Response selection and response execution in task switching: Evidence from a go-signal paradigm

The present study used a go/no-go signal delay (GSD) to explore the role of response-related processes in task switching. A go/no-go signal was presented at either 100 ms or 1,500 ins after the stimulus. Participants were encouraged to use the GSD for response selection and preparation. The data indicate that the opportunity to select and prepare a response (i.e., long GSD) resulted in a substantial reduction of task-shift costs (Experiment 1) and n - 2 task-repetition costs (i.e., backward inhibition; Experiment 2) in the current trial. These results suggest that interference from the preceding trial can be resolved during response selection and preparation. Furthermore, the shift costs and the n - 2 repetition costs after no-go trials with long GSD (i.e., response selection but no execution) were markedly smaller than after go trials. These findings suggest that the interference that gives rise to shift costs and n - 2 repetition costs is related not solely to response selection but also to response execution. Thus, the present study demonstrates dissociable contributions of response selection and response execution to interference effects in task switching.