| The models of memory used in consumer behavior research are based on the assumption that the internal representation of information is in the form of abstract codes. According to these models, the mode or input sense through which information enters human memory is not encoded because it is only meaning that is represented in memory, and mode of input is not part of its meaning. It was argued in this dissertation that internal representations are not abstract, but are mode specific. That is, memory codes are isomorphic with external stimuli. According to this view, auditory information would be represented as an auditory memory trace and visual information would be represented as a visual memory trace.; The predictions of mode specific view of memory were contrasted with the predictions of abstract representation view of memory in a learning experiment. Subjects were exposed to target information either twice in one mode (auditory or visual) or once in each of the two modes (auditory and visual). Based on the findings in learning literature that support the superior retrievability of information that is represented in multiple paths in memory, it was hypothesized that mixed mode presentation should result in greater recall than single mode presentation. This hypothesis was supported both with immediate and delayed recall tasks.; The mode specific view of memory was then used to predict conditions under which pictures would be recalled better than words. It was argued that when words are processed so as to form visual images, then words and pictures would be recalled equally well. If words are processed without generating visual images, then pictures would be recalled better than words. These hypotheses were supported in a delayed recall task.; It was concluded that internal representations are mode specific. It follows from this conclusion that greater learning can be achieved by presenting information in multiple modes than in one mode, even after controlling for the number of exposures. Another conclusion from this research was that pictures would be retrieved better than words as long as people do not form visual images when they process verbal information. |
