Constructing Emoticon Meaning from priming and high dimensional modeling.
Mentor:Curt Burgess, Professor of Psychology, University of California, Riverside
A semantic priming experiment was conducted with word primes and emoticon targets. The strength of word-emoticon relationships was determined by a norming study. There were three word – emoticon relationships, e.g., primary, happy :) or secondary, blushing :) or unrelated, bitter :) . Primes were presented after a fixation point for 500 ms, followed by the emoticon target. Subjects made semantic relatedness judgments. Response times to emoticon targets preceded by a primary meaning were faster than unrelated targets. However, response times to secondary meaning targets were longer than the unrelated targets. These results are discussed from a multiple meaning activation viewpoint (i.e., Simpson & Burgess, 1985) and in the context of high-dimensional memory models such as HAL (Burgess & Lund, 2000).