Abstract
We previously documented an index of object-based attention called the shift direction anisotropy (SDA), a measure of the relative advantage of horizontal (compared to vertical) shifts of object-based attention (Barnas & Greenberg, 2016, 2019). The SDA is typically larger when target stimuli appear in a location across the visual field meridians from the initial locus of attention. Based on these results, we hypothesized that the horizontal shift advantage stemmed from more efficient attentional deployment during interhemispheric (horizontal) versus intrahemispheric (vertical) orienting, primarily due to support from independent pools of attentional resources (Barnas & Greenberg, VSS 2019). Here, we used fMRI of early visual cortex to test this hypothesis. During this experiment, subjects were presented with an L-shaped object in either the upper-left or lower-right corner of the visual field and a centrally presented arrow would cue subjects to attend to the vertex of the object. Subjects performed a target detection task (60% valid, 20% invalid, 20% catch) at one of five possible target locations, with distractors appearing at the four non-target locations. On invalid trials, the target was either horizontally (invalid-horizontal) or vertically (invalid-vertical) displaced from the vertex. Likewise, targets could be located either across the respective visual field meridian (invalid crossing) or not (invalid non-crossing). Brain activation time-locked to cue onset showed that attention was allocated throughout the L-shaped object to both crossing and non-crossing invalid-vertical and invalid-horizontal locations. However, a Bonferroni-corrected within-subject t-test revealed smaller magnitude attentional modulation at crossing invalid-horizontal target locations compared to crossing invalid-vertical locations (t(755) = -3.07; p < 0.005). Thus, the horizontal shift advantage is not likely explained by a post-cue prioritization of invalid horizontal locations but could instead be due to the preferential allocation of attentional resources horizontally prior to cue onset as a result of reduced competition.