Shiyang Ren¹ (), Aysun Duyar¹, Marisa Carrasco¹; ¹New York University

[Background] Endogenous temporal attention is the voluntary deployment of attention to a specific moment, whereas temporal expectation is based on the temporal probability of event onsets. Endogenous temporal attention interacts with temporal expectation. Benefits of temporal attention are highest when stimuli appear early and decrease as onset is delayed. It is unknown whether attentional benefits depend on the timing of the prior trials–i.e. whether stimulus onsets affect observers’ expectations in the subsequent trials and result in differential allocation of attention. [Goal] To investigate potential sequential effects of stimulus onset on the interaction between temporal expectation and attention. [Methods] Two oriented Gabors were presented sequentially at the fovea. Observers performed a 2AFC orientation discrimination task. Participants were instructed to attend to either the first (T1) or second (T2) stimulus (valid), or to both stimuli (neutral). T1 could appear within a temporal window of 1200-1600 ms (expected moment: 1400 ms), and T2 appeared 250 ms after T1 (expected moment: 1650 ms). A response cue indicated the target observers had to discriminate. [Results] There was a significant interaction between temporal attention and expectation. Temporal attention improved performance, and the improvements decreased as the stimulus onset was delayed. The timing of the preceding trial –early, expected, or late onset– did not affect the magnitude of the interaction between expectation and attention in consecutive trials. [Conclusions] Temporal attention benefits decreased as a function of stimulus onset delay. Once observers attended to an early moment, they could not reallocate attention to a later moment, regardless of the timing of the stimulus onset in the preceding trial. Stimulus timing within a trial did not affect temporal expectations in the subsequent trial. Thus, the interaction between temporal attention and expectation was consistent regardless of the trials' sequential order, suggesting that observers optimize performance in each trial.

Acknowledgements: NIH R01-EY019693 to M.C.