The Amsel frustration effect in monkeys

Abstract

Abstraect Four adolescent stumptail macaques were tested in a discrete-trials retractable-lever analogue of the double runway commonly used to study frustration effects in rats. Each trial consisted of fixed-ratio responding on a left lever followed by an identical contingency on a right lever. In agreement with runway data in rats, right-lever responses were faster following nonrein-forcement than following reinforcement of left-lever responses . Problem Although the instrumental learning research of Amsel (1962) and others on frustration has developed into one of the major lines of inquiry in learning research, nearly all of this work has been conducted with rats trained in a double runway. In the present study, we attempted to demonstrate the frustration effect (FE) in monkeys using procedures similar to those in the double-runway studies by Amsel & Roussel (1952) and Wagner (1959). Since the use of an actual runway for monkeys presented practical difficulties, an operant analogue of the double runway was employed. Using a standard monkey test chamber equipped with two retractable levers, we devised a schedule which may be designated "discrete-trials fixed-ratio (left) fixed-ratio (right)," or DT-FRLFRR' In this schedule trials were given in which Ss completed a fixed number of responses on the left lever followed by an equal number of responses on the right lever. The first half of each trial ended with either reward (R trial) or non reward (F trial), and the left and right "halves" of a trial were separated by a mid-trial interval (MTI) during which both levers were retracted. With appropriate cues added, the period of responding on the left lever was considered analogous to the "Alley 1" portion of double-runway trials, the right-lever responding to "Alley 2" responding, and the MTI to "mid-box confinement time". This schedule is a simple extension of the discrete-trials single-FR runway analogue used by Karsh (1964) and Porter and Hug (1965). Me.hod Subjects . The Ss were four adolescent females of the species Macaca speciosa which had previous train-ing on various intermittent reinforcement sched-ules in free-operant conditioning and in discrimination reversal learning in a Y-maze. Apparatus . The apparatus was a Lehigh Valley 1317 oper-ant test chamber equipped with left and right re-tractable levers, 3-color cue lights above each lever, a set of house lights, and a Davis pellet feeder. The Psychon . Sci. , 1965, Vol. 3 feeder dispensed 45-or 97-mg sucrose pellets into a food cup located 3 in below a point midway between the levers, which were 14 in apart. All functions of the test chamber were automatically programmed and recorded. Trial Procedure. Training began with the basic schedule used throughout the experiment-discrete trials in which S completed a fixed number of responses on the left lever followed by an equal number on the right lever. The left lever was inserted into the chamber coincident with the onset of the house lights and a green flashing (10/sec.) left cue light. The latter signal remained on throughout the MTI, during which the left lever was retracted and S was required to wait for insertion of the right lever. The second half of each trial started with the termination of the left cue light, the onset of a white flashing (also 10/sec.) right cue light, and the insertion of the right lever. Five sec. after the final response to the right lever and its retraction, the trial ended with the cessation of both the right cue light and house lights, plunging S into an intertrial period of total darkness which averaged 30 sec. Response measures taken throughout the experiment were the latency of the first lever-press and the time to complete the ratio (time from first to last response), in each half of a trial. These measures were converted to "starting speed" and "ratio speed" measures, respectively, by a reCiprocal (x 100) transformation.