Early and late selection: Visual letter confusions in a bar-probe task

This study is concerned with the early vs late selection issue in visual information processing. Early selection theories assert that selective attention operates in a stage containing rapidly decaying precategorical visual information. Late selection theories argue that selective attention operates in a stage containing abstract categorical information. A prediction following from the assumptions of early selection theories, but not from the assumptions of late selection theories, is that in partial-report bar-probe tasks the function relating the number of visual confusions, resulting from misidentification of the target, to probe delays has to exhibit an inverted-U shape. An experiment is reported in which this prediction was tested. A reliable inverted-U function was obtained. The total pattern of results supports the early selection theories for small probe delays, but for the larger probe delays adequate late selection has to be postulated.