The mechanisms underlying increased dual-task costs in the comparison of modality compatible stimulus-response mappings (e.g., visual-manual, auditory-vocal) and modality incompatible mappings (e.g., visual-vocal, auditory-manual) remain elusive. To investigate whether additional control mechanisms are at work in simultaneously processing two modality incompatible mappings, we applied a transfer logic between both types of dual-task mappings in the context of a mental fatigue induction. We expected an increase in dual-task costs for both modality mappings after a fatigue induction with modality compatible tasks. In contrast, we expected an additional, selective increase in modality incompatible dual-task costs after a fatigue induction with modality incompatible tasks. We tested a group of 45young individuals (19–30 years) in an online pre-post design, in which participants were assigned to one of three groups. The two fatigue groups completed a 90-min time-on-task intervention with a dual task comprising either compatible or incompatible modality mappings. The third group paused for 90 min as a passive control group. Pre and post-session contained single and dual tasks in both modality mappings for all participants. In addition to behavioral performance measurements, seven subjective items (effort, focus, subjective fatigue, motivation, frustration, mental and physical capacity) were analyzed. Mean dual-task performance during and after the intervention indicated a practice effect instead of the presumed fatigue effect for all three groups. The modality incompatible intervention group showed a selective performance improvement for the modality incompatible mapping but no transfer to the modality compatible dual task. In contrast, the compatible intervention group showed moderately improved performance in both modality mappings. Still, participants reported increased subjective fatigue and reduced motivation after the fatigue intervention. This dynamic interplay of training and fatigue effects suggests that high control demands were involved in the prolonged performance of a modality incompatible dual task, which are separable from modality compatible dual-task demands.
