ABSTRACT
In decision research, the formal language of risk and uncertainty is somewhat confusing. Traditionally the two terms have been differentiated according to whether or not precise probabilities can be attached to events. Edwards (1954) gave tossing a coin as an example of an event involving risk and the proposition ‘Will you drink a glass of beer immediately after reading this paper?’ as an example of uncertainty. In the latter case, he suggested, there are no generally accepted rules for assigning a precise probability to the event, whereas most people accept that the probability of the coin landing ‘heads’ is precisely 0.5. To complicate matters, risky and uncertain events (as defined above) which may result in negative consequences are both termed risks in everyday language. The less confusing terms well-defined and ill-defined risks might better distinguish risk and uncertainty in Edwards’ sense. Much early research focused on welldefined risks and used simple gamble tasks to test alternative theories. Such lotteries provide precise probability information, but apart from notable exceptions such as the National Lottery and the gambling casino, most real-life risks are illdefined. That is, the probabilities of uncertain events which may follow a decision are imprecise, vague or ambiguous, and are essentially subjective. Researchers questioning the idea that lotteries display the essential characteristics of realworld risks have begun to focus on alternative tasks. For example, Tversky and Kahneman have departed from the standard lottery in developing Cumulative and Generic Prospect Theory.
