In dual-task (DT) situations task-order coordination (TOC) processes are required to schedule the bottleneck-processing of component tasks as had recently been shown by observations suggesting increased TOC-costs in DT situations comparing variable against fixed order of component tasks.
The current study investigated whether TOC-processes can be improved by training and whether such improvements can operate on a stimulus-unspecific, i.e. higher-order, processing level or not.
For that purpose, participants trained the same two visual-manual component tasks in four different groups, each of which having various requirements on TOC-processes. In pre- and post-training sessions, we compared the TOC-costs between the different training groups for situations with unchanged and with untrained stimuli to test for (stimulus-unspecific) training-related improvements.
Results demonstrate that TOC were improved (i.e., reduced TOC-costs) for participants who trained flexibly to change processing order compared with participants who did not train such flexibly changing order processing. Importantly, the improvements could be shown for situations with new, untrained stimuli, which is indicative for the operation of training on a stimulus-unspecific processing level.
The results are consistent with the Efficient Task Instantiation (ETI) model, which states that two task sets can be conjointly instantiated in working memory after practicing the two corresponding component tasks in DT situations. We assume that training of flexibly adapting the processing order of two component tasks can lead to an efficient, conjoint instantiation of both potential task orders on a higher-order stimulus-unspecific level which can guide efficient task-order scheduling independently on the specifically presented stimuli.