While a considerable amount of research has centered on explaining the risky shift (enhancement of risk acceptance through group discussion; see, for example, Dion, Baron, and Miller, 1970), few studies have investigated the generality of the phenomenon. Three of several ways in which the laboratory context is usually lacking of reality are: (a) The decision consequences are imaginary, as in the ‘Choice-Dilemma’ situations used in most of the risky-shift research; (b) the decision consequences - even if they are real (e.g. money) - are slight; and (c) the decisions are made for oneself (whereas in real life many decisions involve agents acting on behalf of others).

The present study was designed to investigate the effects of the above three factors - type of incentive, magnitude of stake and decision target - using a betting task as the decision situation.