Gender-Selective Effects of the P300 and N400 Components of the Visual

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Available online at www.sciencedirect.com

Vision Research 48 (2008) 917–925 www.elsevier.com/locate/visres

Gender-selective eﬀects of the P300 and N400 components of the visual evoked potential

Scott C. Steﬀensen *, Allison J. Ohran, Daniel N. Shipp, Kimberly Hales, Sarah H. Stobbs, Donovan E. Fleming

Department of Psychology (1050 SWKT), Brigham Young University, Provo, UT 84602, USA

Received 11 May 2007; received in revised form 31 October 2007

Abstract

There is an ongoing controversy regarding the role of gender in modulating components of the human visual-evoked potential (VEP) and event-related potentials (ERPs). Our aim was to further characterize the role of gender on VEPs, ERPs and response performance in an object recognition task. We recorded VEPs and reaction time (RT) in a paradigm wherein subjects responded to a randomly presented ‘‘Relevant” stimulus, and did not respond when presented with ‘‘Irrelevant” or ‘‘Standard” visual stimuli. There was no effect of gender on early components of the VEP or RT to Relevant stimuli. Relevant and Irrelevant stimuli evoked distinct VEP components including the P300, N400 and late-positive (LP) ERPs that were well-discriminated from those of the Standard stimulus. Females were character- ized by greater P300 and N400 responses than males for the Relevant stimulus, but exclusively greater N400 responses for the Irrelevant stimulus. There were no significant gender differences for the LP, or for the latency of any ERP component. Gender differences were not attributed to hemispheric asymmetry, as there were no significant differences in P300 and N400 VEP amplitudes between lateral occipital or parietal electrode positions. These results indicate that the N400 can be elicited in a task requiring the processing of irrelevant, but not unexpected, stimuli and that females process visual information differently than males, perhaps by increased allocation of attentional resources to distracting stimuli. Ó 2008 Elsevier Ltd. All rights reserved.

Keywords: Attention; Event-related potentials; Reaction time; Vision; Gender; P300; N400; Distractor; Object recognition

1. Introduction nent of the VEP is influenced by the probability that a given object will appear, hence it is considered to be an The visual scene contains an unlimited number of event-related potential (ERP), and is a measure of atten- objects, and observers must selectively attend to and differ- tion allocation, as well as the difficulty of discriminating entiate stimuli. Various investigations have demonstrated between the target object and the standard stimuli (Dun- the utility of the visual evoked potential (VEP) in the study can-Johnson & Donchin, 1977; Isreal, Wickens, & Don- of visual information processing (Acosta & Nasman, 1992; chin, 1980; Katayama & Polich, 1998; Kramer & Strayer, Bennington & Polich, 1999; Conill, 1998; Hillyard, Man- 1988; Polich, 1986; Polich & Bloom, 1988; Wickens, Kra- gun, Luck, & Heinze, 1990; Katayama & Polich, 1999; mer, Vanasse, & Donchin, 1983). The N400 component Mertens & Polich, 1997, Naumann et al., 1992; Picton of the VEP is also an ERP and is thought to reflect contex- et al., 2000; Polich, 1998; Pritchard, 1981; Sutton, Braren, tual integration and is often associated with semantic pro- Zubin, & John, 1965). The amplitude of the P300 compo- cessing under non-congruent or unexpected conditions in both visual and auditory paradigms (Finnigan, Humph- reys, Dennis, & Geffen, 2002; Gunter, Friederici, & Schrie- * Corresponding author. Fax: +1 801 422 0602. fers, 2000; Hagoort, 2003; Hagoort & Brown, 1999; E-mail address: scott_steff[email protected] (S.C. Steffensen). Koyama, Nageishi, & Shimokochi, 1992; Kutas &

