Partial reinforcement: a review and critique.

Jenkins, William O.; Stanley, Julian C.

doi:10.1037/h0060772

Cited by 279 publications

(181 citation statements)

References 50 publications

Supporting

Mentioning

158

Contrasting

Unclassified

Order By: Relevance

“…EXPERIMENT II Jenkins and Stanley (1950) have reviewed a large number of studies which support the general conclusion that resistance to extinction is greater after partial than after continuous reinforcement. In practically all the summarized experiments, several groups of subjects were used in the experimental design.…”

Section: Experiments I: Discussionmentioning

confidence: 99%

Resistance‐to‐extinction Functions in the Single Organism

Hearst

1961

J Exper Analysis Behavior

View full text Add to dashboard Cite

Functions relating the number of reinforcements to some aspect of behavior are most often obtained through the use of different groups of subjects, each given different numbers of reinforcements before a test phase that is the same for all groups. The average of the individual measurements in each group then is related to the number of reinforcements given each group. By such a method, the classic data of Williams (1938) and Perin (1942) have shown that both the mean number of extinction responses and the mean time required to attain an "extinction criterion" increase with negative acceleration as the number of reinforcements increases. Williams and Perin plotted the individual data points which enter the mean of each of their reinforcement groups. It is noteworthy that in Williams' experiment, (a) the two highest individual scorers in extinction were members of one of the low-reinforcement groups; and (b) the range of extinction scores in three of the four experimental groups was virtually the same, even though the group mean scores revealed an orderly and significant relation with the number of reinforcements. In Perin's study, all five reinforcement groups displayed their modal number of individual extinction responses at zero, although the mean data indicated a relation very similar to that of Williams.Some recent papers have considered the problem of drawing inferences about individual behavioral functions from group data (Sidman, 1952;Estes, 1956;Schrier, 1958), and there has been some consideration (Spence, 1956;Skinner, 1959) of the relative merits of individual and group functions in behavioral research. In view of these discussions and of the individual eccentricities pointed out in Williams' and Perin's data, the present set of experiments was undertaken (a) to examine a technique for obtaining resistance-to-extinction functions in individual subjects, and (b) to compare these findings with results previously obtained by separate-group methods. The present technique is founded on the pigeon's capacity for very efficient and rapid color discriminations (Ferster & Skinner, 1957) and on an extinction testing procedure analogous to that used by Guttman and Kalish (1956) and Hanson (1959) in the study of individual and small-group generalization gradients in the pigeon. EXPERIMENT I: METHOD SubjectsThe Ss were eight barren, White Carneaux hen pigeons, ranging from 4 to 8 years old.Throughout the experiment, they were maintained at 75% of their free-feeding body weights. ApparatusThe experimental box was a standard commercial model (Foringer), and its essential details have been described elsewhere (Ferster, 1953;Ferster & Skinner, 1957). Except during blackout periods (described below), the box was illuminated by a 110-volt-AC, 6-watt houselight mounted inside the electrical junction box on the rear wall of the chamber.The translucent response key could be illuminated with four pairs of colored Xmas tree lamps (red, green, blue, or orange) mounted behind the key. These lamps were 110 volts, 7.5 watts, whi...

show abstract