ACQUISITION AND MAINTENANCE OF MATCHING WITHOUT A REQUIRED OBSERVING RESPONSE¹

Abstract

Two experiments demonstrated that matching-to-sample performance was improved when an explicit observing response was required to the sample stimulus. The first experiment demonstrated that fewer training sessions were required to establish matching with than matching without such a response. The second experiment demonstrated the dependence of established matching accuracy upon this observing response and the development in two of three subjects of a new overt observing response even when none was required by the experimenter.The matching-to-sample procedure is a conditional discrimination procedure in which a standard (ST) or sample stimulus designates which of a number of comparison (CO) stimuli is correct . Analyses of behavior under this sort of conditional discrimination procedure have generally presumed the ST stimulus to exercise a logically and temporally different role than the CO stimuli. Skinner (1950) described matching-to-sample as "choosing red after being stimulated by red." Cumming and proposed that the ST exercises an "instructional" role controlling the strength of response to each of the CO stimuli. However, an alternative and more parsimonious analysis of matching-to-sample is available, and it seems worthwhile to eliminate this analysis before accepting the other. The terms of this alternative analysis become clearer once the apparatus employed in many matching-to-sample studies is described. Gen-erally, matching-to-sample is studied in the laboratory using relatively few alternative stimuli (e.g., red, green, and blue lights for color matching). Often, a three-stimulus display is used, the center display presenting the ST and the two side displays presenting the COs. Given the few stimuli and the three-key display, it is possible that the entire three- stimulus display might be acting as a unitary stimulus controlling the strengths of the responses "peck right" and "peck left" for a subject exhibiting high matching-to-sample accuracy. Depending upon the distance separating the three keys, a "unitary stimulus" analysis becomes possible: where keys are very close together, training to match-to-sample might merely be strengthening the response to the "larger" color, i.e., the side-keycenter-key area that is of homogeneous color.Although the unitary-stimulus analyses are more parsimonious than the instructionalstimulus analyses, they would be difficult to maintain if accuracy in matching-to-sample was shown to be dependent upon a separate observing response (Wyckoff, 1954) to the ST.In many matching-to-sample experiments such a separate response to the ST has been required (e.g., Riesen and Nissen, 1942;Ferster, 1960; METHOD SubjectsSix male White Carneaux pigeons were maintained throughout the experiment at 80% of their free-feeding weights. Three 435 1968, 11,[435][436][437][438][439][440][441] NUMBER 4 (JULY)

Sustained attention: Stimulus control determined by schedule of cue production interacting with cue duration

“…Variations in both CPR and OR rates across birds appear to be quite uncorrelated with concurrent variations in control by 0.2-sec SDs (Table I). Moreover, previous reports show increased stimulus control accompanying increased CPRs (D'Amato, Etkin, & Fazzaro, 1968) and ORs (Eckerman, Lanson, & Cumming, 1968;Farthing & Opuda, 1974). In contrast, the present results show increased CPRs and ORs accompanied by decreased stimulus control.…”

Enhanced learning of new discriminations after stimulus fading

“…In Condition C, it follows that subjects should observe the relevant features of the stimuli more readily after fading than after traditional discrimination training and, thus, learn C in fewer trials after fading. Evidence supporting such an analysis linking acquisition rate to likelihood of observing behavior is provided by D ' Amato, Etkin, and Fazzarro (1968), Cohen , Looney, Brady, and Aucella (1976), and Eckerman, Lanson, and Cumming (1968). Further confirmation, however, would depend upon direct measurement of observing behavior in each phase of a replication of the current experiment.…”

Attentional enhancement in matching-to-sample: Facilitation in matching acquisition by sample-discrimination training

“…According to Cumming et al, subjects learn a unique coding response to each sample during MTS. Choices between the comparisons are then mediated by the coding responses rather than by the physical characteristics of the samples alone (see also Berryman, Cumming, Cohen, & Johnson, 1965;Eckerman, Lanson, & Cumming, 1968;Maki, Gillund, Hauge, & Siders, 1977). According to this argument, the rate at which birds acquire MTS will be determined by how quickly they learn the coding responses to each sample stimulus (Carter & Werner, 1978), which in turn should be governed by how rapidly they come to attend to the relevant features of the samples.…”