Federmeier, 2000). However, some studies have shown verter. The computer monitored artifacts online and rejected the EEG there may be an analogous N400 elicited in non-linguistic during stimulus trials that contained artifacts associated with excessive paradigms, for example, when faces or relatively complex head, muscle or eye-blink movements. Evoked potentials were acquired in 2 s epochs during each visual stimulus. They began 100 ms prior to pictures are used as an unexpected or novel stimulus (Bar- the presentation of a visual stimulus on the computer monitor and rett, Rugg, & Perrett, 1988; Ganis & Kutas, 2003; Jemel, ended 1900 ms after the stimulus. Each stimulus presentation was George, Olivares, Fiori, & Renault, 1999; Olivares, Igle- replaced by a blank screen. We recorded the subject’s reaction time sias, & Rodriguez-Holguin, 2003; West & Holcomb, (RT) from the time of presentation of the visual stimulus to the time 2002). Traditionally, N400 paradigms involve unprimed the subject pressed the switch. or unexpected targets (Bentin, McCarthy, & Wood, 1985; 2.3. Behavioral task Deacon, Mehta, Tinsley, & Nousak, 1995; Kim, Kim, & Kwon, 2001). At the beginning of the session, each subject read a standard set of A number of studies have demonstrated mixed gender- instructions displayed on the screen that described the task they were to related effects on VEPs, ERPs and auditory evoked poten- perform and showed the visual stimuli that would be presented. A brief tials. Some have found lower amplitudes in males or verbal explanation was given and any questions were answered by the shorter latencies in females for early VEP components, experimenter. Visual stimuli were generated by a PC-type computer and displayed for two seconds on a 1700 monitor, 61 cm in front of including the N50, P100, N100 and N200 (Ehlers, Wall, the subject. The stimulus subtended visual angles of 19.5° horizontally Garcia-Andrade, & Phillips, 2001; Mitchell, Howe, & and 12.2° vertically and was centered on the monitor. The visual stimuli Spencer, 1987), while others have found both increased consisted of three randomly presented 4 4 matrices with 15 right-fac- amplitude and decreased latency for the early components ing arrows (white figures on black background) and one target embed- of the female VEP (Chu, 1987). The ERP components of ded at random positions in the matrix (Fig. 1). Subjects were asked to respond to a diamond embedded in the 4 4 matrix, which served as the VEP have shown similar gender variation. Using both the ‘‘Relevant” stimulus, and to ignore matrices consisting of all visual and auditory oddball tasks, Hoffman and Polich right-facing arrows, which served as the ‘‘Standard” stimulus, or a dia- (1999) reported that females have a larger P300 component mond with lines through it, which served as the ‘‘Irrelevant” stimulus. than males. However, some studies contradict this by The objects in the Relevant and Irrelevant matrices appeared randomly showing there is no significant difference between male in any of the 16 positions of the 4 4 matrix. The subjects were direc- ted to respond, by pressing a hand-switch, to the Relevant stimulus and and female auditory P300 (Polich, 1986) or visual and audi- not to respond to Irrelevant or Standard stimuli. Each session began tory P300 (Sangal & Sangal, 1996). Our aim was to further with a block of 10 practice presentations. Visual feedback regarding tar- evaluate the effects of gender on VEPs and ERPs evoked by get detection and reaction time (RT) was displayed on the computer the presentation of relevant and irrelevant stimuli in a screen to the subjects immediately after each trial. Subjects were shown visual object recognition task. As far as we know, no stud- their RT (measured with a 1 ms precision) when they responded to the Relevant stimulus and a ‘NO’ when they responded to the Irrelevant or ies to date have evaluated late components of the VEP, Standard stimuli. The parameters that were monitored for the subject’s other than the P300, under these conditions. We will show response to the Relevant stimulus included RT, sensitivity (percent hits supportive evidence for gender differences in the P300, as and misses), and accuracy (percent false alarms), but only RT was mea- well as new evidence suggesting that the N400 is a more sured and analyzed statistically. sensitive index of gender differences. 2.4. Visual evoked potentials

2. Methods Visual evoked potentials were obtained from all subjects by averag- ing two seconds of EEG around each visual stimulus. We averaged 54 2.1. Participants stimulus presentations for each of the three matrices (i.e., Relevant, Irrelevant, and Standard stimuli) at each of the electrode positions. Subjects were recruited from undergraduate level neuroscience and We measured the latency and amplitude of each of the peaks of the psychology courses at Brigham Young University and ranged in age from within-subject averaged VEP components N50, P100, N100, N200, 18–25 years (mean age = 22 ± 0.4 years). Fifteen male and 15 female sub- P300, N400 and late-positive (LP). Specific components of each subjects were asked to participate in two 15 min electroencephalographic ject’s averaged VEP were identified by using a digital location grid (EEG) recording sessions wherein different visual stimuli were presented. on the computer, and were analyzed in terms of peak to peak ampli- Each participant gave signed, informed consent (approved by the Brigham tude and peak latency. We then averaged specific components of the Young University Institutional Review Board) prior to the initiation of VEP recorded at all electrode positions across subjects for each stimulus the procedures. condition.

2.2. Electroencephalographic and reaction time recordings 2.5. Data analysis and statistics

Subjects were seated in a sound-attenuated, electrically shielded Inspection of grand averaged potentials at each electrode location of room in front of a 1700 computer monitor. EEG was recorded from the 10–20 International System revealed that the VEP waveforms evoked standard International 10–20 System locations Fz, C3, Cz, C4, T3, at the parietal and occipital electrodes contained the most well-defined T4, P3, Pz, P4, O1, Oz and O2 using electrodes embedded in an elastic combination of early and late components in association with the Rele- cap and electro-gel. These sites were referenced to linked mastoids that vant and Irrelevant visual stimuli. Thus, we limited our statistical analysis were attached with electro-gel to the ears. Cortical potentials were to the specific components of VEPs evoked at Oz, O1, O2, P3 and P4 sites. amplified and filtered (0.3–30 Hz) using a Grass Instruments Model Measures of reaction time (RT) and VEP components at Oz, O1, O2, P3 12c-16-35 neurodata acquisition system, and sampled at 500 Hz (12 and P4 sites were analyzed in 15 male and 15 female subjects with bit voltage resolution) with a Scientific Solutions Labmaster A/D con- ANOVA, with gender as a between-subjects factor and condition (Rele- S.C. Steffensen et al. / Vision Research 48 (2008) 917–925 919

