Mechanisms of short-term memory for serial order.
Item
-
Title
-
Mechanisms of short-term memory for serial order.
-
Identifier
-
AAI9304737
-
identifier
-
9304737
-
Creator
-
Sliwinski, Martin John.
-
Contributor
-
Adviser: Elizabeth Bostock
-
Date
-
1992
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Psychology, Experimental | Psychology, Developmental | Gerontology
-
Abstract
-
Three experiments used a task that measured probed recall of digit strings to explore factors that underlie short-term memory for serial order. Associative and positional probes were used to measure recall of order information. A version of Buschke's (1963a) missing scan probe was used to measure recall of item information. Experiment 1 analyzed data from 24 young adults (mean age = 22). Experiment 2 analyzed data from 4 young subjects tested over several sessions to determine whether the pattern of results observed in the aggregated data from Experiment 1 describes the performance of individual subjects as well. Experiment 3 compared recall and serial order effects of young (ages 18-26) and aged (ages 65-80) subjects. Typical serial position curves were observed for all probe types, for two levels of distractor difficulty and for two list lengths. While the comparison of serial position curves from the probed recall of order and item information did not reveal any significant differences, there was a considerable difference in recall accuracy. This difference between recall accuracy of item and order information was exaggerated in the aged, suggesting a specific impairment of memory for order in aged. Consistent with previous research, positional codes were found to be more basic than associative codes for serial order memory. However, there was some indication that additional encoding operations, such as chunking, came into play for the long stimulus lists. Application of a quantitative model (Estes' perturbation model, Lee & Estes, 1981) which relies on positional coding of serial order met with mixed success. The model fit the serial position data from the positional probe well, especially for the short stimulus list. However, the model had difficulty predicting recall for associative probes and the long stimulus lists, providing a further indication that there are important aspects of serial order memory not captured by an exclusively positional account.
-
Type
-
dissertation
-
Source
-
PQT Legacy CUNY.xlsx
-
degree
-
Ph.D.
