- Email: [email protected]
Event-Related Potentials with the Stroop colour-word task: Timing of semantic conflict
International Journal of Psychophysiology 72 (2009) 246–252
Contents lists available at ScienceDirect
International Journal of Psychophysiology j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / i j p s y c h o
Event-Related Potentials with the Stroop colour-word task: Timing of semantic conflict Montserrat Zurrón ⁎, María Pouso, Mónica Lindín, Santiago Galdo, Fernando Díaz Department of Clinical Psychology and Psychobiology, University of Santiago de Compostela, Galicia, Spain
a r t i c l e
i n f o
Article history: Received 31 July 2008 Received in revised form 19 December 2008 Accepted 7 January 2009 Available online 13 January 2009 Keywords: Event-Related Potentials Stroop task Temporal Principal Components Analysis Semantic conflict
a b s t r a c t Event-Related Potentials (ERPs) elicited by congruent and incongruent colour-word stimuli of a Stroop paradigm, in a task in which participants were required to judge the congruence/incongruence of the two dimensions of the stimuli, were recorded in order to study the timing of the semantic conflict. The reaction time to colour-word incongruent stimuli was significantly longer than the reaction time to congruent stimuli (the Stroop effect). A temporal Principal Components Analysis was applied to the data to identify the ERP components. Three positive components were identified in the 300–600 ms interval in response to the congruent and incongruent stimuli: First P3, P3b and PSW. The factor scores corresponding to the First P3 and P3b components were significantly smaller for the incongruent stimuli than for the congruent stimuli. No differences between stimuli were observed in the factor scores corresponding to the PSW or in the ERP latencies. We conclude that the temporal locus of the semantic conflict, which intervenes in generating the Stroop effect, may occur within the time interval in which the First P3 and P3b components are identified, i.e. at approximately 300–450 ms post-stimulus. We suggest that the semantic conflict delays the start of the response selection process, which explains the longer reaction time to incongruent stimuli. © 2009 Elsevier B.V. All rights reserved.
1. Introduction The Stroop interference effect is one of the most robust and bestknown interference effects in psychology. It was observed by J.R. Stroop in 1935 when participants were shown drawings of coloured squares and incongruent colour words (names of colours written in an incongruent colour with their meaning, e.g. the word ‘red’ written in blue). Stroop observed that the participants required more time to say the colour of the colour-incongruent words than the colour of the squares. Stroop (1935) attributed the longer response time to the interference provoked by the conflict between the meaning of the word and the colour in which it was written. Since these initial experiments were carried out, the task used to provoke the Stroop interference effect has been modified on numerous occasions. These modifications to the initial task designed by Stroop have made it possible to verify the robustness of the phenomenon and to determine the variables affecting its magnitude (see MacLeod 1991, for a review). In paradigms of the Stroop colour-word type, the participant usually carries out a verbal task (saying the colour of the stimulus), or a manual task to verify the reaction time (pressing a specific button according to the colour of the stimulus). Although different methods have been used to
⁎ Corresponding author. Department of Clinical Psychology and Psychobiology, University of Santiago de Compostela, Campus Sur s/n. 15782 Santiago de Compostela, Galicia, Spain. Tel.: +34 981 563100x13707; fax: +34 981 528071. E-mail address: [email protected] (M. Zurrón). 0167-8760/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.ijpsycho.2009.01.002
measure the Stroop effect, the general aim is to compare the time taken by the participant to carry out the task in response to incongruent stimuli (stimuli with two dimensions or characteristics that are contradictory, such as the word ‘red’ written in blue), with the time taken in response to congruent (e.g. the word ‘red’ written in red) and/or neutral stimuli (e.g.: the word ‘phase’ or the nonword ‘XXXX’ written in any colour, or patches of colour). The increase in response time to incongruent stimuli is considered to be due to the fact that processing the word, a relatively automatic process, creates conflict with the processing of the colour, which interferes in the processing of this dimension. Different hypotheses have been proposed to explain how word interference operates in relation to colour. MacLeod (1991) concluded that the “parallel distributed processing” (PDP) model, proposed by Cohen et al. (1990), best explains the 18 empirical characteristics of the Stroop Effect (see MacLeod, 1991 for a description of the 18 characteristics). According to this model, colour-word stimuli activate two processing pathways (simultaneously, but to different strengths), one for the word and another for the colour, and each pathway is formed by various processing units. Intersections may occur between the two processing pathways at the level of different units. The interference would be the result of conflict between units of different pathways at the locus of one or more intersections. The model allows that the effect of Stroop interference can be generated by different types of conflict. The PDP model therefore reconciles explanations of the Stroop effect that were considered antagonistic, such as the response conflict and the semantic conflict hypotheses.
M. Zurrón et al. / International Journal of Psychophysiology 72 (2009) 246–252
The response conflict hypothesis considers that the Stroop interference is produced by the conflict between two response plans, one for the colour and another for the word (Dyer, 1973a; Morton, 1969; Morton and Chambers, 1973; Posner and Snyder, 1975), whereas the semantic conflict hypothesis considers that the interference is produced by the conflict between two semantic codes, one generated for the word and another for the colour (Luo, 1999; Seymur, 1977). De Houwer (2003) and Van Veen and Carter (2005) have investigated the hypothesis that the Stroop effect may be caused by the two types of conflict. Three types of stimuli were presented in both studies: incongruent stimuli that provoked semantic and response conflicts, incongruent stimuli that provoked semantic conflict but no response conflict, and congruent stimuli. It was found that the reaction time to the first type of stimulus was longer than the reaction time to the second, and both were longer than the reaction time to the congruent stimuli. These findings support the hypothesis that both types of conflict intervene in generating the Stroop effect. Van Veen and Carter (2005) found, by use of functional magnetic resonance imaging (fMRI), that the cortical networks activated by each type of conflict are not overlaid. The Stroop test has been successfully adapted to record EventRelated Potentials (ERPs). The interest in complementing reaction time measurements with ERP recordings, in studying the Stroop effect, lies in the fact that ERP recording enables precise analysis of the timing of the cognitive processes that the subject uses in processing information and therefore the conflict. Several studies have shown that the amplitude of the ERPs elicited by congruent and incongruent colour-word stimuli differs in the range of latency between 300–450 ms, in Stroop tasks in which the participants were required to respond to the colour of the stimuli (Eppinger et al., 2007; Hanslmayr et al., 2008; Houston et al., 2004; Ilan and Polich, 1999; Kray et al., 2005; Liotti et al., 2000; MarkelaLerenc et al., 2004; McNeely et al., 2003; Potter et al., 2002; Qiu et al., 2006; Rebai et al., 1997; West, 2003, 2004; West and Alain, 1999, 2000a, 2000b; West et al., 2005) and in tasks in which participants were required to decide on the congruence/incongruence of the two dimensions of colour-word stimuli (Mager et al. 2007). Some authors have identified the P300 component for this range of latencies, and have shown that the amplitude of this component elicited by incongruent stimuli is smaller than the amplitude elicited by congruent stimuli (Houston et al., 2004; Ilan and Polich, 1999; Potter et al., 2002). In other studies (Eppinger et al., 2007; Hanslmayr et al., 2008; Kray et al., 2005; McNeely et al., 2003; Qiu et al., 2006; Rebai et al., 1997; West, 2003, 2004; West and Alain, 1999, 2000a, 2000b) a negative-going wave denominated N400 or N450 or Ni (negativity for incompatible trials) has been identified, with less positive voltage values (as negative voltages are rarely reached) for incongruent stimuli than for congruent stimuli. Yet other studies have identified both components (Mager et al. 2007; West et al., 2005), and finally, Liotti et al. (2000) and Markela-Lerenc et al. (2004) opted to quantify the amplitude by the voltage values obtained within different time windows, without specifying components, and obtained the same results. In addition to the disparity as regards the ERP components identified in this range of latencies, no consensus has been reached as regards the topographic distribution of the differences in amplitude between the ERPs elicited by the two types of stimuli. On the one hand Mager et al. (2007), Ilan and Polich (1999) and Potter et al. (2002) observed the greatest differences in amplitude at central–parietal electrodes, for a manual congruence/incongruence judgment task, a manual response to colour, and a verbal response to colour, respectively, whereas Hanslmayr et al. (2008), Houston et al. (2004),Markela-Lerenc et al. (2004) and West and Alain (1999, 2000a,b), observed the greatest differences in amplitude at centrofrontal locations for manual responses to colour. Finally, Liotti et al.
247
(2000),Eppinger et al. (2007) and Kray et al. (2005) observed that the differences in amplitude were widely distributed across the scalp for a manual response to colour. As regards the meaning of such differences in the ERP amplitudes in response to the two types of stimulus, the N400 (N450 or Ni) wave has been linked to the stimulus-related conflict (Eppinger et al., 2007; Kray et al., 2005; McNeely et al., 2003; West, 2003, 2004). This interpretation is very general and does not specify the stage of processing of the word-colour stimulus at which the conflict is produced, or in other words, what type of conflict in relation to the stimulus is involved in the Stroop effect. In the range of latencies between 300 and 450 ms it is therefore necessary to determine which of the ERP component(s) display differences in amplitude in response to the two types of stimuli, to clarify their topographical distribution, and to specify the underlying conflict. Tasks in which congruence/incongruence must be judged (Dyer, 1973b; Luo, 1999; Mager et al., 2007; Treisman and Fearnley, 1969; Zysset et al., 2001) have the advantage (relative to the colour-word response tasks) of guaranteeing that the intersection between the two processing pathways (colour and word), and therefore the conflict between them for incongruent stimuli, will occur before the motor response is produced, because both responses (whether congruent or not) require that the intersection has already been produced. In these tasks the semantic conflict associated with generation of the Stroop effect is therefore favoured with respect to the response conflict, better enabling analysis of the effect of the semantic conflict on the ERPs. The aims of the present study were therefore to determine: 1) the temporal locus of the colour-word semantic conflict involved in the Stroop effect; 2) the ERP component(s) associated with or modulated by the semantic conflict, and 3) the scalp distribution of such components. For this, we recorded ERPs elicited by colour-word stimuli in a congruence/incongruence judgment task, and applied temporal Principal Components Analysis to the ERP data. 2. Materials and methods 2.1. Participants Twenty one university students (7 female) aged between 19 and 24 (Mean: 20.95; SD = 1.75), with a medium–high socioeconomic status, participated as volunteers in the study. All had normal or corrected to normal vision (but no contact lenses). All were healthy with no history of neurological or psychiatric disorders or drug abuse, and were not receiving medication at the time of participating in the study (nor had received any in the preceding weeks). The participants were also required to abstain from drugs/alcohol/caffeine and nicotine prior to testing, and none reported fatigue caused by lack of sleep. None of the participants were familiar with the protocols used in the study. All participants were self-reported right-handed. 2.2. Task and stimuli The participants were presented with 104 stimuli separated into two blocks (i.e. each comprised 52 stimuli), with a rest period of 90 s between them. The stimuli were the Spanish words “azul” (blue), “verde” (green), “rojo” (red) and “gris” (grey) (each word was displayed 26 times), and were written in one of these four colours on a black background, so that the stimulus was either congruent with the colour and meaning of the word (the colour and meaning coincided), or incongruent (the word and colour did not coincide). Half of the stimuli were congruent and half incongruent. The stimuli were presented in a random order, and each word was displayed for 200 ms with an SOA of 2000 ms. The screen was black during the 1800 ms between the end of one stimulus and the onset of the next.
248
M. Zurrón et al. / International Journal of Psychophysiology 72 (2009) 246–252
The images were displayed on a flat 19-inch monitor placed 1 m from the subject's face. The words were presented in the centre of the screen in upper case Helvetica 2 typeface; each letter was 2 cm wide and 3.5 cm high, and subtended at an angle of 1.15 × 2.01° of arc. The participants were asked to press a button with their right thumb if both dimensions of the stimulus were incongruent, and to press another button with their left thumb if both dimensions of the stimulus were congruent. Prior to recording, a trial run was carried out with six stimuli to establish whether the subject understood the instructions and responded correctly. 2.3. EEG recording All of the participants were seated in a comfortable chair in an electrically-shielded laboratory with attenuated sound and lighting levels, and were instructed not to move during the recording session and to look at the centre of the screen. Each participant was fitted with a cap with a nose reference and a frontopolar ground, for recoding the electroencephalographic (EEG) activity at 30 electrode sites corresponding to the 10–20 system. Eye movements were recorded simultaneously with additional electrodes placed above and below the left eye (VEOG) and at the outer canthus of each eye (HEOG). All impedances were reduced to 5 KΩ or less in order to obtain good quality recordings, and care was taken to avoid damaging the subject's scalp. The EEG signals were passed through a 0.1–30 Hz band-pass filter and amplified at 22.5 K before being sampled at 500 Hz; the digitalized signal was then stored. Ocular artefacts were corrected off-line with the algorithm of Semlitsch et al. (1986). Pre-stimulus (200 ms) and post-stimulus (1400 ms) segments were extracted from the EEG, and the pre-stimulus baseline was corrected. Signals exceeding ±100 μV were automatically excluded from the averages. For the averages only the epochs corresponding to the stimuli with correct responses were taken into account. The number of epochs included in the ERP averages for each type of stimulus (congruent and incongruent) was between 45 and 50, and there were no differences between the number of epochs for either type of stimulus.
The tPCA provides a statistical decomposition of the brain electrical patterns superposed on the scalp-recorded data (Dien andFrishkoff, 2004). A covariance-matrix-based tPCA was applied for both conditions (congruent and incongruent stimuli). The decision regarding selection of the number of components (or factors) was based on the scree test (Cattell, 1966). Extracted factors were then subjected to Promax rotation. Dien (1998) showed that the use of Promax rotation improves the accuracy of the results and reduces problems such as misallocation of variance (see also Dien andFrishkoff, 2004). The tPCA provides two matrices, one for factor loadings and another for factor scores. The first shows the load of each factor through time, and the second provides information about the extent to which each factor is present in the averaged ERPs (i.e. at each electrode site). The factor scores are transformed values of the original voltages and thus can be used as a measure of the amplitude. The scores of each Temporal Factor were examined by repeated measures analysis of variance (ANOVAs), with two within-subject factors: Congruence (with 2 levels: congruent and incongruent) and Electrode (with 3 levels: Fz, Cz and Pz). As regards the latencies of the ERP components, the peak latencies were analyzed by repeated measures ANOVAs, with two withinsubject factors: Congruence (with 2 levels: congruent and incongruent) and Electrode (with 2 or 3 levels), and by paired sample ttests, depending on the component. When the ANOVA for the latencies and for the factor scores revealed significant results, pairwise comparison of means (with Bonferroni corrections) was carried out to identify the source of the differences. Greenhouse–Geisser corrections to the degrees of freedom were applied in all cases in which the condition of sphericity was not met. In these cases, the original degrees of freedom are presented together with the corrected p and ε values. Differences were considered significant at p ≤ 0.05. All statistical analyses were carried out with SPSS (version 15.0). 3. Results 3.1. Behavioural results
2.4. Statistical analysis Paired sample t-tests were used to compare the reaction times (RTs) and the number of errors in response to congruent and incongruent stimuli. A temporal Principal Components Analysis (tPCA) was applied to the ERP data to ensure correct identification of the ERP components.
The RT was significantly longer in response to the incongruent stimuli (Mean = 623 ms; SD = 84) than in response to the congruent stimuli (Mean = 563 ms; SD = 92), t(13) = −7.2, p b 0.0001. There was no significant difference in the number of errors between the two types of stimuli (Congruent: Mean = 2.7; SD = 2.3; Incongruent: Mean = 2.3; SD = 2.1).
Fig. 1. Temporal Factor (TF) loadings (Y-axis) in the Principal Components Analysis for ERP data. X-axis: Time (ms)
M. Zurrón et al. / International Journal of Psychophysiology 72 (2009) 246–252
3.2. Temporal Principal Components Analysis (tPCA) Fig. 1 shows the Temporal Factor (TF) loadings, and Fig. 2 illustrates the First P3, P3b, and PSW components in the grand-averaged EventRelated Potential waveforms, at Fz, Cz and Pz electrode sites, and the
249
voltage maps showing the topographical distribution of the amplitude (μV) on the peak latencies of the three components. The mean factor scores of TF2, TF7 and TF4, at the Fz, Cz, and Pz electrode sites and the mean latencies (ms) of First P3, P3b, and PSW, elicited by the congruent and incongruent stimuli are shown in
Fig. 2. Grand average ERP waveforms at Fz, Cz and Pz for congruent (red line) and incongruent (blue line) stimuli, and the voltage maps showing the topographical distribution of the amplitude (μV) on the peak latencies of the First P3, P3b, and PSW components.
250
M. Zurrón et al. / International Journal of Psychophysiology 72 (2009) 246–252
Table 1 Mean factor scores for TF2 (First P3), TF7 (P3b), and TF4 (PSW), at the Fz, Cz, and Pz electrode sites, and mean latencies (ms) of the First P3, P3b, and PSW components elicited by the congruent and incongruent stimuli. Standard deviations are shown in brackets. TF: Temporal Factor, PSW: Positive Slow Wave. Factor scores
Congruent
Incongruent
TF2 (First P3)
TF7 (P3b)
TF4 (PSW)
TF2 (First P3)
TF7 (P3b)
TF4 (PSW)
Fz Cz Pz
− 0.04 (0.97) − 0.08 (1.15) 0.74 (1.08)
0.09 (0.85) 0.68 (0.91) 0.9 (1.07)
− 0.40 (0.87) 0.47 (0.72) 0.98 (0.91)
− 0.24 (0.75) − 0.35 (1.08) 0.42 (1.01)
− 0.11 (0.88) 0.42 (0.96) 0.58 (1.17)
− 0.33 (0.86) 0.37 (0.53) 0.84 (0.76)
Latencies
First P3
P3b
PSW
First P3
P3b
PSW
308 (18.1)
380 (23) 379 (27.5) 384 (33.3)
311 (19.6)
397 (56.5) 395 (53.3) 396 (57.6)
578 (62.8) 578 (59.3)
Fz Cz Pz
565 (62.8) 559 (63.5)
Table 1; the F values in repeated-measures ANOVAs for the factor scores of the TF2, TF7, and TF4, and for the latencies of P3b and PSW are shown in Table 2. The tPCA identified eight TF, which explained 91.7% of total variance. Taking into account the temporal range of the factor loadings and that the scores for the factors were positive and maximal at parietal electrodes, three of these TF appeared to correspond to the ERP components observed in the 300–600 ms interval. The TF2 corresponded to a posterior positive component, which was denominated First P3. The peak latency of the First P3 component was measured at Pz in the 290–330 ms interval. The TF7 corresponded to the positive P3b central-parietal component described by Sutton et al. (1965). The peak latency of P3b was measured at Fz, Cz and Pz in the 360–430 ms interval. The TF4 corresponded to a parietal–central Positive Slow Wave, which was denominated PSW. The peak latency of PSW was measured at Cz and Pz in the 540–600 ms the interval. Temporal Factor 2 (corresponding to First P3 component) showed significantly higher scores in response to congruent stimuli than incongruent stimuli, and significantly higher at the Pz than at the Cz and Fz electrode sites. Temporal Factor 7 (corresponding to P3b component) presented significantly higher scores in response to congruent stimuli than incongruent stimuli, and significantly higher at the Pz and Cz than at the Fz electrode site. Temporal Factor 4 (corresponding to the PSW component) showed significantly higher scores at Pz than Cz and Fz, and significantly higher at the Cz than at the Fz electrode site. No differences were found for the factor scores in response to the congruent and incongruent stimuli, at the electrodes evaluated.
Table 2 F values in repeated-measures ANOVA for the factor scores for TF2 (First P3), TF7 (P3b), and TF4 (PSW), and for the latencies of P3b and PSW (ε = epsilon value, C = Congruence factor, E = Electrode factor). Factor scores
TF2 (First P3)
TF7 (P3b)
TF4 (PSW)
ANOVA (C × E) C E C×E
5.43⁎ 12.18⁎⁎⁎ (ɛ: 0.77) 1.21 (ɛ: 0.77)
5.85⁎ 10.27⁎⁎ (ɛ: 0.68) 0.75 (ɛ: 0.57)
0.34 27.33⁎⁎⁎ (ɛ: 0.66) 5.71⁎⁎
P3b
PSW
2.33 0.31 0.35 (ɛ: 0.73)
1.82 0.57 1.44
The latency of First P3 component did not show any differences in response to congruent and incongruent stimuli (t(13) = −0.45, p b 0.66). The latencies of the P3b and PSW components were not different, either in response to congruent or incongruent stimuli, or among electrode sites. 4. Discussion 4.1. Task performance A clear Stroop effect was obtained, as the RT to the incongruent stimuli was significantly longer than the RT to the congruent stimuli. The stimulation paradigm used, adapted to the recording to ERP and with congruence/incongruence judgment response, was therefore suitable for studying the effect of Stroop colour-word interference. This also indicates that with congruence/incongruence judgment tasks, the conflict between the semantic code of the word that is read and the semantic code activated by the colour perceived plays an essential role in generating the Stroop effect, which supports the importance of this conflict in colour-response Stroop tasks. The task for judging the congruence/incongruence of both dimensions (chromatic and linguistic) of the stimulus requires the controlled processing of both dimensions (see criteria for automaticity/control in: Kahneman and Treisman, 1984; Neumann, 1984; Shiffrin, 1988). However, in the response to the colour of the stimuli, it is assumed that processing of the linguistic dimension is relatively automatic. Therefore, the Stroop effect is obtained both when the processing of the linguistic dimension is relatively automatic and when the processing is controlled. Therefore, although the degree of automation of a task affects the interference, so that the most automatic task interferes in the least automatic task (MacLeod, 1991; MacLeod and Dunbar, 1988), it is not necessary for one of the two dimensions of the stimulus to be processed automatically (or in a relatively automatic way) for the Stroop effect to be generated. 4.2. ERPs
Latency ANOVA (C × E) C E C×E
3.3 Latency of ERP components
–
⁎⁎⁎p b 0.0001, ⁎⁎p b 0.01, ⁎p b 0.05. For the factor scores, the degrees of freedom for the Congruence factor were 1/20, and for the Electrode factor and the Congruence × Electrode interaction were 2/40. For the latencies, the degrees of freedom for the Congruence factor was 1/15 in P3b and 1/16 in PSW, and for the Electrode factor and the Congruence × Electrode interaction were 2/30 in P3b and 1/16 in PSW.
4.2.1. First P3 and P3b The scores for the TF2 and TF7 were significantly smaller for incongruent stimuli than for congruent stimuli, and therefore the amplitudes of the ERP components –First P3 and P3b– were smaller for incongruent than for congruent stimuli. Given the latency, polarity and topography of the First P3, we consider that it belongs to the P300 “family”, and we will therefore interpret it as such. Taking into account that the P300 amplitude is negatively correlated with the difficulty of the task (Donchin, 1981; Donchin and Coles, 1988; Johnson, 1986; Verleger, 1988), the smaller First P3 and P3b amplitudes may reflect greater difficulty in processing the incongruent stimuli than the congruent stimuli. This greater difficulty may be the result of the
M. Zurrón et al. / International Journal of Psychophysiology 72 (2009) 246–252
semantic conflict generated by incongruent stimuli. The smaller amplitude of First P3 and P3b in response to incongruent stimuli may be considered as an index of the differential semantic processing of these stimuli relative to the congruent stimuli, and therefore the smaller amplitude in the interval between 300 and 450 ms in response to incongruent stimuli may indicate the temporal locus of the semantic conflict, i.e. the intersection between the unit that processes the meaning of the word and the unit that processes the meaning of the colour might occur around 300 and 450 ms post-stimulus. The differences in amplitude between the ERP elicited by both stimuli are found at the three midline electrodes, in accordance with the results of Eppinger et al. (2007), Kray et al. (2005) and Liotti et al. (2000), and in contrast with the results of other studies in which greater differences in amplitude were observed at centro-parietal (Ilan and Polich, 1999; Potter et al., 2002) and centro-frontal electrodes (Hanslmayr et al., 2008; Houston et al., 2004; MarkelaLerenc et al., 2004 West and Alain, 1999, 2000a,b). Such discrepancies may be due to the combination of two aspects in which the studies differ: the different measures of amplitude considered (amplitude of the P300 component, or of the N400 wave or of the mean voltage in a time window); and different types of response to the word-colour stimuli (manual compared with verbal response, response to colour of the stimulus compared with congruence judgment). In the present congruence/incongruence judgment task, we did not observe any independent TF that corresponds to a N400 (N450 or Ni) ERP component. This, in conjunction with the significantly different TF2 and TF7 scores for congruent and incongruent stimuli, allow us to conclude that, in this type of task, the “N400” wave is not a component of the ERPs. The results therefore indicate that, at least in congruence judgment tasks, the differences in amplitude between the two types of stimuli in the range 300–450 ms may be related to the difference in amplitude of the P300 complex. One may wonder if the N400 (N450 or Ni) ERP wave, described in previous studies involving colour response tasks (Eppinger et al., 2007; Hanslmayr et al., 2008; Kray et al., 2005; McNeely et al., 2003; Qiu et al., 2006; Rebai et al., 1997; West, 2003, 2004; West and Alain, 1999, 2000a, 2000b) corresponds to a real component or also reflects differences in the P300 complex amplitude. In light of the above, in future studies in which ERPs are evaluated in Stroop tasks, with different types of response to word-colour stimuli, it would be advisable to use tPCA for correct identification of the ERP components. There were no differences in latencies of First P3 and P3b between the two types of stimuli. The latency of P300 has received particular interest in studies of ERPs elicited by Stroop tasks. In this respect, the results of the present study are consistent with those of previous studies in which Stroop colour response tasks for incongruent and congruent (or neutral) stimuli were used (Atkinson et al., 2003; Duncan-Johnson and Kopell, 1980, 1981; Grapperon et al., 1998; Ilan and Polich, 1999; Lavoie, 1999; Rosenfeld and Skogsberg, 2006; West and Alain, 2000a). It has been suggested (Atkinson et al., 2003; Duncan-Johnson and Kopell, 1981; Ilan and Polich, 1999; Lavoie, 1999; Rosenfeld and Skogsberg, 2006) that the lack of any differences in the latencies of this component rules out any possible influence of stages prior to the elaboration of responses in the generation of the Stroop effect. This argument is based on the fact that the latency of P3b was initially considered as an index of the stimulus evaluation time that is independent of the response processes (Donchin, 1981; Kutas et al., 1977). It was reasoned that if conflict is produced between two motor programmes, the latency of this component should not differ in response to incongruent and congruent (or neutral) stimuli. However, in light of the results of the present study, we consider that the absence of differences in the latencies of the ERPs elicited by congruent and incongruent stimuli cannot be used as data to support the response conflict rather than the semantic conflict, as in the present study a congruence/incongruence judgment task was used,
251
thereby minimizing the effect that the conflict between the responses may have on Stroop interference. Consequently, the absence of differences in latencies of the components of the ERPs studied does not rule out the relevance of the semantic intersection stage in the genesis of the Stroop effect. Pfefferbaum et al. (1986) and Verleger (1997) proposed that in complex tasks, evaluation of the stimulus and selection of the response should be processed serially, so that selection of the response begins after P300; in contrast, in simple tasks, evaluation of the stimulus and selection of the response may be processed in parallel, and therefore in the temporal interval in which P300 appears, the subject will also be selecting the response. If we consider that in a congruence/incongruence task the response to congruent stimuli is simple, because of the lack of conflict, the process of selecting the response may begin during the process of evaluating the stimulus, i.e. in the temporal interval of the P300 complex. However, the semantic conflict of the incongruent stimulus should be resolved before the response is selected. In other words, the incongruent stimulus requires that the processes of evaluating the stimulus and selection of the response occur serially, so that selection of the response will begin after evaluation of the stimulus (after the P300 complex). As a result, a delay in the start of the selection of the response to the incongruent stimulus, with respect to the congruent stimulus, will explain the delay in the reaction time to the former, with respect to the latter. Furthermore, this is consistent with a recent interpretation of P300, in which this component reflects some process that mediates between the perceptual analysis of the stimuli and initiation of the response processes (Verleger et al., 2005). 4.2.2. The Positive Slow Wave (PSW) There were no differences in either the latency or amplitude of the parieto-central PSW component between conditions. Mager et al. (2007) used a Stroop congruence–incongruence judgment task, and also observed a positive component after P300 with central–parietal distribution and a latency peak at 550 ms. This late positive component has also been identified in colour response tasks (West and Alain, 1999, 2000a,b), and was denominated Sustained Potential (SP) by West and colleagues (McNeely et al., 2003; West, 2003, 2004; West et al., 2005). A similar positive component was identified by Squires et al. (1975), in oddball tasks in which the subject was required to count the target stimuli; this component appeared after P300 (P3b), with a maximum amplitude at Pz, and did not reach positive values at Fz. Although it is not clear which cognitive process is related to the PSW component, in order for this to be elicited, it appears necessary to use tasks in which the subject has to do more than detect a given stimulus, and for example, has to select one from amongst a number of possible answers (Perchet and García-Larrea, 2000). García-Larrea and Cézanne-Bert (1998) considered that in general terms, the PSW component is a non-specific activity that indicates completion of a synchronized operation immediately after detection of the target. The data obtained in the present study also appear to support the relationship between the PSW component obtained from the Stroop task with processes that are subsequent to categorization of the stimuli as congruent or incongruent. The latency of the PSW elicited by congruent stimuli is coincident with the RT, whereas for incongruent stimuli the latency was shorter than the RT and therefore the present results do not support the interpretation of West et al. (2005), who related the SP component to the process of selecting the response. We suggest that the PSW may be associated with review of the stimulus categorization, a process that may occur in parallel with the readiness and/or execution of the motor response. In summary, without ruling out the idea that a conflict between two response plans may intervene in the generation of the Stroop effect, in a congruence/incongruence judgment task, the incongruent
252
M. Zurrón et al. / International Journal of Psychophysiology 72 (2009) 246–252
colour-word stimuli provoked a conflict between two semantic codes (colour and word) which lengthened the reaction time to these stimuli (Stroop effect). This semantic conflict was manifested as a smaller amplitude of First P3 and P3b elicited by incongruent stimuli relative to those elicited by congruent stimuli, and therefore the temporal locus of the semantic conflict may occur at between 300 and 450 ms post-stimulus. It is suggested that the semantic conflict delays the start of the selection of the response, which may explain the longer reaction time to incongruent stimuli. References Atkinson, C.M., Drysdale, K.A., Fulham, W.R., 2003. Event-related potential to Stroop and reverse Stroop stimuli. Int. J. Psychophysiol. 47, 1–21. Cattell, R.B., 1966. The scree test for the number of factors. Behav. Res. 1, 245–276. Cohen, J.D., Dunbar, K., McClelland, J.L., 1990. On the control of automatic processes: a parallel distributed processing account of the Stroop effect. Psychol. Rev. 97, 332–361. De Houwer, J., 2003. On the role of stimulus–stimulus and stimulus–response compatibility in the Stroop effect. Mem. Cognit. 31, 353–359. Dien, J., 1998. Addressing misallocation of variance in principal component analysis of event-related potentials. Brain Topogr. 11, 43–55. Dien, J., Frishkoff, G.A., 2004. Principal components analysis of ERPs data. In: Handy, T.C. (Ed.), Event-Related Potentials. A Methods Handbook. MIT Press, Cambridge, pp. 189–207. Donchin, E., 1981. Surprise!...Surprise? Psychopysiology 18, 493–513. Donchin, E., Coles, M.G.H., 1988. Is the P300 component a manifestation of context updating? Behav. Brain Sci. 11, 355–372. Duncan-Jonhson, C.C., Kopell, B.S., 1980. The locus of interference in the Stroop task: when you read “blue”, do you see “red”? Psychophysiology 17, 308–309. Duncan-Jonhson, C.C., Kopell, B.S., 1981. The Stroop effect: brain potentials localize the source of interference. Science 214, 938–940. Dyer, F.N., 1973a. The Stroop phenomenon and its use in the study of perceptual, cognitive and response processes. Mem. Cognit. 1, 106–120. Dyer, F.N., 1973b. Same and different judgments for word-color pairs with “irrelevant” words or colors: evidence for word-code comparisons. J. Exp. Psychol. 98, 102–108. Eppinger, B., Kray, J., Mecklinger, A., John, O., 2007. Age differences in task switching and response monitoring: evidence from ERPs. Biol. Psychol. 75, 52–67. García-Larrea, L., Cézanne-Bert, G., 1998. P3, positive slow wave and working memory load: a study on the functional correlates of slow wave activity. EEG & Clin. Neurophysiol. 108, 260–273. Grapperon, J., Vidal, F., Leni, P., 1998. The contribution of cognitive evoked potentials to knowledge of mechanism on the Stroop test. Neurophysiol. Clin. 28, 207–220. Hanslmayr, S., Pastötter, B., Bäuml, K., Gruber, S., Winber, M., Klimesch, W., 2008. The electrophysiological dynamics of interference during the Stroop task. J. Cogn. Neurosci. 20, 215–225. Houston, R.B., Bauer, L.O., Hesselbrock, V.M., 2004. Effects of borderline personality disorder features and a family history of alcohol or drug dependence on P300 in adolescents. Int. J. Psychophysiol. 53, 57–70. Ilan, A.B., Polich, J., 1999. P300 and response time from a manual Stroop task. Clin. Neurophysiol. 110, 367–373. Jonhson Jr., R., 1986. A triarchic model of P300 amplitude. Psychopgysiology 23, 367–384. Kahneman, D., Treisman, A., 1984. Changing views of attention and automaticity. In: Parasuraman, R., Davis, D.R. (Eds.), Varieties of Attention. Academic Press, New York, pp. 29–61. Kray, J., Eppinger, B., Mecklinger, A., 2005. Age differences in attentional control: an event-related potential approach. Psychophysiology 42, 407–416. Kutas, M., McCarthy, G., Donchin, E., 1977. Augmenting mental chronometry: the P300 as a measure of stimulus evaluation time. Science 197, 792–795. Lavoie, M.E., 1999. Toward a functional explanation of the locus of the Stroop interference: a psychophysiological study. Brain Cogn. 40, 167–170. Liotti, M., Woldorff, M.G., Perez, R., Mayberg, H.S., 2000. An ERP study of the temporal course of the Stroop color-word interference effect. Neuropsychologia 38, 701–711. Luo, C.R., 1999. Semantic competition as the basis of Stroop interference: evidence from color-word matching task. Psychol. Sci. 10, 35–40. MacLeod, C.M., 1991. Half a century of research on the Stroop effect: an integrative review. Psychol. Bull. 109, 163–203. MacLeod, C.M., Dunbar, K., 1988. Training and Stroop-like interference: evidence for a continuum automaticity. J. Exp. Psychol. Learn. Mem. Cognit. 10, 304–315. Mager, R., Bullinger, A.H., Brand, S., Schmidlin, M., Schärli, H., Müller-Spahn, F., Störmer, R., Falkenstein, M., 2007. Age-related changes in cognitive conflict processing: an event-related potential study. Neurobiol. Aging 28, 1925–1935.
Markela-Lerenc, J., Ille, N., Kaiser, S., Fiedler, P., Mundt, C., Weisbrod, M., 2004. Prefrontal-cingulate activation during executive control: which comes first? Cogn. Brain Res. 18, 278–287. McNeely, H., West, R., Christensen, B.K., Alain, C., 2003. Neurophysiological evidence for disturbances of conflict processing in patients with schizophrenia. J. Abnorm. Psychol. 112, 1–9. Morton, J., 1969. Categories of interference: verbal mediation and conflict in card sorting. Br. J. Psychol. 60, 329–346. Morton, J., Chambers, S.M., 1973. Selective attention to words and colours. Q. J. Exp. Psychol. 25, 387–397. Neumann, O., 1984. Automatic processing: a review of recent findings and a plea for an old theory. In: Prinz, W., Sanders, A.F. (Eds.), Cognition and Motor Processes. Springer Verlag, Berlin, pp. 255–293. Perchet, C., García-Larrea, L., 2000. Visuospatial attention and motor reaction in children: an electrophysiological study of the “Posner” paradigm. Psychophysiology 37, 231–241. Pfefferbaum, A., Christensen, C., Ford, J.M., Kopell, B.S., 1986. Apparent response incompatibility effects on P3 latency depend on the task. EEG & Clin. Neurophysiol. 64, 424–437. Posner, M.I., Snyder, C.R.R., 1975. Attention and cognitive control. In: Solso, R.L. (Ed.), Information Processing and Cognition. The Loyola Symposium, Erlbaum, Hillsdale NJ, pp. 55–85. Potter, D.D., Jory, S.H., Bassett, M.R.A., Barrett, K., Mychalkiw, W., 2002. Effect of mild head injury on event-related potential correlates of Stroop task performance. J. Int. Neuropsychol. Soc. 8, 828–837. Qiu, J., Luo, Y., Wang, Q., Zhang, F., Zhang, Q., 2006. Brain mechanism of Stroop interference effect in Chinese characters. Brain Res. 1072, 186–193. Rebai, M., Bernard, C., Lannou, J., 1997. The Stroop test evokes a negative brain potential, the N400. Int. J. Neurosci. 91, 85–94. Rosenfeld, J.P., Skogsberg, K.R., 2006. P300-based Stroop study with low probability and target Stroop oddballs: the evidence still favours the response selection hypothesis. Int. J. Psychophysiol. 60, 240–250. Semlitsch, H.V., Anderer, P., Schuster, P., Presslich, O., 1986. A solution for reliable and valid reduction of ocular artefacts applied to the P300 ERP. Psychophysiology 23, 695–703. Seymur, P.H., 1977. Conceptual encoding and locus of the Stroop effect. Q. J. Exp. Psychol. 29, 245–265. Shiffrin, R.M., 1988. Attention. In: Atkinson, R.C., Hernstein, R.J., Lindzey, G., Luce, R.D. (Eds.), Steven's Handbook of Experimental Psychology. Wiley, New York, pp. 739–911. Squires, N.K., Squires, K., Hillyard, S.A., 1975. Two varieties of long-latency positive waves evoked by unpredictable auditory stimuli in man. EEG & Clin. Neurophysiol. 38, 387–401. Stroop, J.R., 1935. Studies of interference in serial verbal reactions. J. Exp. Psychol. 18, 643–662. Sutton, S., Braren, M., Zubin, J., John, E.R., 1965. Information delivery and the sensory evoked potential. Science 150, 1187–1188. Treisman, A., Fearnley, S., 1969. The Stroop test: selective attention to colours and words. Nature 222, 437–439. Van Veen, V., Carter, C.S., 2005. Separating semantic conflict and response conflict in the Stroop task: a functional MRI study. NeuroImage 27, 497–504. Verleger, R., 1988. Event-related potentials and memory: a critique of the context updating hypothesis and an alternative interpretation of P3. Behav. Brain Sci. 11, 341–354. Verleger, R., 1997. On the utility of P3 latency as an index of mental chronometry. Psychophysiology 34, 131–156. Verleger, R., Jaskowski, P., Wascher, E., 2005. Evidence for an integrative role of P3b in linking reaction to perception. J. Psychophysiol. 19, 165–181. West, R., 2003. Neural correlates of cognitive control and conflict detection in the Stroop and digit-location tasks. Neuropsychologia 41, 1122–1135. West, R., 2004. The effects of aging on controlled attention and conflict processing in the Stroop task. J. Cogn. Neurosci. 16, 103–113. West, R., Alain, C., 1999. Event-related neural activity associated with the Stroop task. Cogn. Brain Res. 8, 157–164. West, R., Alain, C., 2000a. Age-related decline in inhibitory control contributes to the increased Stroop effect observed in older adults. Psychophysiology 37, 179–189. West, R., Alain, C., 2000b. Effects of task context and fluctuations of attention on neural activity supporting performance of Stroop task. Brain Res. 873, 102–111. West, R., Jakubek, K., Wymbs, N., Perry, M., Moore, K., 2005. Neural correlates of conflict processing. Exp. Brain Res. 167, 38–48. Zysset, S., Müller, K., Lohmann, G., von Cramon, D.Y., 2001. Color-word matching Stroop task: separating interference and response conflict. NeuroImage 13, 29–36.
Contents lists available at ScienceDirect
International Journal of Psychophysiology j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / i j p s y c h o
Event-Related Potentials with the Stroop colour-word task: Timing of semantic conflict Montserrat Zurrón ⁎, María Pouso, Mónica Lindín, Santiago Galdo, Fernando Díaz Department of Clinical Psychology and Psychobiology, University of Santiago de Compostela, Galicia, Spain
a r t i c l e
i n f o
Article history: Received 31 July 2008 Received in revised form 19 December 2008 Accepted 7 January 2009 Available online 13 January 2009 Keywords: Event-Related Potentials Stroop task Temporal Principal Components Analysis Semantic conflict
a b s t r a c t Event-Related Potentials (ERPs) elicited by congruent and incongruent colour-word stimuli of a Stroop paradigm, in a task in which participants were required to judge the congruence/incongruence of the two dimensions of the stimuli, were recorded in order to study the timing of the semantic conflict. The reaction time to colour-word incongruent stimuli was significantly longer than the reaction time to congruent stimuli (the Stroop effect). A temporal Principal Components Analysis was applied to the data to identify the ERP components. Three positive components were identified in the 300–600 ms interval in response to the congruent and incongruent stimuli: First P3, P3b and PSW. The factor scores corresponding to the First P3 and P3b components were significantly smaller for the incongruent stimuli than for the congruent stimuli. No differences between stimuli were observed in the factor scores corresponding to the PSW or in the ERP latencies. We conclude that the temporal locus of the semantic conflict, which intervenes in generating the Stroop effect, may occur within the time interval in which the First P3 and P3b components are identified, i.e. at approximately 300–450 ms post-stimulus. We suggest that the semantic conflict delays the start of the response selection process, which explains the longer reaction time to incongruent stimuli. © 2009 Elsevier B.V. All rights reserved.
1. Introduction The Stroop interference effect is one of the most robust and bestknown interference effects in psychology. It was observed by J.R. Stroop in 1935 when participants were shown drawings of coloured squares and incongruent colour words (names of colours written in an incongruent colour with their meaning, e.g. the word ‘red’ written in blue). Stroop observed that the participants required more time to say the colour of the colour-incongruent words than the colour of the squares. Stroop (1935) attributed the longer response time to the interference provoked by the conflict between the meaning of the word and the colour in which it was written. Since these initial experiments were carried out, the task used to provoke the Stroop interference effect has been modified on numerous occasions. These modifications to the initial task designed by Stroop have made it possible to verify the robustness of the phenomenon and to determine the variables affecting its magnitude (see MacLeod 1991, for a review). In paradigms of the Stroop colour-word type, the participant usually carries out a verbal task (saying the colour of the stimulus), or a manual task to verify the reaction time (pressing a specific button according to the colour of the stimulus). Although different methods have been used to
⁎ Corresponding author. Department of Clinical Psychology and Psychobiology, University of Santiago de Compostela, Campus Sur s/n. 15782 Santiago de Compostela, Galicia, Spain. Tel.: +34 981 563100x13707; fax: +34 981 528071. E-mail address: [email protected] (M. Zurrón). 0167-8760/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.ijpsycho.2009.01.002
measure the Stroop effect, the general aim is to compare the time taken by the participant to carry out the task in response to incongruent stimuli (stimuli with two dimensions or characteristics that are contradictory, such as the word ‘red’ written in blue), with the time taken in response to congruent (e.g. the word ‘red’ written in red) and/or neutral stimuli (e.g.: the word ‘phase’ or the nonword ‘XXXX’ written in any colour, or patches of colour). The increase in response time to incongruent stimuli is considered to be due to the fact that processing the word, a relatively automatic process, creates conflict with the processing of the colour, which interferes in the processing of this dimension. Different hypotheses have been proposed to explain how word interference operates in relation to colour. MacLeod (1991) concluded that the “parallel distributed processing” (PDP) model, proposed by Cohen et al. (1990), best explains the 18 empirical characteristics of the Stroop Effect (see MacLeod, 1991 for a description of the 18 characteristics). According to this model, colour-word stimuli activate two processing pathways (simultaneously, but to different strengths), one for the word and another for the colour, and each pathway is formed by various processing units. Intersections may occur between the two processing pathways at the level of different units. The interference would be the result of conflict between units of different pathways at the locus of one or more intersections. The model allows that the effect of Stroop interference can be generated by different types of conflict. The PDP model therefore reconciles explanations of the Stroop effect that were considered antagonistic, such as the response conflict and the semantic conflict hypotheses.
M. Zurrón et al. / International Journal of Psychophysiology 72 (2009) 246–252
The response conflict hypothesis considers that the Stroop interference is produced by the conflict between two response plans, one for the colour and another for the word (Dyer, 1973a; Morton, 1969; Morton and Chambers, 1973; Posner and Snyder, 1975), whereas the semantic conflict hypothesis considers that the interference is produced by the conflict between two semantic codes, one generated for the word and another for the colour (Luo, 1999; Seymur, 1977). De Houwer (2003) and Van Veen and Carter (2005) have investigated the hypothesis that the Stroop effect may be caused by the two types of conflict. Three types of stimuli were presented in both studies: incongruent stimuli that provoked semantic and response conflicts, incongruent stimuli that provoked semantic conflict but no response conflict, and congruent stimuli. It was found that the reaction time to the first type of stimulus was longer than the reaction time to the second, and both were longer than the reaction time to the congruent stimuli. These findings support the hypothesis that both types of conflict intervene in generating the Stroop effect. Van Veen and Carter (2005) found, by use of functional magnetic resonance imaging (fMRI), that the cortical networks activated by each type of conflict are not overlaid. The Stroop test has been successfully adapted to record EventRelated Potentials (ERPs). The interest in complementing reaction time measurements with ERP recordings, in studying the Stroop effect, lies in the fact that ERP recording enables precise analysis of the timing of the cognitive processes that the subject uses in processing information and therefore the conflict. Several studies have shown that the amplitude of the ERPs elicited by congruent and incongruent colour-word stimuli differs in the range of latency between 300–450 ms, in Stroop tasks in which the participants were required to respond to the colour of the stimuli (Eppinger et al., 2007; Hanslmayr et al., 2008; Houston et al., 2004; Ilan and Polich, 1999; Kray et al., 2005; Liotti et al., 2000; MarkelaLerenc et al., 2004; McNeely et al., 2003; Potter et al., 2002; Qiu et al., 2006; Rebai et al., 1997; West, 2003, 2004; West and Alain, 1999, 2000a, 2000b; West et al., 2005) and in tasks in which participants were required to decide on the congruence/incongruence of the two dimensions of colour-word stimuli (Mager et al. 2007). Some authors have identified the P300 component for this range of latencies, and have shown that the amplitude of this component elicited by incongruent stimuli is smaller than the amplitude elicited by congruent stimuli (Houston et al., 2004; Ilan and Polich, 1999; Potter et al., 2002). In other studies (Eppinger et al., 2007; Hanslmayr et al., 2008; Kray et al., 2005; McNeely et al., 2003; Qiu et al., 2006; Rebai et al., 1997; West, 2003, 2004; West and Alain, 1999, 2000a, 2000b) a negative-going wave denominated N400 or N450 or Ni (negativity for incompatible trials) has been identified, with less positive voltage values (as negative voltages are rarely reached) for incongruent stimuli than for congruent stimuli. Yet other studies have identified both components (Mager et al. 2007; West et al., 2005), and finally, Liotti et al. (2000) and Markela-Lerenc et al. (2004) opted to quantify the amplitude by the voltage values obtained within different time windows, without specifying components, and obtained the same results. In addition to the disparity as regards the ERP components identified in this range of latencies, no consensus has been reached as regards the topographic distribution of the differences in amplitude between the ERPs elicited by the two types of stimuli. On the one hand Mager et al. (2007), Ilan and Polich (1999) and Potter et al. (2002) observed the greatest differences in amplitude at central–parietal electrodes, for a manual congruence/incongruence judgment task, a manual response to colour, and a verbal response to colour, respectively, whereas Hanslmayr et al. (2008), Houston et al. (2004),Markela-Lerenc et al. (2004) and West and Alain (1999, 2000a,b), observed the greatest differences in amplitude at centrofrontal locations for manual responses to colour. Finally, Liotti et al.
247
(2000),Eppinger et al. (2007) and Kray et al. (2005) observed that the differences in amplitude were widely distributed across the scalp for a manual response to colour. As regards the meaning of such differences in the ERP amplitudes in response to the two types of stimulus, the N400 (N450 or Ni) wave has been linked to the stimulus-related conflict (Eppinger et al., 2007; Kray et al., 2005; McNeely et al., 2003; West, 2003, 2004). This interpretation is very general and does not specify the stage of processing of the word-colour stimulus at which the conflict is produced, or in other words, what type of conflict in relation to the stimulus is involved in the Stroop effect. In the range of latencies between 300 and 450 ms it is therefore necessary to determine which of the ERP component(s) display differences in amplitude in response to the two types of stimuli, to clarify their topographical distribution, and to specify the underlying conflict. Tasks in which congruence/incongruence must be judged (Dyer, 1973b; Luo, 1999; Mager et al., 2007; Treisman and Fearnley, 1969; Zysset et al., 2001) have the advantage (relative to the colour-word response tasks) of guaranteeing that the intersection between the two processing pathways (colour and word), and therefore the conflict between them for incongruent stimuli, will occur before the motor response is produced, because both responses (whether congruent or not) require that the intersection has already been produced. In these tasks the semantic conflict associated with generation of the Stroop effect is therefore favoured with respect to the response conflict, better enabling analysis of the effect of the semantic conflict on the ERPs. The aims of the present study were therefore to determine: 1) the temporal locus of the colour-word semantic conflict involved in the Stroop effect; 2) the ERP component(s) associated with or modulated by the semantic conflict, and 3) the scalp distribution of such components. For this, we recorded ERPs elicited by colour-word stimuli in a congruence/incongruence judgment task, and applied temporal Principal Components Analysis to the ERP data. 2. Materials and methods 2.1. Participants Twenty one university students (7 female) aged between 19 and 24 (Mean: 20.95; SD = 1.75), with a medium–high socioeconomic status, participated as volunteers in the study. All had normal or corrected to normal vision (but no contact lenses). All were healthy with no history of neurological or psychiatric disorders or drug abuse, and were not receiving medication at the time of participating in the study (nor had received any in the preceding weeks). The participants were also required to abstain from drugs/alcohol/caffeine and nicotine prior to testing, and none reported fatigue caused by lack of sleep. None of the participants were familiar with the protocols used in the study. All participants were self-reported right-handed. 2.2. Task and stimuli The participants were presented with 104 stimuli separated into two blocks (i.e. each comprised 52 stimuli), with a rest period of 90 s between them. The stimuli were the Spanish words “azul” (blue), “verde” (green), “rojo” (red) and “gris” (grey) (each word was displayed 26 times), and were written in one of these four colours on a black background, so that the stimulus was either congruent with the colour and meaning of the word (the colour and meaning coincided), or incongruent (the word and colour did not coincide). Half of the stimuli were congruent and half incongruent. The stimuli were presented in a random order, and each word was displayed for 200 ms with an SOA of 2000 ms. The screen was black during the 1800 ms between the end of one stimulus and the onset of the next.
248
M. Zurrón et al. / International Journal of Psychophysiology 72 (2009) 246–252
The images were displayed on a flat 19-inch monitor placed 1 m from the subject's face. The words were presented in the centre of the screen in upper case Helvetica 2 typeface; each letter was 2 cm wide and 3.5 cm high, and subtended at an angle of 1.15 × 2.01° of arc. The participants were asked to press a button with their right thumb if both dimensions of the stimulus were incongruent, and to press another button with their left thumb if both dimensions of the stimulus were congruent. Prior to recording, a trial run was carried out with six stimuli to establish whether the subject understood the instructions and responded correctly. 2.3. EEG recording All of the participants were seated in a comfortable chair in an electrically-shielded laboratory with attenuated sound and lighting levels, and were instructed not to move during the recording session and to look at the centre of the screen. Each participant was fitted with a cap with a nose reference and a frontopolar ground, for recoding the electroencephalographic (EEG) activity at 30 electrode sites corresponding to the 10–20 system. Eye movements were recorded simultaneously with additional electrodes placed above and below the left eye (VEOG) and at the outer canthus of each eye (HEOG). All impedances were reduced to 5 KΩ or less in order to obtain good quality recordings, and care was taken to avoid damaging the subject's scalp. The EEG signals were passed through a 0.1–30 Hz band-pass filter and amplified at 22.5 K before being sampled at 500 Hz; the digitalized signal was then stored. Ocular artefacts were corrected off-line with the algorithm of Semlitsch et al. (1986). Pre-stimulus (200 ms) and post-stimulus (1400 ms) segments were extracted from the EEG, and the pre-stimulus baseline was corrected. Signals exceeding ±100 μV were automatically excluded from the averages. For the averages only the epochs corresponding to the stimuli with correct responses were taken into account. The number of epochs included in the ERP averages for each type of stimulus (congruent and incongruent) was between 45 and 50, and there were no differences between the number of epochs for either type of stimulus.
The tPCA provides a statistical decomposition of the brain electrical patterns superposed on the scalp-recorded data (Dien andFrishkoff, 2004). A covariance-matrix-based tPCA was applied for both conditions (congruent and incongruent stimuli). The decision regarding selection of the number of components (or factors) was based on the scree test (Cattell, 1966). Extracted factors were then subjected to Promax rotation. Dien (1998) showed that the use of Promax rotation improves the accuracy of the results and reduces problems such as misallocation of variance (see also Dien andFrishkoff, 2004). The tPCA provides two matrices, one for factor loadings and another for factor scores. The first shows the load of each factor through time, and the second provides information about the extent to which each factor is present in the averaged ERPs (i.e. at each electrode site). The factor scores are transformed values of the original voltages and thus can be used as a measure of the amplitude. The scores of each Temporal Factor were examined by repeated measures analysis of variance (ANOVAs), with two within-subject factors: Congruence (with 2 levels: congruent and incongruent) and Electrode (with 3 levels: Fz, Cz and Pz). As regards the latencies of the ERP components, the peak latencies were analyzed by repeated measures ANOVAs, with two withinsubject factors: Congruence (with 2 levels: congruent and incongruent) and Electrode (with 2 or 3 levels), and by paired sample ttests, depending on the component. When the ANOVA for the latencies and for the factor scores revealed significant results, pairwise comparison of means (with Bonferroni corrections) was carried out to identify the source of the differences. Greenhouse–Geisser corrections to the degrees of freedom were applied in all cases in which the condition of sphericity was not met. In these cases, the original degrees of freedom are presented together with the corrected p and ε values. Differences were considered significant at p ≤ 0.05. All statistical analyses were carried out with SPSS (version 15.0). 3. Results 3.1. Behavioural results
2.4. Statistical analysis Paired sample t-tests were used to compare the reaction times (RTs) and the number of errors in response to congruent and incongruent stimuli. A temporal Principal Components Analysis (tPCA) was applied to the ERP data to ensure correct identification of the ERP components.
The RT was significantly longer in response to the incongruent stimuli (Mean = 623 ms; SD = 84) than in response to the congruent stimuli (Mean = 563 ms; SD = 92), t(13) = −7.2, p b 0.0001. There was no significant difference in the number of errors between the two types of stimuli (Congruent: Mean = 2.7; SD = 2.3; Incongruent: Mean = 2.3; SD = 2.1).
Fig. 1. Temporal Factor (TF) loadings (Y-axis) in the Principal Components Analysis for ERP data. X-axis: Time (ms)
M. Zurrón et al. / International Journal of Psychophysiology 72 (2009) 246–252
3.2. Temporal Principal Components Analysis (tPCA) Fig. 1 shows the Temporal Factor (TF) loadings, and Fig. 2 illustrates the First P3, P3b, and PSW components in the grand-averaged EventRelated Potential waveforms, at Fz, Cz and Pz electrode sites, and the
249
voltage maps showing the topographical distribution of the amplitude (μV) on the peak latencies of the three components. The mean factor scores of TF2, TF7 and TF4, at the Fz, Cz, and Pz electrode sites and the mean latencies (ms) of First P3, P3b, and PSW, elicited by the congruent and incongruent stimuli are shown in
Fig. 2. Grand average ERP waveforms at Fz, Cz and Pz for congruent (red line) and incongruent (blue line) stimuli, and the voltage maps showing the topographical distribution of the amplitude (μV) on the peak latencies of the First P3, P3b, and PSW components.
250
M. Zurrón et al. / International Journal of Psychophysiology 72 (2009) 246–252
Table 1 Mean factor scores for TF2 (First P3), TF7 (P3b), and TF4 (PSW), at the Fz, Cz, and Pz electrode sites, and mean latencies (ms) of the First P3, P3b, and PSW components elicited by the congruent and incongruent stimuli. Standard deviations are shown in brackets. TF: Temporal Factor, PSW: Positive Slow Wave. Factor scores
Congruent
Incongruent
TF2 (First P3)
TF7 (P3b)
TF4 (PSW)
TF2 (First P3)
TF7 (P3b)
TF4 (PSW)
Fz Cz Pz
− 0.04 (0.97) − 0.08 (1.15) 0.74 (1.08)
0.09 (0.85) 0.68 (0.91) 0.9 (1.07)
− 0.40 (0.87) 0.47 (0.72) 0.98 (0.91)
− 0.24 (0.75) − 0.35 (1.08) 0.42 (1.01)
− 0.11 (0.88) 0.42 (0.96) 0.58 (1.17)
− 0.33 (0.86) 0.37 (0.53) 0.84 (0.76)
Latencies
First P3
P3b
PSW
First P3
P3b
PSW
308 (18.1)
380 (23) 379 (27.5) 384 (33.3)
311 (19.6)
397 (56.5) 395 (53.3) 396 (57.6)
578 (62.8) 578 (59.3)
Fz Cz Pz
565 (62.8) 559 (63.5)
Table 1; the F values in repeated-measures ANOVAs for the factor scores of the TF2, TF7, and TF4, and for the latencies of P3b and PSW are shown in Table 2. The tPCA identified eight TF, which explained 91.7% of total variance. Taking into account the temporal range of the factor loadings and that the scores for the factors were positive and maximal at parietal electrodes, three of these TF appeared to correspond to the ERP components observed in the 300–600 ms interval. The TF2 corresponded to a posterior positive component, which was denominated First P3. The peak latency of the First P3 component was measured at Pz in the 290–330 ms interval. The TF7 corresponded to the positive P3b central-parietal component described by Sutton et al. (1965). The peak latency of P3b was measured at Fz, Cz and Pz in the 360–430 ms interval. The TF4 corresponded to a parietal–central Positive Slow Wave, which was denominated PSW. The peak latency of PSW was measured at Cz and Pz in the 540–600 ms the interval. Temporal Factor 2 (corresponding to First P3 component) showed significantly higher scores in response to congruent stimuli than incongruent stimuli, and significantly higher at the Pz than at the Cz and Fz electrode sites. Temporal Factor 7 (corresponding to P3b component) presented significantly higher scores in response to congruent stimuli than incongruent stimuli, and significantly higher at the Pz and Cz than at the Fz electrode site. Temporal Factor 4 (corresponding to the PSW component) showed significantly higher scores at Pz than Cz and Fz, and significantly higher at the Cz than at the Fz electrode site. No differences were found for the factor scores in response to the congruent and incongruent stimuli, at the electrodes evaluated.
Table 2 F values in repeated-measures ANOVA for the factor scores for TF2 (First P3), TF7 (P3b), and TF4 (PSW), and for the latencies of P3b and PSW (ε = epsilon value, C = Congruence factor, E = Electrode factor). Factor scores
TF2 (First P3)
TF7 (P3b)
TF4 (PSW)
ANOVA (C × E) C E C×E
5.43⁎ 12.18⁎⁎⁎ (ɛ: 0.77) 1.21 (ɛ: 0.77)
5.85⁎ 10.27⁎⁎ (ɛ: 0.68) 0.75 (ɛ: 0.57)
0.34 27.33⁎⁎⁎ (ɛ: 0.66) 5.71⁎⁎
P3b
PSW
2.33 0.31 0.35 (ɛ: 0.73)
1.82 0.57 1.44
The latency of First P3 component did not show any differences in response to congruent and incongruent stimuli (t(13) = −0.45, p b 0.66). The latencies of the P3b and PSW components were not different, either in response to congruent or incongruent stimuli, or among electrode sites. 4. Discussion 4.1. Task performance A clear Stroop effect was obtained, as the RT to the incongruent stimuli was significantly longer than the RT to the congruent stimuli. The stimulation paradigm used, adapted to the recording to ERP and with congruence/incongruence judgment response, was therefore suitable for studying the effect of Stroop colour-word interference. This also indicates that with congruence/incongruence judgment tasks, the conflict between the semantic code of the word that is read and the semantic code activated by the colour perceived plays an essential role in generating the Stroop effect, which supports the importance of this conflict in colour-response Stroop tasks. The task for judging the congruence/incongruence of both dimensions (chromatic and linguistic) of the stimulus requires the controlled processing of both dimensions (see criteria for automaticity/control in: Kahneman and Treisman, 1984; Neumann, 1984; Shiffrin, 1988). However, in the response to the colour of the stimuli, it is assumed that processing of the linguistic dimension is relatively automatic. Therefore, the Stroop effect is obtained both when the processing of the linguistic dimension is relatively automatic and when the processing is controlled. Therefore, although the degree of automation of a task affects the interference, so that the most automatic task interferes in the least automatic task (MacLeod, 1991; MacLeod and Dunbar, 1988), it is not necessary for one of the two dimensions of the stimulus to be processed automatically (or in a relatively automatic way) for the Stroop effect to be generated. 4.2. ERPs
Latency ANOVA (C × E) C E C×E
3.3 Latency of ERP components
–
⁎⁎⁎p b 0.0001, ⁎⁎p b 0.01, ⁎p b 0.05. For the factor scores, the degrees of freedom for the Congruence factor were 1/20, and for the Electrode factor and the Congruence × Electrode interaction were 2/40. For the latencies, the degrees of freedom for the Congruence factor was 1/15 in P3b and 1/16 in PSW, and for the Electrode factor and the Congruence × Electrode interaction were 2/30 in P3b and 1/16 in PSW.
4.2.1. First P3 and P3b The scores for the TF2 and TF7 were significantly smaller for incongruent stimuli than for congruent stimuli, and therefore the amplitudes of the ERP components –First P3 and P3b– were smaller for incongruent than for congruent stimuli. Given the latency, polarity and topography of the First P3, we consider that it belongs to the P300 “family”, and we will therefore interpret it as such. Taking into account that the P300 amplitude is negatively correlated with the difficulty of the task (Donchin, 1981; Donchin and Coles, 1988; Johnson, 1986; Verleger, 1988), the smaller First P3 and P3b amplitudes may reflect greater difficulty in processing the incongruent stimuli than the congruent stimuli. This greater difficulty may be the result of the
M. Zurrón et al. / International Journal of Psychophysiology 72 (2009) 246–252
semantic conflict generated by incongruent stimuli. The smaller amplitude of First P3 and P3b in response to incongruent stimuli may be considered as an index of the differential semantic processing of these stimuli relative to the congruent stimuli, and therefore the smaller amplitude in the interval between 300 and 450 ms in response to incongruent stimuli may indicate the temporal locus of the semantic conflict, i.e. the intersection between the unit that processes the meaning of the word and the unit that processes the meaning of the colour might occur around 300 and 450 ms post-stimulus. The differences in amplitude between the ERP elicited by both stimuli are found at the three midline electrodes, in accordance with the results of Eppinger et al. (2007), Kray et al. (2005) and Liotti et al. (2000), and in contrast with the results of other studies in which greater differences in amplitude were observed at centro-parietal (Ilan and Polich, 1999; Potter et al., 2002) and centro-frontal electrodes (Hanslmayr et al., 2008; Houston et al., 2004; MarkelaLerenc et al., 2004 West and Alain, 1999, 2000a,b). Such discrepancies may be due to the combination of two aspects in which the studies differ: the different measures of amplitude considered (amplitude of the P300 component, or of the N400 wave or of the mean voltage in a time window); and different types of response to the word-colour stimuli (manual compared with verbal response, response to colour of the stimulus compared with congruence judgment). In the present congruence/incongruence judgment task, we did not observe any independent TF that corresponds to a N400 (N450 or Ni) ERP component. This, in conjunction with the significantly different TF2 and TF7 scores for congruent and incongruent stimuli, allow us to conclude that, in this type of task, the “N400” wave is not a component of the ERPs. The results therefore indicate that, at least in congruence judgment tasks, the differences in amplitude between the two types of stimuli in the range 300–450 ms may be related to the difference in amplitude of the P300 complex. One may wonder if the N400 (N450 or Ni) ERP wave, described in previous studies involving colour response tasks (Eppinger et al., 2007; Hanslmayr et al., 2008; Kray et al., 2005; McNeely et al., 2003; Qiu et al., 2006; Rebai et al., 1997; West, 2003, 2004; West and Alain, 1999, 2000a, 2000b) corresponds to a real component or also reflects differences in the P300 complex amplitude. In light of the above, in future studies in which ERPs are evaluated in Stroop tasks, with different types of response to word-colour stimuli, it would be advisable to use tPCA for correct identification of the ERP components. There were no differences in latencies of First P3 and P3b between the two types of stimuli. The latency of P300 has received particular interest in studies of ERPs elicited by Stroop tasks. In this respect, the results of the present study are consistent with those of previous studies in which Stroop colour response tasks for incongruent and congruent (or neutral) stimuli were used (Atkinson et al., 2003; Duncan-Johnson and Kopell, 1980, 1981; Grapperon et al., 1998; Ilan and Polich, 1999; Lavoie, 1999; Rosenfeld and Skogsberg, 2006; West and Alain, 2000a). It has been suggested (Atkinson et al., 2003; Duncan-Johnson and Kopell, 1981; Ilan and Polich, 1999; Lavoie, 1999; Rosenfeld and Skogsberg, 2006) that the lack of any differences in the latencies of this component rules out any possible influence of stages prior to the elaboration of responses in the generation of the Stroop effect. This argument is based on the fact that the latency of P3b was initially considered as an index of the stimulus evaluation time that is independent of the response processes (Donchin, 1981; Kutas et al., 1977). It was reasoned that if conflict is produced between two motor programmes, the latency of this component should not differ in response to incongruent and congruent (or neutral) stimuli. However, in light of the results of the present study, we consider that the absence of differences in the latencies of the ERPs elicited by congruent and incongruent stimuli cannot be used as data to support the response conflict rather than the semantic conflict, as in the present study a congruence/incongruence judgment task was used,
251
thereby minimizing the effect that the conflict between the responses may have on Stroop interference. Consequently, the absence of differences in latencies of the components of the ERPs studied does not rule out the relevance of the semantic intersection stage in the genesis of the Stroop effect. Pfefferbaum et al. (1986) and Verleger (1997) proposed that in complex tasks, evaluation of the stimulus and selection of the response should be processed serially, so that selection of the response begins after P300; in contrast, in simple tasks, evaluation of the stimulus and selection of the response may be processed in parallel, and therefore in the temporal interval in which P300 appears, the subject will also be selecting the response. If we consider that in a congruence/incongruence task the response to congruent stimuli is simple, because of the lack of conflict, the process of selecting the response may begin during the process of evaluating the stimulus, i.e. in the temporal interval of the P300 complex. However, the semantic conflict of the incongruent stimulus should be resolved before the response is selected. In other words, the incongruent stimulus requires that the processes of evaluating the stimulus and selection of the response occur serially, so that selection of the response will begin after evaluation of the stimulus (after the P300 complex). As a result, a delay in the start of the selection of the response to the incongruent stimulus, with respect to the congruent stimulus, will explain the delay in the reaction time to the former, with respect to the latter. Furthermore, this is consistent with a recent interpretation of P300, in which this component reflects some process that mediates between the perceptual analysis of the stimuli and initiation of the response processes (Verleger et al., 2005). 4.2.2. The Positive Slow Wave (PSW) There were no differences in either the latency or amplitude of the parieto-central PSW component between conditions. Mager et al. (2007) used a Stroop congruence–incongruence judgment task, and also observed a positive component after P300 with central–parietal distribution and a latency peak at 550 ms. This late positive component has also been identified in colour response tasks (West and Alain, 1999, 2000a,b), and was denominated Sustained Potential (SP) by West and colleagues (McNeely et al., 2003; West, 2003, 2004; West et al., 2005). A similar positive component was identified by Squires et al. (1975), in oddball tasks in which the subject was required to count the target stimuli; this component appeared after P300 (P3b), with a maximum amplitude at Pz, and did not reach positive values at Fz. Although it is not clear which cognitive process is related to the PSW component, in order for this to be elicited, it appears necessary to use tasks in which the subject has to do more than detect a given stimulus, and for example, has to select one from amongst a number of possible answers (Perchet and García-Larrea, 2000). García-Larrea and Cézanne-Bert (1998) considered that in general terms, the PSW component is a non-specific activity that indicates completion of a synchronized operation immediately after detection of the target. The data obtained in the present study also appear to support the relationship between the PSW component obtained from the Stroop task with processes that are subsequent to categorization of the stimuli as congruent or incongruent. The latency of the PSW elicited by congruent stimuli is coincident with the RT, whereas for incongruent stimuli the latency was shorter than the RT and therefore the present results do not support the interpretation of West et al. (2005), who related the SP component to the process of selecting the response. We suggest that the PSW may be associated with review of the stimulus categorization, a process that may occur in parallel with the readiness and/or execution of the motor response. In summary, without ruling out the idea that a conflict between two response plans may intervene in the generation of the Stroop effect, in a congruence/incongruence judgment task, the incongruent
252
M. Zurrón et al. / International Journal of Psychophysiology 72 (2009) 246–252
colour-word stimuli provoked a conflict between two semantic codes (colour and word) which lengthened the reaction time to these stimuli (Stroop effect). This semantic conflict was manifested as a smaller amplitude of First P3 and P3b elicited by incongruent stimuli relative to those elicited by congruent stimuli, and therefore the temporal locus of the semantic conflict may occur at between 300 and 450 ms post-stimulus. It is suggested that the semantic conflict delays the start of the selection of the response, which may explain the longer reaction time to incongruent stimuli. References Atkinson, C.M., Drysdale, K.A., Fulham, W.R., 2003. Event-related potential to Stroop and reverse Stroop stimuli. Int. J. Psychophysiol. 47, 1–21. Cattell, R.B., 1966. The scree test for the number of factors. Behav. Res. 1, 245–276. Cohen, J.D., Dunbar, K., McClelland, J.L., 1990. On the control of automatic processes: a parallel distributed processing account of the Stroop effect. Psychol. Rev. 97, 332–361. De Houwer, J., 2003. On the role of stimulus–stimulus and stimulus–response compatibility in the Stroop effect. Mem. Cognit. 31, 353–359. Dien, J., 1998. Addressing misallocation of variance in principal component analysis of event-related potentials. Brain Topogr. 11, 43–55. Dien, J., Frishkoff, G.A., 2004. Principal components analysis of ERPs data. In: Handy, T.C. (Ed.), Event-Related Potentials. A Methods Handbook. MIT Press, Cambridge, pp. 189–207. Donchin, E., 1981. Surprise!...Surprise? Psychopysiology 18, 493–513. Donchin, E., Coles, M.G.H., 1988. Is the P300 component a manifestation of context updating? Behav. Brain Sci. 11, 355–372. Duncan-Jonhson, C.C., Kopell, B.S., 1980. The locus of interference in the Stroop task: when you read “blue”, do you see “red”? Psychophysiology 17, 308–309. Duncan-Jonhson, C.C., Kopell, B.S., 1981. The Stroop effect: brain potentials localize the source of interference. Science 214, 938–940. Dyer, F.N., 1973a. The Stroop phenomenon and its use in the study of perceptual, cognitive and response processes. Mem. Cognit. 1, 106–120. Dyer, F.N., 1973b. Same and different judgments for word-color pairs with “irrelevant” words or colors: evidence for word-code comparisons. J. Exp. Psychol. 98, 102–108. Eppinger, B., Kray, J., Mecklinger, A., John, O., 2007. Age differences in task switching and response monitoring: evidence from ERPs. Biol. Psychol. 75, 52–67. García-Larrea, L., Cézanne-Bert, G., 1998. P3, positive slow wave and working memory load: a study on the functional correlates of slow wave activity. EEG & Clin. Neurophysiol. 108, 260–273. Grapperon, J., Vidal, F., Leni, P., 1998. The contribution of cognitive evoked potentials to knowledge of mechanism on the Stroop test. Neurophysiol. Clin. 28, 207–220. Hanslmayr, S., Pastötter, B., Bäuml, K., Gruber, S., Winber, M., Klimesch, W., 2008. The electrophysiological dynamics of interference during the Stroop task. J. Cogn. Neurosci. 20, 215–225. Houston, R.B., Bauer, L.O., Hesselbrock, V.M., 2004. Effects of borderline personality disorder features and a family history of alcohol or drug dependence on P300 in adolescents. Int. J. Psychophysiol. 53, 57–70. Ilan, A.B., Polich, J., 1999. P300 and response time from a manual Stroop task. Clin. Neurophysiol. 110, 367–373. Jonhson Jr., R., 1986. A triarchic model of P300 amplitude. Psychopgysiology 23, 367–384. Kahneman, D., Treisman, A., 1984. Changing views of attention and automaticity. In: Parasuraman, R., Davis, D.R. (Eds.), Varieties of Attention. Academic Press, New York, pp. 29–61. Kray, J., Eppinger, B., Mecklinger, A., 2005. Age differences in attentional control: an event-related potential approach. Psychophysiology 42, 407–416. Kutas, M., McCarthy, G., Donchin, E., 1977. Augmenting mental chronometry: the P300 as a measure of stimulus evaluation time. Science 197, 792–795. Lavoie, M.E., 1999. Toward a functional explanation of the locus of the Stroop interference: a psychophysiological study. Brain Cogn. 40, 167–170. Liotti, M., Woldorff, M.G., Perez, R., Mayberg, H.S., 2000. An ERP study of the temporal course of the Stroop color-word interference effect. Neuropsychologia 38, 701–711. Luo, C.R., 1999. Semantic competition as the basis of Stroop interference: evidence from color-word matching task. Psychol. Sci. 10, 35–40. MacLeod, C.M., 1991. Half a century of research on the Stroop effect: an integrative review. Psychol. Bull. 109, 163–203. MacLeod, C.M., Dunbar, K., 1988. Training and Stroop-like interference: evidence for a continuum automaticity. J. Exp. Psychol. Learn. Mem. Cognit. 10, 304–315. Mager, R., Bullinger, A.H., Brand, S., Schmidlin, M., Schärli, H., Müller-Spahn, F., Störmer, R., Falkenstein, M., 2007. Age-related changes in cognitive conflict processing: an event-related potential study. Neurobiol. Aging 28, 1925–1935.
Markela-Lerenc, J., Ille, N., Kaiser, S., Fiedler, P., Mundt, C., Weisbrod, M., 2004. Prefrontal-cingulate activation during executive control: which comes first? Cogn. Brain Res. 18, 278–287. McNeely, H., West, R., Christensen, B.K., Alain, C., 2003. Neurophysiological evidence for disturbances of conflict processing in patients with schizophrenia. J. Abnorm. Psychol. 112, 1–9. Morton, J., 1969. Categories of interference: verbal mediation and conflict in card sorting. Br. J. Psychol. 60, 329–346. Morton, J., Chambers, S.M., 1973. Selective attention to words and colours. Q. J. Exp. Psychol. 25, 387–397. Neumann, O., 1984. Automatic processing: a review of recent findings and a plea for an old theory. In: Prinz, W., Sanders, A.F. (Eds.), Cognition and Motor Processes. Springer Verlag, Berlin, pp. 255–293. Perchet, C., García-Larrea, L., 2000. Visuospatial attention and motor reaction in children: an electrophysiological study of the “Posner” paradigm. Psychophysiology 37, 231–241. Pfefferbaum, A., Christensen, C., Ford, J.M., Kopell, B.S., 1986. Apparent response incompatibility effects on P3 latency depend on the task. EEG & Clin. Neurophysiol. 64, 424–437. Posner, M.I., Snyder, C.R.R., 1975. Attention and cognitive control. In: Solso, R.L. (Ed.), Information Processing and Cognition. The Loyola Symposium, Erlbaum, Hillsdale NJ, pp. 55–85. Potter, D.D., Jory, S.H., Bassett, M.R.A., Barrett, K., Mychalkiw, W., 2002. Effect of mild head injury on event-related potential correlates of Stroop task performance. J. Int. Neuropsychol. Soc. 8, 828–837. Qiu, J., Luo, Y., Wang, Q., Zhang, F., Zhang, Q., 2006. Brain mechanism of Stroop interference effect in Chinese characters. Brain Res. 1072, 186–193. Rebai, M., Bernard, C., Lannou, J., 1997. The Stroop test evokes a negative brain potential, the N400. Int. J. Neurosci. 91, 85–94. Rosenfeld, J.P., Skogsberg, K.R., 2006. P300-based Stroop study with low probability and target Stroop oddballs: the evidence still favours the response selection hypothesis. Int. J. Psychophysiol. 60, 240–250. Semlitsch, H.V., Anderer, P., Schuster, P., Presslich, O., 1986. A solution for reliable and valid reduction of ocular artefacts applied to the P300 ERP. Psychophysiology 23, 695–703. Seymur, P.H., 1977. Conceptual encoding and locus of the Stroop effect. Q. J. Exp. Psychol. 29, 245–265. Shiffrin, R.M., 1988. Attention. In: Atkinson, R.C., Hernstein, R.J., Lindzey, G., Luce, R.D. (Eds.), Steven's Handbook of Experimental Psychology. Wiley, New York, pp. 739–911. Squires, N.K., Squires, K., Hillyard, S.A., 1975. Two varieties of long-latency positive waves evoked by unpredictable auditory stimuli in man. EEG & Clin. Neurophysiol. 38, 387–401. Stroop, J.R., 1935. Studies of interference in serial verbal reactions. J. Exp. Psychol. 18, 643–662. Sutton, S., Braren, M., Zubin, J., John, E.R., 1965. Information delivery and the sensory evoked potential. Science 150, 1187–1188. Treisman, A., Fearnley, S., 1969. The Stroop test: selective attention to colours and words. Nature 222, 437–439. Van Veen, V., Carter, C.S., 2005. Separating semantic conflict and response conflict in the Stroop task: a functional MRI study. NeuroImage 27, 497–504. Verleger, R., 1988. Event-related potentials and memory: a critique of the context updating hypothesis and an alternative interpretation of P3. Behav. Brain Sci. 11, 341–354. Verleger, R., 1997. On the utility of P3 latency as an index of mental chronometry. Psychophysiology 34, 131–156. Verleger, R., Jaskowski, P., Wascher, E., 2005. Evidence for an integrative role of P3b in linking reaction to perception. J. Psychophysiol. 19, 165–181. West, R., 2003. Neural correlates of cognitive control and conflict detection in the Stroop and digit-location tasks. Neuropsychologia 41, 1122–1135. West, R., 2004. The effects of aging on controlled attention and conflict processing in the Stroop task. J. Cogn. Neurosci. 16, 103–113. West, R., Alain, C., 1999. Event-related neural activity associated with the Stroop task. Cogn. Brain Res. 8, 157–164. West, R., Alain, C., 2000a. Age-related decline in inhibitory control contributes to the increased Stroop effect observed in older adults. Psychophysiology 37, 179–189. West, R., Alain, C., 2000b. Effects of task context and fluctuations of attention on neural activity supporting performance of Stroop task. Brain Res. 873, 102–111. West, R., Jakubek, K., Wymbs, N., Perry, M., Moore, K., 2005. Neural correlates of conflict processing. Exp. Brain Res. 167, 38–48. Zysset, S., Müller, K., Lohmann, G., von Cramon, D.Y., 2001. Color-word matching Stroop task: separating interference and response conflict. NeuroImage 13, 29–36.