Abstract
The stimulus prefix effect, in which a non-recallable item placed at the beginning of a to-be-remembered string of items creates interference, was studied using a 1-s simultaneous visual method of presentation. Two variants on the standard prefix condition were used; in one, the prefix, o, was presented in a different colour from the memory string of seven digits and, in the other, the prefix consisted of three identical zeroes. The three prefix conditions were compared with two control conditions in which either seven or eight digits were presented. No evidence of prefix effects was found in the recall data over serial positions 1–4, for which retrieval was highly efficient, and it was concluded that the action of response set may exclude the prefix from a response buffer (c.f. Morton, 1970). There was also no evidence that the specialized prefixes led to an attenuation of the interference found in the remaining positions. The latter finding is considered to be contrary to a unit-formation hypothesis (Kahneman, 1973) but is consistent with a mode of operation of the logogen model (Morton, 1970).
Get full access to this article
View all access options for this article.