Escalation: The Role of Problem Recognition and Cognitive Bias

Decision Sciences, Aug 2007 by Keil, Mark, Depledge, Gordon, Rai, Arun

This view of selective perception as being a product of situation and individual experience dovetails with Hogarth's (1987) four-part description of what selective perception is: (i) people structure problems on the basis of their own experience, (ii) anticipations of what one expects to see bias what one does see, (iii) people seek information consistent with their own views/hypotheses, and (iv) people downplay or disregard conflicting evidence.

The first element of Hogarth's definition is consistent with an influential study conducted by Dearborn and Simon (1958), which claimed that managers selectively perceive information according to their functional background. However, Walsh (1988) was unable to find supporting evidence for this in a study that was designed to replicate and extend Dearborn and Simon's original work.

Other researchers have found that functional background "tends to restrict the areas to which [managers] pay attention" and might lead them to ignore certain stimuli (Beyer et al., 1997, p. 730). Furthermore, Waller, Huber, and Glick (1995) determined that functional background affected the kinds of changes that executives perceive in the effectiveness of their organization, but not its environment.

Attempting to reconcile the sometime contradictory findings of prior research, Bunderson and Sutcliffe (1995) suggest that individuals develop knowledge structures based on their functional background, but that this functional orientation can be modified by situational factors. Specifically, they suggest three situational factors that can moderate the effect of an individual's functional background on the issues that they perceive as important: (i) situational strength, which is "the extent to which interpretation of a certain situation is clear" (p. 461); (ii) individual accountability; and (iii) firm strategy.

Taken together, the results of these studies appear to indicate that selective perception can play an important role in the information to which an individual attends and can thus, in theory, influence decision making. Because escalation of commitment involves a decision-making situation in which an individual must choose whether or not to continue with a previously chosen course of action, it is reasonable to investigate the role of selective perception in this context. While selective perception has long been recognized as a potentially strong cognitive bias, we could find no prior research that has investigated whether it affects decision making in escalation situations.

Illusion of control

Langer (1975, p. 313) defines illusion of control as "an expectancy of a personal success probability inappropriately higher than the objective probability would warrant." In a series of experiments, Langer found that an illusion of control can be induced through the introduction of an aspect of a skill situation into a purely chance event.

Prior research has shown that people will often attribute a favorable outcome to their skills in managing a situation, even when the outcome is purely fortuitous (Jones et al., 1972; Harvey & Weary, 1985). However, when the outcome is unfavorable, they will attribute that to some chance factor outside of their control (Langer, 1975; Hogarth, 1987). Thus, perhaps unsurprisingly, illusion of control tends to occur more often in individuals with a prior history of success (Langer & Roth, 1975; Duhaime & Schwenk, 1985). Langer and Roth (1975), for example, conducted an experiment in which subjects were asked to predict the outcome of 30 successive coin tosses. Subjects were given predetermined feedback on the correctness of their predictions. In one treatment condition, the pattern of initial feedback suggested that their predictions were quite accurate, whereas in another treatment condition, the initial pattern of feedback suggested that their predictions were quite inaccurate. In both treatment conditions, subjects were told they had made 15 accurate predictions and 15 inaccurate predictions. The results of the experiment showed that, when subjects were given initial feedback suggesting they were accurate predictors of coin tosses, they developed a greater illusion of control over their ability to predict the outcome of coin tosses. The results of Langer and Roth's experiment suggest that a previous history of success in predicting a chance event can promote an illusion of control.