Albino rats were trained in a straight runway for 110 days, two trials/day, with an intertrial interval of 3-5 min. One trial each day ended in large reward (10 standard Noyes pellets for one group, 1 ml of 30% sucrose solution for another) and the other trial ended with small reward (1 pellet or 3% sucrose). The daily sequence of reward magnitudes, either small-large (SL) or large-small (LS), was determined by pseudorandom orders, so that a block of 8 days contained 4 SL and 4 LS days. For 20 days following this random-sequence training, all rats received fixed sequences, always SL or always LS, with the same reward type (standard pellets or sucrose solutions) as that used during random-sequence training. Differential running for different reward magnitudes (patterning) did not develop during random-sequence ttaining with either reward type. The fixed-sequence training generated patterning on Trial 2 but not on Trial 1. The data are discussed in the light of several possible sources of the stimulus control of patterning, and implications of the data for a sequential account of the successive negative contrast effect are briefly considered.

No abstract