Abstract
Prosaccades entail a response to the spatial location of a stimulus, whereas antisaccades require decoupling the normally direct relations between stimulus and response (i.e., 180° spatial transformation). Notably, prosaccades are characterized by an invariant undershooting bias: a finding linked to a movement time and/or an energy minimization strategy (Gillen et al. 2013: Exp Brain Res). It is, however, unclear whether antisaccades are similarly characterized by an invariant endpoint bias. To address that issue, we had participants (N=12) complete pro- and antisaccades to briefly presented stimuli in separate blocks (i.e., proximal and distal) that included an equal number of targets but differed with respect to the magnitude of their eccentricities. In the proximal block target eccentricities were 3, 5.5, 8, 10.5, and 13o, whereas in the distal block eccentricities were 10.5, 13, 15.5, 18, and 20.5o. Results for the proximal block indicated that antisaccades to the 3.0o and 5.0o targets produced larger amplitudes than their matched prosaccade counterparts, whereas the 10.5o and 13.0o targets produced the converse pattern (amplitudes for the 8o target did not differ between tasks). In turn, results for the distal block produced a null between-task difference for the 10.5o target; however, antisaccade amplitudes for the remaining targets were less than their prosaccade counterparts. Thus, pro- and antisaccades were characterized by distinct endpoint strategies. In particular, we propose that the top-down requirements of decoupling the spatial relations between stimulus and response renders the mediation of antisaccades via relational visual cues.
