Abstract
Attempts to measure the termination of the icon by a patterned mask have been criticized on the grounds that S may confuse icon disappearance with mask arrival. Exp. I shows that Ss successfully detect actual temporal overlap of target and mask of as little as 10 msec. Exp. II shows that any information extracted from this target-mask integral is negligible compared to that extracted from targets alone. These results vindicate estimates of iconic readout rate which employ backward masking to control stimulus information availability.
Get full access to this article
View all access options for this article.
