Runway Performances as a Function of Reinforcement Schedule and Alley Length

The present study proposed to arrange and integrate the classical conditioning and the instrumental learning. 48 rats were trained to run to water in a straight alley and randomly assigned into six groups of 8 each, namely, G1: 32 in.-100%; G2: 32 in.-50; G3: 8 in.100%; G4: 8 in.-50%; G5,: 4 in-100%; and G6: 4 in.-50%. A factorial design was employed with three lengthes of alley and two levels of reinforcement schedules. The data were collected in ways of: (1) measures of blocks of four trials for 12 acquisition days; (2) measures of blocks of four trials for 9 extinction days; (3) measures of the first trial of 12 acquisition days; and (4) measures of the first trial of 9 extinction days. Further, four recordings of each section of different alley were measured, i. e., the starting speed; the running speed of the first 4 in section of the runway; the running speed of the second 4 in. section of the runway; and the running speed of the last 24 in. section of the runway. Namely, there were four recordings of 32 in. alley groups, three recordings of 8 in. alley groups and only two recordings of 4 in. alley groups. Analyses of variance indicated that starting speed was a most sensitive measure of all, since it yielded significant and consistent major effects. For the extinction performances, the findings were as we expected that there was a greater resistance to extinction under partial reinforcement schedule. It was originally hypothetized that the 8 in. and 4 in. alley groups might be comparable to the classical conditioning and hence the partial reinforcement should not result in significant and consistent effect as that in the last section of the runway. However, the results indicated that during the acquisition, the effect of the reinforcement schedule was significant in the shorter alley groups. Anyway, the results were consistent with the previous instrumental reward studies. That is, the effects of variation of the reinforcemennt schedule in instrumental conditioning situation may depend upon which portion of the response chain is measured. We also confirmed the prediction that the response latency depends on the response speed.