PICTORIAL AND VERBAL ENCODING IN SHORT-TERM MEMORY.
Abstract
What is remembered, and the speed of retrieval and systematic errors in retrieval depend on the coding of stimulus information at the time of storage in memory. The study attempted to manipulate the encoding modality of pictorial and verbal material by manipulating the modality of the expected retrieval test (visual or verbal recognition). The stimuli were schematic faces with well-learned nonsense names, constructed so that pictorial similarity correlated negatively with verbal similarity. It was hypothesized that Ss would perform more accurately or faster when encoding modality was the same as the response (test) modality, irrespective of stimulus modality. In Exp. 1, 8 faces were presented on each trial, one was cued, and S attempted to select that one from a set of 2 or 4 faces of names. The data of 2 Ss supported the hypothesis, the data of 2 Ss showed consistent preference for visual coding. In Exp. 2, a single face or name was presented followed by another one; S responded 'same' if they had the same name, 'different' otherwise. Expectancy of test modality was manipulated by having 79% of the second stimuli as faces or 79% as names. Slower reaction times were obtained for all Ss when the second stimulus was the less frequent modality, and RT was less for 'different' responses when the stimuli differed on more than one attribute in the encoding (second-stimulus) modality, regardless of the modality of the first stimulus, or the particular modality of the second stimulus. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 1968
- Accession Number
- AD0683732
Entities
People
- Barbara Gans Tversky
Organizations
- University of Michigan