Dissociable effects of auditory attention switching and stimulus–response compatibility

A level of 65 dB SPL on every ear measured with a calibrated artificial head (with ear coupler) by HEAD acoustics.

Previous results indicated worse performance on response-incongruent trials (Koch et al., 2011). However, congruency was not the focus of the present study, and including congruency in the data analyses did not affect the interpretation of the present data.

Different RT-trimming procedures did not change the results in qualitative terms.

Because RT were overall somewhat (but not significantly) longer with animal names, it is important to exclude that the effect of stimulus material on auditory switch costs was just due to higher costs in trials associated with longer RT. To do so, we calculated the proportional switch costs for each stimulus material by dividing switch costs by RT on repetitions. Importantly, the re-analysis with this proportional measure of switch costs confirmed the effect of stimulus material on switch costs, t(34) = 2.703, p < 0.05.

Since the mapping of relevant gender and relevant ear varied unpredictably from trial to trial, a change in the relevant gender did not necessarily go along with a change of the relevant ear. We assumed that a repetition benefit of the relevant gender was most pronounced if the relevant ear was also repeated. Therefore, we conducted a post hoc analysis to examine potential influences of changes of the relevant ear. RTs and error rates were submitted to separate ANOVAs using transition, S–R compatibility, ear transition, and stimulus material as independent variables. We only report the relevant main effects and interactions regarding ear transition. In the RTs, the main effect of ear transition was significant, F(1,34) = 26.604, MSE = 2.956, p < 0.001, \(\eta_{\text{p}}^{2}\) = 0.44, indicating longer RTs in ear repetitions than in ear switches (1,065 vs. 1,032 ms) and thus costs if the relevant ear repeats (i.e., “ear-repetition costs”) of 33 ms. The interaction of transition and ear transition was also significant, F(1,34) = 10.052, MSE = 28.860, p < 0.05, \(\eta_{\text{p}}^{2}\) = 0.23, indicating larger attention switch costs when the target stimulus was presented to the other ear than to the same ear (80 vs. 40 ms). In the error rates, the main effect of ear transition was also significant, F(1,34) = 6.732, MSE = 0.001, p < 0.05, \(\eta_{\text{p}}^{2}\) = 0.17, indicating higher error rates in ear repetitions than in ear switches (5.4 vs. 4.6 %) and thus also ear-repetition costs of 0.8 %. Additionally, the interaction of ear transition and stimulus material was significant, F(1,34) = 8.785, MSE = 0.001, p < 0.05, \(\eta_{\text{p}}^{2}\) = 0.21, indicating that these ear-repetition costs were much larger for number words than for animal names (2.7 vs. 0.1 %). Ear transitions varied fully independently from gender transitions, so that the auditory attention switch costs are further magnified by ear switches, but they are also present in ear repetitions. However, for the present purpose, it is noteworthy that this influence refers to attention switch costs, whereas the S–R compatibility effect does not seem to be modulated by switch or repetition of ear of target presentation.