Programming strategies for rapid aiming movements under simple and choice reaction time conditions

Abstract

Increases in reaction time (RT) as a function of response complexity have been shown to differ between simple and choice RT tasks. Of interest in the present study was whether the influence of response complexity on RT depends on the extent to which movements are programmed in advance of movement initiation versus during execution (i.e., online). The task consisted of manual aiming movements to one or two targets (one- vs. two-element responses) under simple and choice RT conditions. The probe RT technique was employed to assess attention demands during RT and movement execution. Simple RT was greater for the two- than for the single-target responses but choice RT was not influenced by the number of elements. In both RT tasks, reaction times to the probe increased as a function of number of elements when the probe occurred during movement execution. The presence of the probe also caused an increase in aiming errors in the simple but not choice RT task. These findings indicated that online programming was occurring in both RT tasks. In the simple RT task, increased executive control mediated the integration between response elements through the utilization of visual feedback to facilitate the implementation of the second element.

Acknowledgments

We thank Stuart Klapp for his helpful comments on an earlier version of this article.

Notes

It could be argued that more accurate responses contain more submovements and that simple RT could increase as the number of submovements increase. However, the occurrence of submovements has been said to be due to feedback-based corrections during movement execution and not programming prior to response initiation.

It is possible that the two-target movement time advantage observed in the present experiment was due to a speed–accuracy trade-off since target hit rates at the first target were less in the two- than in the one-element conditions. However, there were no differences in target hit rates at the first target between simple and choice RT conditions, and, therefore, differences in reaction time and movement time between RT tasks cannot be explained by speed–accuracy trade-offs.

It is also possible that visually based online corrections in the simple RT condition disrupted the integration between the first and second elements and that this led to a reduction in the two-target advantage. Such a disruption would probably have led to an increase in pause times between the first and second elements. However, since pause times did not differ between simple and choice RT conditions, it seems that movements were integrated during execution of the first element in the simple RT condition.

A secondary task that involved a manual response was used since higher degrees of interference have been reported for dual tasks involving two manual responses than for those involving a manual and a vocal response (McLeod, 1978; Pashler, 1990).

Ketelaars et al. (1999) have also demonstrated that interference from a secondary task prior to movement reversal occurs in a narrow time window.

It is also possible that the probe task interfered with visual control processes during the first element, and this led to increased error rates at the first target.