Oral Presentation (3) February 19

Japanese macaques' short-term memory for serial items: An item selection strategy based on the least memory strength

Ohshiba Nobuaki
Baika Women's College, Japan

Two Japanese macaque (Macaca fuscata) learned 2-item delayed matching-to-samples (DMTSS), and one of them learned 3-item one. An apparatus measured 75 ~ 75 ~ 75 cm, whose front panel housed a CRT with a touch screen installed. A personal computer controlled the presentation of stimuli, recording of the subjects' responses, and delivery of food rewards. In the training, three different colored rectangles (red, green, blue, or yellow) appeared one by one (memorization phase), and then those three rectangles were presented simultaneously (recall phase). The subject was required to touch those three items following the order of presentation in the memorization phase. When the subject touched a appropriate rectangle, it disappeared immediately. A trial in which the subject selected a wrong item was terminated and repeated until correct choices were done. The configuration of the stimulus items at the recall phase was randomized and counterbalanced across trials. A session was consisted of 96 non-correction trials. The criterion for acquisition was set at 85% in two consecutive sessions. After training was complete, a probe test was administered, in which one of the three rectangles was replaced with another that did not appear at the memorization phase. Results showed that the later serial position the replacing item occupied, the lower the accuracy became, and that the subject tended to select the replacing item in earlier serial positions than expected. Examining in detail the subject's choice patterns suggested that he select the items associated with the smallest memory strength. A subset test was conducted in which only two of the three items that appeared at the recall phase were presented at the memorization phase. The subject's performance was better when the first and the third items were presented than otherwise, revealing a symbolic distance effect. These results were also consistent with the memory-strength-guided strategies.

 

OHSHIBA, Nobuaki
Department of Human Sciences,
Baika Women's Collegue
2-19-5, Shukunosho, Ibaraki-shi, Osaka
567-8578, Japan
ohshiba@baika.ac.jp