Gain-loss frequency and final outcome in the Soochow Gambling Task: A Reassessment

Behavioral decision making literature suggests that decision makers are guided less by final outcome than by immediate gain-loss. However, studies of the Iowa Gambling Task (IGT) under dynamic and uncertain conditions reveal very different conclusions about the role of final outcome. Another research group designed a similar yet simpler game, the Soochow Gambling Task (SGT), which demonstrated that, in dynamic decision making, the effect of gain-loss frequency is more powerful than that of final outcome. Further study is needed to determine the precise effect of final outcome on decision makers. This experiment developed two modified SGTs to explore the effect of final outcome under the same gain-loss frequency context.

Each version of the SGT was performed by twenty-four undergraduate Soochow University students. A large-value (+/- $200, +/- $550 and +/- $1050) and a small-value (+/- $100, +/- $150 and +/- $650) contrast of SGT were conducted to investigate the final outcome effect. The computerized SGT was launched to record and analyze the choices of the participants.

The results of both SGT versions consistently showed that the preferred decks A and B to decks C and D. Analysis of learning curves also indicated that, throughout the game, final outcome had a minimal effect on the choices of decision makers.

Experimental results indicated that, in both the frequent-gain context and the frequent-loss context, final outcome has little effect on decision makers. Most decision makers are guided by gain-loss frequency but not by final outcome.