Transformation of Visual Memory Revealed by Latency of Location-Specific Matching
Abstract
This research addressed extent to which the processing of items with well-learned names differs from that of items without well-learned names. To address this question, a new paradigm was created that does not require items to have names and does not require a naming response. In many short term visual memory paradigms, a bar marker designates a display location and the task is to name the item in that location. In the new paradigm the marker is replaced by a comparison item positioned above a display location. Instead of naming the item in the location, the subject's task is to say yes if the item in the indicated location is identical to the probe, and no otherwise. Data were collected from six subjects in a large (75,000 trial) experiment. A multiple regression model was formulated and applied to these data to compensate for non-orthogonalities in the experimental design that could not be avoided. Several models of the processing underlying performance were also generated and tested. Regression and modeling results were then used in data analyses. Analyses indicated that (1) changes in the underlying representations for forms with and without well-learned names take place; (2) these changes are qualitatively similar for the two kinds of material; and (3) the representation of items without well-learned names is searched more slowly than the representation for named items. Keywords: Visual image memory.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 15, 1988
- Accession Number
- ADA198712
Entities
People
- David L. Turock
Organizations
- University of Pennsylvania