The lab examines interactions between response selection processes using dual-task procedures. Here, our focus has been to determine whether two unrelated responses can be selected simultaneously. This is a topic of active debate, and a sizeable body of evidence is consistent with the notion that dual-task performance is limited by the existence of a “central bottleneck” that operates on a single task at a time. However, there have been recent demonstrations that, with sufficient practice, the dual-task cost can be eliminated. Our initial work set out to determine whether the absence of dual-task costs could be accounted for by models that included a central bottleneck. We withheld certain combinations of stimuli for the two tasks to determine whether subjects had learned a compound stimulus-response mapping (i.e., learned to treat the two stimuli and two responses as single entities). When the withheld combinations were introduced after several sessions of training, no costs were observed. Moreover, we introduced 50 ms asynchronies between the two stimuli, which produced only minimal changes in performance. Had subjects learned to group their responses or stagger the bottleneck processes to eliminate the dual-task costs, introducing the asynchrony should have strongly affected reaction times. Instead, it appears that individuals are able to perform the two tasks in parallel.
In follow-up work, we investigated whether this dual-task ability extends to different combinations of tasks. Intriguingly, we found that dual-task costs strongly depend on the combinations of the two tasks. When a task using auditory stimuli and requiring manual response was paired with a task using visual stimuli and requiring vocal responses (AM/VV), dual-task costs were robust after 16 sessions of training. In contrast, these costs were essentially eliminated after 10 sessions of training when a task using auditory stimuli and requiring vocal responses was paired with a task using visual stimuli and requiring manual responses (AV/VM). The difference in dual-task costs is particularly striking because the stimuli and responses are highly similar for the two groups and because the single-task reaction times are nearly identical. Thus, this finding is not easily accounted for by any of the dominant theories of response selection. We are presently working to isolate the source of the greater dual-task costs associated with particular combinations of modality-pairings.
Hazeltine, E., Ruthruff, E., & Remington, R. W. (in press). The role of input and output modality pairings in dual-task performance: Evidence for content-dependent central interference. Cognitive Psychology.
Hazeltine, E. & Ruthruff, E. (2005). Modality pairing effects and the response selection bottleneck. Psychological Research.
Ruthruff, E., Hazeltine, E., & Remington, R. (2005). Residual Dual-Task Cost after Practice: What Does it Mean?. Psychological Research.
Hazeltine, E., Teague, D., & Ivry, R. B. (2002). Simultaneous dual-task performance reveals parallel response selection after practice. Journal of Experimental Psychology: Human Perception and Performance, 28(3), 527-545.
David Meyer has proposed EPIC, a powerful model of dual-task control.
Hal Pashler's seminal work brought bottleneck models to the fore of dual-task research.
Eric Ruthruff is a close collaborator who has written extensively about the bottleneck model.
Eric Schumacher performs brain imaging experiments examining the neural substrates of response selection and executive processes.
Torsten Schubert is a friend and colleague at Humbolt University in Berlin who studies similar questions using both behavioral and neuroimaging techniques.
