switch cost in voluntary task switching. We reasoned that voluntary task switching requires the selection of
random task sequences, which necessitates the active inhibition of previously executed tasks. The asymmetric
switch cost was used as a marker for persisting activation. Participants switched voluntarily between color naming
and word naming. One group was instructed to select unpredictable task sequences. The other group was
not instructed to do so. When participants were instructed to be unpredictable, no asymmetric switch cost was
observed. When participants were not instructed to be unpredictable, an asymmetric switch cost was observed.
We conclude that the amount of persisting activation in voluntary task switching is limited and that the switch
cost in voluntary task switching reflects the time needed for reconfiguring the cognitive system from one task
to another rather than the time needed to compensate for persisting activation.