Effect of Criterional Learning-Set Training Where Problem Objects are Encased in Plexiglas Bins

Abstract

Extended criterional learning-set (LS) training prior to test on a series of 6-tr/LS problems provided a significant, but transitory, advantage for Ss trained in a situation where problem objects were encased in Plexiglas bins. Criterional training appeared to compensate, at least partially, for those characteristics of this method of problem presentation which render squirrel monkey Ss learning-setless when given just 6-tr/LS training (Rumbaugh, Ternes, & Abordo, 1965). Rumbaugh, Ternes, and Abordo ( 1965 ) have reported that, when squirrel monkeys were trained in a 6-tr/problem learning set (LS; Harlow, 1949) situation where problem objects were presented to Ss either on Plexiglas bases or in Plexiglas bins, secured to the stimulus tray by sets of tracks, Ss remained learningsetless even after 500 problems. The purpose of the present study was to compare performance of the Plexiglas-bin-training group of that study with that of another group given extensive criterional LS training, prior to encounter of a set of 200 6-tr/problems, and to determine whether by so doing this kind of training might effectively offset those conditions which had deterred the formation of LS in the former group. To the degree that criterional training might, for example, compensate for the relatively low-level learning ability of the squirrel monkey (Rumbaugh & McQueeney, 1963) and/or reduce the difficulty of the learning task (Stevens, 1964) the development of LS should be facilitated.Ss were four experimentally naive, male squirrel monkeys whose characteristics, maintenance conditions, and adaptation procedures in every way paralleled those of Ss in the Rumbaugh, Ternes, and Abordo study; the test apparatus and stimulus tray used for their training was that used in their Plexiglas-bin-training condition.Ss first learned to criterion the square vs circle problem at rates which proved very equivalent to those of Ss in the Plexiglas-bin group, the t test for mean scores of the two groups falling far short of significance ( t = .09, df = 7, p > .05 ) . Ss then commenced criterional-LS training consisting of 1800 trials (equivalent in trial count to 300 6-tr/LS problems) in which each of a series of object-quality discrimination problems was learned to the intraproblem criterion of 20 responses correct wichin a span of 25 trials. They then received the set -----

Do Some Taxa Have Better Domain-General Cognition than others? A Meta-Analysis of Nonhuman Primate Studies

“…Second, because tests directly measure performance, not intrinsic abilities, contextual confounds can obscure true differences. In procedure 15 , where several apes performed poorly, the relevant objects were encased in plexi-glass bins, a situation which was later shown to be highly distracting for some species (Rumbaugh & McCormack, 1967; see also Ternes, Abordo, & Rumbaugh, 1965). The final, and most important, reason that taxonomic overlap does not weaken our claim of taxonomic differences is that evolutionary theory fully expects that developmental and genetic differences will produce substantial phenotypic variation.…”