Some cross-cultural data on probability learning

Extended training in a probability learning task was given to 17 Ss from two non-western cultures. Ss came from Moscow University, USSR and Cuttington, Liberia, a small rural African town. Long-term probability matching was obtained in both groups. However, the trial-to-trial changes in response proportions varied markedly between groups; Liberian Ss tended to follow the reinforcing events, while responses by the Russian Ss depended more heavily on their own preceding responses. Some negative recency effect was obtained for both groups. This effect decreased somewhat over trials, but was still present at the termination of training.