The Effect of Saccadic Response Time on the Reference Frame of Inhibition of Return

We have previously shown that oculomotor inhibition of return is predominantly retinotopic (gaze-centred) immediately following a saccadic eye movement, and predominantly spatiotopic (world-centred) at longer post-saccadic delays (Mathôt & Theeuwes, 2010a). This claim has been contested by Hilchey et al. (2012), who re-analysed our data after removing very fast saccades (<100ms) and observed that removing these saccades attenuated the effect of interest. Here we report an elaborate re-analysis of that same dataset. We show that recent conflicting reports on the reference frame of inhibition of return can likely be reconciled by a previously unrecognised effect of saccadic response time. Oculomotor inhibition of return is predominantly retinotopic at short post-saccadic delays, particularly for fast saccades. In contrast, oculomotor inhibition of return is predominantly spatiotopic at longer post-saccadic delays, particularly for slow saccades. With respect to previous studies, differences in paradigm and trial exclusion criteria may have resulted in different distributions of saccadic response times, thus leading to different conclusions regarding the reference frame of inhibition of return. Finally, we argue that the effects reported here are not specific to inhibition of return, but reflect general mechanisms that are observed across a wide range of attentional, visuomotor, and visuotactile phenomena.