Abstract
In a series of three experiments, the role of attention in visuospatial enumeration was examined in order to better understand differences in the mechanisms involved in subitizing and estimation. In the first experiment, 0 to 9 target items were presented during a brief masked display (150 ms), either with distractors of a different color or without these distractors, to see whether feature-based attention could be used effectively. Accurate performance in the 0 to 3 range was found only in the no-distractor condition, indicating a failure of the subitizing system in the presence of distractors. The data suggests that the approximate number system may be used for the entire range (0 to 9) when distractors are present. However, even the approximate number system was impaired in the presence of distractors, in the form of underestimation for high numerosities. These results suggest that color cannot be used to efficiently segregate multiple targets for enumeration. In a second experiment, distractors of a different color were previewed and to-be-enumerated items appeared as abrupt onsets. In this condition, performance with distractors was no longer impaired in either the subitizing or estimation ranges. Because targets and distractors still differed in color, featural inhibition of the previewed distractor color may have been critical to improved performance rather than onset alone. In a third experiment, targets and distractors shared the same color and could only be identified by differences in time of onset. Performance in the distractor condition remained strong in the subitizing range, suggesting that exogenous attentional mechanisms can provide efficient access to multiple objects. In contrast, impairments in estimation returned when targets were only defined by temporal onset. Results are discussed with regard to the differing roles of attention and feature processing in subitizing and estimation.