Win-stay and win-shift lever-press strategies in an appetitively reinforced task for rats.

Two experiments examined acquisition of win-stay, win-shift, lose-stay, and lose-shift rules by which hungry rats could earn food reinforcement. In Experiment 1, two groups of rats were trained in a two-lever operant task that required them to follow either a win-stay/lose-shift or a win-shift/lose-stay contingency. The rates of acquisition of the individual rules within each contingency differed: lose-shift and lose-stay rules were acquired faster than win-stay and win-shift rules. Contrary to a number of previous reports, the win-shift rule was acquired less rapidly than any of the other rules. In Experiment 2, the four rules were taught separately, but subjects still acquired the win-shift rule more slowly than any of the other rules.