EFFECT OF PARTIALLY COVERING STRING ARRAYS ON PATTERNED STRING PERFORMANCE OF PLATYRRHINE MONKEYS.

Abstract

This study involved a new type of patterned string task in which a delay period was imposed between string presentation and opportunity to respond. In Experiments I and II, six squirrel monkeys were tested on parallel and crossed string problems in a Wisconsin General Test Apparatus. After the parallel or crossed string pattern was viewed by the subject for five seconds, one of three conditions was carried out: (a) a cover was placed over the ends of the strings thereby obscuring the food cup at the end of one string (far cover); (b) a cover was placed over the center portion of the strings allowing the subject to view the food cup but breaking the visual continuity of the strings (middle cover); (c) a cover was placed behind the string pattern thereby not obstructing the subject's view (no cover). After placing the cover according to one of three conditions, a Plexiglas screen was raised either immediately or after a five-second delay thus allowing the subject to respond. Results indicated that squirrel monkeys committed more errors under the far cover condition than under the no cover condition on parallel string patterns. On the crossed string pattern, squirrel monkeys manifested more position preference during the middle cover condition than during the no cover condition or during the far cover condition. In Experiment III, six capuchin monkeys were tested under 11 conditions on crossed string patterns. Eleven conditions were used to vary the lighting and the location of covering during the 12-second delay. Conditions 1 through 10 were conducted either in light or in dark when a cover was imposed during the beginning, middle, or end four seconds or during the beginning or end eight seconds. Regardless of light or dark conditions, capuchin monkeys manifested more correct responses when the full pattern was visible during the last four or the last eight seconds before the response. The result suggests that information received in the early part of the delay interval was used less efficiently than did that in the last part of the delay interval.