Answers are grounded in this paper and cite their sources.
In 4 experiments, Ss were instructed to circle each letter t found during the course of reading. Contrary to the unitization reading model (Drewnowski & Healy, 1977), the word the did not show a high proportion of detection errors when a normal reading pattern was disrupted with (a) a right-to-left reading direction (Experiment 1), (b) a large interword space preceding target locations (Experiments 2 and 3), (c) a large interword space after target locations (Experiment 3), or (d) a 5-letter viewing window (Experiment 4). As a result, a parafoveal-processing hypothesis was implicated that describes the large number of errors normally found in the word the because of a reading process that often skips parafoveally identified words. Consequently, a refinement of the unitization model was proposed
No takes yet. Share an insight, caveat, or question.
Hadley et al. (1991) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: