Elsevier

Neuropsychologia

Volume 38, Issue 5, 1 May 2000, Pages 701-711
Neuropsychologia

An ERP study of the temporal course of the Stroop color-word interference effect

https://doi.org/10.1016/S0028-3932(99)00106-2Get rights and content

Abstract

The electrophysiological correlates of the Stroop color-word interference effect were studied in eight healthy subjects using high-density Event-Related Potentials (ERPs). Three response modalities were compared: Overt Verbal, Covert Verbal, and Manual. Both Overt Verbal and Manual versions of the Stroop yielded robust Stroop color-word interference as indexed by longer RT for incongruent than congruent color words. The Incongruent vs Congruent ERP difference wave presented two effects. A first effect was a medial dorsal negativity between 350–500 ms post-stimulus (peak at 410 ms). This effect had a significantly different scalp distribution in the Verbal and Manual Stroop versions, with an anterior–medial focus for overt or covert speech, and a broader medial–dorsal distribution for the manual task. Dipole source analysis suggested two independent generators in anterior cingulate cortex. Later on in time, a prolonged positivity developed between 500–800 ms post-stimulus over left superior temporo-parietal scalp. This effect was present for all the three response modalities. A possible interpretation of these results is that Stroop color-word interference first activates anterior cingulate cortex (350–500 ms post-stimulus), followed by activation of the left temporo-parietal cortex, possibly related to the need of additional processing of word meaning.

Introduction

The Stroop color interference task is among the most extensively studied paradigms in cognitive psychology. The classical behavioral effect consists of a lengthening in reaction time to color naming when the word meaning and the presentation do not match (i.e., they are “incongruent”) relative to when they correspond (i.e., they are “congruent”) [31]. At a psychological level, the Stroop effect has been best explained in terms of response competition. Longer reaction time and greater interference is present when the irrelevant attribute of the stimulus (the word meaning) is analyzed faster (i.e. is more automatic) than the relevant attribute (color), and the unwanted response is therefore available first (“race” model) [21].

Lesion correlation data in stroke patients [28], [35], and neuroimaging studies in healthy volunteers [6], [8], [11], [13], [20], [22], [33] have greatly improved our understanding of regions in the brain mediating the attentional demands involved in the Stroop color word interference. More specifically, the anterior cingulate cortex appears to be heavily involved, although other areas have been less consistently reported, such as inferior frontal cortex, parietal cortex, posterior cingulate and motor and premotor regions. The original Positron Emission Tomography (PET) findings generally supported the construct of an anterior attentional system involved in target detection and response selection, especially when confronted with conflicting stimulus or response biases [24], [25].

Considerable interest in the Stroop Task derives from its growing utility as a diagnostic and research tool to probe executive function in frontal lobe injury [28], [35] and psychiatric disease. In depression and schizophrenia, Stroop performance is generally impaired. PET abnormalities in the anterior cingulate have been described both at rest and during attentional performance [5], [9], [12], [14], [17], [18]. In depression, abnormal resting state hypometabolism and hypoperfusion have been reported in all portions of the anterior cingulate, including the ventral subgenual [12], [18], the rostral [17], and the dorsal [5], [18] partitions. The dorsal anterior cingulate abnormality in both depression and schizophrenia corresponds to the region implicated in selective attention by previous PET studies in normals [24], [25].

In spite of the well-replicated finding of anterior cingulate involvement in the Stroop color interference task delineated by cognitive neuroimaging studies using PET [6], [8], [11], [13], [22], [33] and recently functional Magnetic Resonance Imaging (fMRI) [20], several key aspects of the neuropsychology of the Stroop effect are still unknown. The first concerns the temporal course of the activations in the network of brain regions involved in the execution of the Stroop task. PET is limited by long integration time (40–120 s), and fMRI is limited by the long duration of the hemodynamic response (several seconds), although Event-Related fMRI appear as a promising technique to improve its temporal resolution [7], [27]. Event-Related Potentials (ERPs) possess exquisite temporal resolution (ms level), but only coarse spatial resolution — improved by the use of high-density electrode arrays [34]. ERPs have the potential to identify the timing, order of activation, and dynamic orchestration of brain regions during the unfolding of the Stroop task. Secondly, because of the limited temporal resolution, PET and conventional fMRI only allow block paradigm design. This has been the case for all the available neuroimaging studies of the Stroop effect, whereas most typically blocks of “incongruent” color words have been contrasted to blocks of “congruent” color words [6], [8], [11], [13], [20], [22], [33]. It can be argued that under these conditions the Stroop Task sums up to a rather different task. First, the elements of switching and unpredictability are absent. Second, the “incongruent” blocks appear intuitively to require a different attentional set, a higher level of sustained attention and arousal, and possibly a different processing strategy altogether (effortful rather than automatic). In addition, in “congruent” blocks the word color may become irrelevant to the task, since the more automatic processing of the word meaning alone is sufficient to correctly perform the task. In contrast to PET and MRI, ERPs capitalize on selective averaging of different stimulus types within the same experimental block (i.e., “congruent” words and “incongruent” words), allowing mixed-trials analysis of the Stroop task in its unadulterated and classical form. Finally, alternative response modalities during the Stroop task that are less problematic in terms of motion artifacts in fMRI, such as manual and covert vocal, have not been much explored before in any imaging modality [2], [15], [26].

