The effects of prefeeding in plain and chained fixed ratio schedules of reinforcement

Abstract

The effects of prefeeding were studied using pigeons reinforced on either plain fixed ratio (FR) 64 or chained FR 32, FR 32 schedules of reinforcement. It was found that (1) the response rate increased as a func tion of the number of responses completed in the ratio, (2) the rate was higher for the chained FR group, (3) the rates decreased as a function of the amount prefed, (4) the latency of the first response in the ratio increased as a function of the amount prefed, (5) the prefeeding had a much greater effect on the rate in the first part than in the later part of the fixed ratio sequence, and (6) the prefeeding had a much greater effect on the response latency than on the response rate. It was concluded that the multiplicative rule of the effect of feeding schedules is inapplicable to behavior changes resulting from different segments of a single schedule of reinforcement. Cumming, Berryman & Nevin (1965) indicate that response rate is a multiplicative function of the feeding schedule. 2 Stated symbolically, Ri =CXRj' where R is the response rate, i and j indicate two different feeding schedules, and C is a constant. This implies that the ratio of the two rates equals a constant, In other words, manipulation of a variable which affects the absolute value of the rates will leave the ratio unaffected; R. R. ~=~=C R. R. ' J,m J,n where m and n indicate two different values of some other variable. Cross multiplication results in the following "corollary": The ratio of the response rate obtained under two different conditions of one variable is unaffected by changes in the feeding schedule; R. R. training stimuli in a stimulus generalization test; the rule does not seem to be applicable in that situation (Jenkins, Pascal, & Walker,1958; Thomas & King,1959). The present experiment is concerned with a test of the applicability of the multiplicative rule to the response rate in different segments of a single schedule of reinforcement ; e.g. the first and last 16 responses of a fixed ratio 64 schedule of reinforcement. Method Eight Silver King pigeons at 70% ad lib weight and standard operant conditioning equipment were used. The chained fixed ratio (FR) group was exposed to the following schedule of food reinforcement: During the first component, S-, the pigeon response key was dark until S waited for 30 sec. without responding. During the second component, S~, a horizontal white line was projected behind the key until the subject made 32 responses. During the final component, si, a vertical white line was projected until the S made 32 more responses. Then the food magazine was presented for 3 sec. and the key was darkened. The schedule of reinforcement for the plain FR group was the same except that the S2 component was eliminated and the number of responses required in the sr component was increased to 64. After the behavior had stabilized, the testing phase was begun. For three sessions, the Ss were not prefed, then for three sessions they were prefed 15 gm of food; and finally for three sessions, they were prefed 30 gm. Prefeeding was done 3 hr. prior to the session. In addition to the usual cumulative response records, two other measures were obtained by the use of running time meters: (1) the average latency ofthe first response following the offset of S-was recorded and (2) the 64 responses which intervene between Sand the reinforcement were divided into four blocks of 16 responses and the average response rate was computed separat'~ly for each of the four blocks (actually the rate for the first block of responses was computed using responses 2 through 16 since it is not meaningful to include the time required for the completion of the first response following S-). Results and Discussion Figure 1 shows that for both groups and all amounts of prefeeding, the response rate was a monotonically increasing function of the number of responses completed in the fixed ratio (cf. Thomas, 1964). For all blocks and all amounts of prefeeding, the response rate was higher for the chained than for the plain FR group; further research will be needed to determine the 285