an abduction system will reflect the circumstantial patterns. The Tversky and Kahneman approach, at least implicitly, suggests that the goal in designing decision support systems should be to turn people away from heuristic approaches.
