A Molecular Analysis of Choice on Concurrent‐chains Schedules

Abstract

Six pigeons responded on concurrent-chains schedules with either independent or interdependent equal variable-interval schedules in the initial links and unequal variable-interval schedules, always in a 2:1 ratio, in the terminal links. Relative response rates in the initial links increased across conditions as initial-link duration was shortened and decreased across conditions as terminal-link duration was shortened, replicating previous findings. Responses in the initial links were recorded in 5-s bins, and local or molecular relative response rates were calculated in order to ascertain how relative response rate varied as a function of time since the onset of the initial links. Two distinct molecular patterns were found. With interdependent initial links, relative response rates for the preferred key were elevated for the first 10 or 20s of the initial links and then declined to an asymptotic value. With independent initial links, a negative recency effect was found similar to that reported by Killeen (1970). These two molecular patterns were related to the different momentary reinforcement probabilities resulting from independent and interdependent scheduling.Key words: choice, molecular analysis, conditioned reinforcement, delay-reduction hypothesis, concurrent-chains schedules, variable-interval schedules, forced-choice procedure, key peck, pigeonsThe concurrent-chains schedule has been used widely in research during the last 25 years (see Fantino, 1977, for a review). Although it was originally developed for study of conditioned reinforcement (Autor, 1960(Autor, , 1969, the concurrent-chains schedule has more often been used to investigate preference between different schedules of reinforcement, a task to which the procedure seems particularly well suited because it allows a measure of choice uncontaminated by the particular patterns of responding generated by the schedules that are chosen. In this procedure, the subject responds on two concurrently available alternatives (the initial links or "choice phase") usually correlated with equal variable-interval (VI) schedules. Responses on each alternative occasionally produce another stimulus, correlated with entry into the terminal link of the chain on that alternative (the "outcome phase"). Responses in the terminal link are reinforced with food. The independent variable has generally involved some difference in the conditions arranged during the terminal links, such as differing rates of reinforcement. in-145 1987, 48, 145-159 NUMBER 1 (JULY)

Choice Behavior in Transition: Development of Preference for the Higher Probability of Reinforcement

“…Transient effects of reinforcement on subsequent responses have been observed in a number of other experiments on choice, but the effects of individual reinforcers are not always in the same direction. Fantino and Royalty (1987) reviewed the results of several studies of this type and presented some additional data from concurrent-chains schedules. They noted that positive recency effects have been found in studies that used concurrent schedules and an interdependent scheduling of reinforcers (e.g., Menlove, 1975;Morgan, 1974).…”

“…They noted that positive recency effects have been found in studies that used concurrent schedules and an interdependent scheduling of reinforcers (e.g., Menlove, 1975;Morgan, 1974). A negative recency effect (a transient decrease in the probability of a response on the same key) has been found in studies with concurrent-chains schedules and independent scheduling of reinforcers (e.g., Fantino & Royalty, 1987;Killeen, 1970). With other combinations (concurrent schedules with independent scheduling or concurrent-chains schedules with interdependent scheduling), neither positive nor negative recency effects have been observed.…”

Concurrent-chain performance in transition: Effects of terminal-link duration and individual reinforcers

“…However, other studies have found the opposite: a negative recency effect, in which there was a lower response percentage on the alternative that had just delivered a reinforcer (Fantino & Royalty, 1987;Killeen, 1970). Fantino and Royalty proposed that negative recency effects are most likely to be seen when the initial links are two independent VI schedules, because immediately after a reinforcer is delivered for one alternative, the probability of reinforcement is temporarily increased for the opposite alternative.…”

Regularities in responding during performance of a complex choice task

“…Such molar measures of responding provide the foundation for many modern theories of operant conditioning and have been particularly important in clarifying how animals process temporal cues and make response decisions during conditioning (for review, see Staddon & Cerutti, 2003). A minority of researchers has attempted to supplement this approach with more molecular measures of responding during learning (Blough, 1963;Fantino & Royalty, 1987;Newland, 1997;Palya, 1992;Shimp, 1992;Silva & Pear, 1995;Silva, Pear, Tait, & Forest, 1996;Weiss, 1970), and to use such measures to develop alternative theories of conditioning (Galbicka, 1992;Misak & Cleaveland, 2011;Shimp, 1966;Shimp, Fremouw, Ingebritsen, & Long, 1994;Tanno & Silberberg, 2012). For instance, Silva and Pear (1995) analyzed measures of pigeon's spatial trajectories as the birds learned to respond to fixed-time (FT) and fixed-interval (FI) reinforcement schedules and found that movement stereotypy varied as a function of when rewards were delivered and on whether rewards were contingent on pecking.…”