The effects of visual dimensions on attentional dynamics

Poster Presentation 43.344: Monday, May 22, 2023, 8:30 am – 12:30 pm, Banyan Breezeway
Session: Attention: Features

Rachel Eddings1 (), Aaron Buss1; 1University of Tennessee, Knoxville

In this project, we explored the associations between selective attention and shifting attention across visual dimensions. Twenty-eight college-age students were administered two tasks, the dimensional change card sort (DCCS) and the triad task. In the DCCS, participants were prompted to match a test card by either shape or color to one of two target locations. Between trials, the needs to change attention across dimensions and the presence of stimulus-response conflict was manipulated. For the triad, participants were shown a Reference object and then two test objects - Identity and Distractor. Identity matched the example exactly along one dimension (shape or color) and was maximally different along the other. Distractor was similar to varying degrees (15 to 63 degrees) along both dimensions to the Reference, but not an exact match for either dimension. In the DDCS, participants were faster for color when the dimension repeated over multiple trials, but not for shape. For triad, participants were more accurate the less similar the distractor was along both dimensions. They also showed greater improvement when the relevant dimension of the Distractor was less similar to the Reference, more so when Identity matched for shape than color. Between tasks, the relative dimensional bias was associated such that the bias to attend to color in the triad task was correlated with the repetition benefit for color in the DCCS task. Overall, we are seeing evidence that dimensional attention is impacted by the dimension that is being attended; however, regardless of attending to color or shape, you are using a common mechanism for different attentional functions such as selecting a dimension or shifting attention to a dimension. Attentional functions are not parcellated out by task demands, but rather by dimension.