s1. Earlier studies have shown that changes in the difficulty of sensory discrimination in a choice reaction time task result in a prolongation of the peak latency for several components of the long-latency event-related potential (ERP). With the use of the technique of response-locked averaging, we have previously shown that manipulation of the difficulty of sensory discrimination also affects response execution as assessed by the interval between the ERP and onset of the response. In the present paper we examine the hypothesis that changing the compatibility of the responses may also affect the difficulty of the discrimination, as well as the execution of the response, as assessed by the interval between stimulus onset and the ERP. Such an effect of response compatibility would provide further evidence for the close integration of motor and sensory processes in the performance of choice reaction time tasks. 2. We continuously recorded the electroencephalogram (EEG) from the scalp and the electromyogram (EMG) from the responding muscles in both compatible and noncompatible visual choice reaction time tasks. In the compatible task subjects responded to a lateralized visual stimulus with the hand ipsilateral to the stimulus, whereas in the noncompatible task they responded with the contralateral hand. EEG and EMG responses were analyzed and averaged off-line, aligning the waveforms by either stimulus onset (stimulus-synchronized averages) or response onset (response-synchronized averages), and averaged separately for both correct and incorrect response outcomes. 3. Response times were significantly faster for frequent stimuli than rare stimuli and were significantly faster to rare stimuli in the compatible than the noncompatible condition. In responses to the frequent stimuli (where both hands were required to respond), the right hand was slightly but consistently faster than the left hand. The right hand also accounted for 83% of the errors made. 4. Stimulus-synchronized and response-synchronized ERPs to either frequent or rare stimuli had a similar appearance for correct responses in both the compatible and noncompatible conditions. The coupling of the response to the ERP for the rare stimuli, however, was different for the two conditions: the response occurred later relative to the ERP components in the response-synchronized average in the noncompatible condition compared with the compatible condition. By contrast, the coupling of the ERPs to the onset of the stimulus was the same in the two conditions. 5. Stimulus-synchronized averages for error responses in which the rare tone was mistaken for a frequent tone showed early sensory processing (as judged by the ERPs) that was similar to that of correct responses to the rare stimuli. After the apparent positive (P2) component of the cerebral response, however, the processing differed, with a superimposed broad negativity possibly reflecting awareness by the subject that a mistake had been made. By contrast, the response-synchronized averages for these error trials appeared like those to frequent stimuli, with the response being coupled to the P2 component of the cerebral response. 6. These results suggest that response compatibility affects response selection processes but does not alter sensory discrimination. However, despite the similarly tight coupling of the response to the ERP in both the compatible and noncompatible conditions, the response occurred later relative to the ERPs in the noncompatible condition. This suggests that different components of the ERP are responsible for triggering the response in different circumstances. Our observations on the error trials suggests that the decision to respond (on these trials) is based on the occurrence of cerebral events that are evoked by either rare or frequent stimuli, whereas this decision (on correct response trials) is based on cerebral events elicited only by the rare stimuli.
The Rapid Extraction of Gist—Early Neural Correlates of High-level Visual Processing