Three published ERP studies have addressed the Stroop effect. The first used only two color words, employed manual responses, and reported exclusively early color selection effects [26]. The second is a recent study using a covert version of the original Stroop task [2]. The third used three color words and employed only manual responses [15]. All of these studies only employed several electrodes, in the midline only. The paucity of studies employing the Stroop paradigm may be partly explained by the concern with contamination of the cognitive effects with overt speech artifacts and other motion artifacts frequently observed during speech [30], [36].

The aim of the present study was to investigate the temporal course of known brain activations during the conventional, mixed-trial Stroop color-word paradigm using high-density (64 channel) ERP recordings. An additional goal of the study was to directly contrast within the same subjects and sessions the conventional version of the Stroop task (Overt Speech) with two alternative response modalities: a covert color naming condition, and a four-choice manual version of the task. This approach allowed us to examine the dependence of the Stroop effect brain activations on the particular response modality employed, as well as being of general interest for future studies of the Stroop effect using brain activity measures that are highly vulnerable to motion artifacts, such as ERPs and fMRI.

Section snippets

Subjects

Eight healthy volunteers (right-handed, age 27.6±6.8 years, three men, five women) participated in the study. Subjects had no history of current or past neurological or psychiatric illness, vision was normal or corrected-to-normal. Informed consent was obtained from all subjects according to the norms of the University of Texas Health Science Center Institutional Review Board.

Stimuli and task

Subjects were seated in a reclining chair facing a monitor placed at 40 cm from their eyes. They were presented with the

Task performance

In the mean Reaction Time analysis (Table 1), a very robust Stroop color-word interference effect was obtained both in the Vocal version of the task, F(1,7)=104.8, P<0.0001, mean effect size=85 ms; and in the manual version of the task (Manual4), F(1,7)=47.8, P<0.0001, mean effect size=110 ms. For both choice and simple RT versions of the task, vocal and manual RTs were not statistically different. Similarly, RTs to the congruent color-words (Manual4) were the same as RTs to the words in the

Discussion

In the present study, robust behavioral and electrophysiological effects of color-word interference were obtained across various versions (overt, covert, manual) of a mixed-trial Stroop task. Strong behavioral effects were obtained both in the standard vocal version of the task and in a manual choice-RT version of the task. Electrophysiological effects were obtained in all three Stroop conditions, including the covert vocal conditions for which no behavior was available. Across the three tasks,

Conclusion

The results reported here would not have been possible without the selective averaging capabilities provided by the ERP technique, allowing the extraction of the specific contribution of congruent and incongruent words within the same blocks of trials, and without the high temporal resolution provided by ERPs, indicating the exact temporal course and the highly dynamic nature of these effects.

This study clarifies the timing of activations during the Stroop color-word interference effect. An

Acknowledgements

The authors wish to thank Dr. Christie Thomas for helpful comments in earlier versions in of the manuscript.

References (37)

  • C.J Bench et al.

    The anatomy of melancholia — focal abnormalities of cerebral blood flow in major depression

    Psychological Medicine

    (1992)
  • M.A Burock et al.

    Randomized event-related experimental designs allow for extremely rapid presentation rates using functional MRI

    Neuroreport

    (1998)
  • C.S Carter et al.

    Anterior cingulate gyrus dysfunction and selective attention deficits in schizophrenia: a [15O]H2O PET study during single-trial Stroop task performance

    American Journal of Psychiatry

    (1997)
  • S.W Derbyshire et al.

    Pain and Stroop interference tasks activate separate processing modules in anterior cingulate cortex

    Experimental Brain Research

    (1998)
  • W.C Drevets et al.

    Subgenual prefrontal cortex abnormalities in mood disorders

    Nature

    (1997)
  • M.S George et al.

    Regional brain activity when selecting response despite interference: An H215O PET study of the Stroop and emotional Stroop

    Human Brain Mapping

    (1994)
  • M.S George et al.

    Blunted left cingulate activation in mood disorder subjects during a response interference task (the Stroop)

    Journal of Neuropsychiatry & Clinical Neurosciences

    (1997)
  • J Grapperon et al.

    The contribution of cognitive evoked potentials to knowledge of mechanisms on the Stroop test

    Neurophysiologie Clinique

    (1988)
  • Cited by (0)

    Presented as poster at the 1999 Meeting of the Cognitive Neuroscience Society. Supported by a NARSAD Young Investigator Award to ML

    1

    Current address: Rotman Center, Baycrest Center for Geriatric Care, Toronto, Canada.

    View full text