An Eye Movement Study of the Fan Effect: Evidence for a Multiple-Access Model

Two studies are reported looking at the relationship between eye movements and the fan effect. The first replicates Anderson (1974) and the second replicates Radvansky, Spieler, and Zacks (1993). These experiments require participants to judge whether they had studied a sentence asserting that a subject (person or object) is in a location. Neither experiment found an effect of fan of either subject or location on the duration of the first fixation but there is an effect of both variables on number of fixations and duration of subsequent fixations. The fan effects are the same, independent of the order in which the terms (subject and location) are presented or fixated. This is evidence for a multiple-access model that requires both terms be encoded in order to retrieve the sentence trace rather than a single-access model in which retrieval is made from one term and checked against the other. Two models are fit to these data under the assumption that, after the first fixation there is a race between a retrieval process that generates a response and an asynchronous process that generates further eye movements.