Visual backward masking as measured by voice reaction time

Eriksen, Charles W.; Eriksen, Barbara A. · 1972 · Crossref

DOI: 10.3758/bf03212835

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates visual backward masking by utilizing voice reaction time (RT) as the dependent variable, rather than the traditional measures of percent correct identification or detection. The research aims to distinguish between two competing theoretical explanations for masking: the "erasure" hypothesis, which posits that a mask replaces the target in visual memory, and the "integration" hypothesis, which suggests that the mask degrades the target stimulus through perceptual superposition. By measuring latency, the authors sought to determine whether masking is an all-or-none phenomenon or a continuous, gradually decreasing effect as the interstimulus interval (ISI) increases. The experiment involved three subjects who viewed capital letters (A, E, R, S) presented for a duration of 350 msec. A black ring mask, varying in size (small, medium, large), was presented at various stimulus onset asynchronies (SOAs) ranging from 0 to 300 msec. Crucially, the target letter remained visible when the mask appeared, a design intended to preclude the possibility of the target being erased from the visual store. Subjects were instructed to vocalize the target letter as quickly as possible, with RT recorded in milliseconds. The smallest ring was expected to produce masking effects, while larger rings served as controls for general second-stimulation effects. The results demonstrated that vocal RT was reliably increased when the target was simultaneously encircled by the small masking ring, with the effect diminishing as the SOA increased. The masking function was monotonic, with RT returning to baseline levels at an SOA of 150 msec. In contrast, the larger control rings produced smaller, nonmonotonic effects on RT. Analysis of RT variances revealed no increase in variance at mid-range SOAs, which would have been predicted by the "psychological moment" hypothesis (a variant of integration theory suggesting masking occurs on a subset of trials). Instead, the variance decreased consistently with mean latency, supporting the view that masking is a progressively diminishing impairment on every trial. The findings are significant because they are incompatible with the erasure interpretation of backward masking; since the target remained visible, the increased processing time must result from degradation of the iconic representation rather than replacement. The data support an integration account where the mask interferes with the processing of the target trace, with the degree of interference decreasing as more time elapses for the target to be processed before the mask appears. This study establishes RT as a sensitive measure for masking phenomena and provides evidence that masking involves a continuous degradation of perceptual information rather than a discrete erasure event.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-25; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• time perception