The orienting response describes a complex of behavioural, sensory, and physiological changes that are elicited by novel or significant stimuli.  In cognitive terms, orienting has been said to reflect the allocation of processing resources from a limited resource capacity^1,2.  The size of the orienting response, such as that reflected in the amplitude of an electrodermal response, thus reflects the amount of resources used to process the stimulus.

Secondary task reaction time (RT) has also been said to reflect the amount of resources used to process a stimulus³.  Secondary RT is measured by asking subjects to perform two tasks simultaneously: (a) the primary orienting task, and (b) a RT task.  The size of the slowing in secondary RT is taken to reflect the amount of processing resources required by the primary orienting task.

Consistent with resource allocation models, a lengthening of secondary RT during stimuli that elicit increased electrodermal orienting has been demonstrated by several investigators⁴.  However, a paradigm which employs a discrimination and counting task results in a dissociation between these measures^5,6,7,8,9,10.

Dawson et al. (1989)⁵ required subjects to count the number of longer-than-usual tones (7 s vs. 5 s) presented to one ear (task-relevant) and to ignore tones presented to the other ear (task-irrelevant).  Simultaneously, subjects made a speeded RT response to a visual probe.  Electrodermal orienting was larger during task-relevant tones than during task-irrelevant tones.  However, secondary RT to probes presented 150 ms following tone onset was slower during task-irrelevant tones than during task-relevant tones.  Thus, electrodermal orienting suggested that more resources were allocated during task-relevant tones, but secondary RT suggested that more resources were allocated during task-irrelevant tones.

Using a similar paradigm, Siddle et  al. (1996)¹⁰ required one group of subjects to count the number of longer-than-usual task-relevant tones and another group to simply count the number of task-relevant tones irrespective of its duration.  Again, electrodermal orienting was larger during task-relevant tones in both groups.  Secondary RT was slower during task-irrelevant tones than during tak-relevant tones only at a 250 ms probe position when subjects counted the number of task-relevant tones.  The same difference was reliable only at a 150 ms probe position when subjects counted the number of longer-than-usual task-relevant tones.

The dissociation effect is problematic for resource allocation models of orienting.  It also places doubt on the use of autonomic orienting and secondary RT as reliable measures of processing resources when investigating cognitive workload demands in experimental and applied settings.  The present study replicated the manipulation of counting task requirements as done by Siddle et al. (1996), but used visual orienting stmuli instead of auditory orienting stimuli.  In addition, only secondary RT was measured because the basic dissociation effect with visual orienting stimuli has been previously demonstrated^8,9.
 

Correspondence to: D.Neumann@ccrn.uwa.edu.au