2004 年 22 巻 2 号 p. 168-173
This study examined the effect of reinforcer devaluation on within-session response decreases. After 16 rats were trained to press a lever for food pellets, pellets flavored with either cinnamon or cocoa powder were associated with lithium chloride. The rats' responses were then reinforced by one flavor of the pellets during 45-minute sessions in a continuous reinforcement (CRF) schedule for 8 days. The pellet flavor was alternated daily. At the start of the session the response rates were decreased by the taste-aversion learning. However, the manipulation did not alter the rate of the response decrease. The response rates were described by a linear function of the cumulative number of food pellets consumed within sessions. Reinforcer devaluation systematically altered the x- and y-axis intercepts but did not alter the slopes of the regression lines. These results suggest that taste-aversion learning lowered the initial response rate (y-axis intercept), but did not alter the rate of the response decrease produced by the food pellets consumed (slope), thereby reducing the cumulative number of pellets that would reduce the response rate to zero (x-axis intercept). These findings also revealed that the effects of the reinforcer devaluation on the parameters of the regression lines could be dissociated from the hunger attenuation reported elsewhere (Aoyama, 2000).