Peer-reviewed article
Probing the use of visual search as a tool to enhance reading speed.
Francesco Carabba, Chiara Borsati, Giulia Fontana, Eric Altieri, Michele Vicovaro et al. ·
Summary
Written by AI, not yet AICS's view ·Key points
Limitations
Abstract
The reading mechanism is associated with perceptual, attentional, and oculomotor processes that partly overlap with functions attributed to the magnocellular-dorsal system. On this basis, we examined whether a home-based visual-search training program, without phonological, orthographic, or reading components, could influence reading-related performance in adult typical readers. Participants were assigned to one of three training versions that differed in the visual features of Gabor patterns to be detected (orientation, phase, or configuration), or to a no-training control group. Training performance indicated that the three visual-search versions differed in difficulty: The orientation task was relatively easy, the phase task was the most perceptually demanding, and the configuration task was intermediate. Regression analyses controlling for pretest performance showed lower adjusted posttest reading times, relative to the no-training group, for the phase group in word reading and lexical decision, and for the configuration group in nonword reading. No consistent training-related advantage was observed for accuracy. These results suggest that demanding visual-search training may be associated with changes in some timed reading-related outcomes, but the effects were outcome-specific and should be interpreted as preliminary. Further work is needed to disentangle task difficulty, search mode, and training dose, and to determine whether similar effects generalize to children or to readers with dyslexia. OPEN PRACTICES STATEMENT: The raw trial-level data for all experimental conditions are publicly available in the OSF repository ( https://doi.org/10.17605/OSF.IO/XTY4S ). The repository also contains the code for the analyses described in the Sensitivity Analysis and Type I Error Rate section. The experiment and analyses were not preregistered.