RELEVANT IRRELEVANT STANDARD

+ RELEVANT 10 µV IRRELEVANT - STANDARD

C3 C4

T3 T4

P3 P4

Oz O1 O2

0 400 800 1200

Fig. 1. Montage of grand averaged VEPs during an object recognition task. The insets above the montage of grand averaged VEPs show the three 4 4 matrices that were randomly presented at 3 s intervals during the 15 min recording session (i.e., Relevant, Irrelevant, and Standard stimuli). One object in the Relevant matrix of right-pointing arrows is a diamond symbol and one object of the Irrelevant matrix is a variation of a diamond and an arrow. Regardless of matrix, these elements are readily distinguished and ‘‘pop-out” from the other 15 elements of each matrix. The Standard matrix consists of all right-pointing arrows. Male and female subjects were instructed to press a hand-switch when the Relevant stimulus was randomly presented, but to not respond when either the Irrelevant or Standard stimuli were presented. Grand averaged VEPs represent the average of all individual subject VEPs. VEPs elicited by Relevant, Irrelevant, and Standard stimuli were compared at all electrode sites in all subjects. The parietal and occipital electrode sites evinced the most well-deﬁned combination of early (i.e., task-dependent) and late (task-independent) components of the VEP. Negative voltage is plotted downward. F corresponds to frontal, C to central, T to temporal, P to parietal, and O to occipital. Fz, Cz, Pz, and Oz are located at the frontal, central, parietal and occipital midline, respectively.

vant, Irrelevant, and Standard stimuli) as a within-subjects factor. Paired- matrices) for response selection and gender conditions. Values are sample t-tests were performed to compare RTs and VEP measurements expressed as means ± SEM. Statistical signiﬁcance was expressed at the across stimulus presentations (i.e., Relevant, Irrelevant, and Standard P < 0.05 level. 920 S.C. Steﬀensen et al. / Vision Research 48 (2008) 917–925

3. Results A STIMULUS P300 120 RELEVANT 3.1. Late components of the VEP are diﬀerentially P100 IRRELEVANT 80 STANDARD modulated by response selection P200 LP 40

Fig. 1 shows a montage of superimposed grand averaged 0 N200N400 VEPs for all subjects recorded from 12 sites on the stan- -40 N50 dard International 10–20 System in association with the -80 presentation of the Relevant, Irrelevant and Standard stimuli. Inspection of the recordings at each electrode location -120 mean RT 5 µV revealed that the VEP waveforms evoked at the parietal -160 and occipital electrodes contained the most well-defined -200 N100 combination of early and late components in association with the Relevant and Irrelevant stimuli. While the early 0 500 1000 components (i.e., N50, P100, N100, P200 and N200) of TIME (ms) the averaged VEP waveforms were unaffected by type of visual stimulus presented, the late components of the aver- B aged VEP waveforms evinced significant amplitude differ- RELEVANT 20 IRRELEVANT ences across stimulus conditions (Fig. 2A). The amplitude STANDARD V) of the P300 component of the waveform appeared to be µ much greater in association with the Relevant stimulus 15 than with Irrelevant and Standard stimuli. Most impor- tantly, the Irrelevant stimulus elicited a distinct N400/LP 10 * voltage excursion on the VEP waveform that was not pro- * duced by Relevant or Standard stimuli. Fig. 2B summarizes the effects of response selection on the amplitudes of 5 each component of the VEP. The amplitude of the Rele- VEP AMPLITUDE ( vant P300 VEP component was significantly greater than 0 that obtained from Irrelevant (P = 5.36E-06, N50 P100 N100 P200 N200 P300 N400 LP t(2,14) = 5.56) or Standard (P = 2.68E-06, t(2,14) = 5.81) Fig. 2. Differential modulation of late components of the visual evoked stimuli, the Relevant N400 VEP component was not signif- potential by type of stimulus presentation. (A) This graph shows icantly greater than the Irrelevant (P = 0.889, t(2,14) = 0.14) superimposed grand averaged VEPs of combined male and female subjects generated at the Oz site by 54 random presentations of each of or Standard (P = 0.173, t(2,14) = 1.40) stimulus, and the Irrelevant LP VEP component was significantly greater the Relevant, Irrelevant, and Standard matrices within the session. The averaged VEP in this subject consisted of multiple components which than either the Relevant (P = 2.54E-06, t(2,14) = 5.83) or were identified by their respective positions on the waveform, relative to Standard (P = 2.16E-04, t(2,14) = 4.23) stimuli. the time of stimulus presentation (dashed vertical line represents time of presentation of the visual stimulus). We identified 8 distinctive alternat- 3.2. Greater P300 and N400 VEP amplitudes in female vs ing positive/negative peaks on the VEP waveform which occurred at male subjects characteristic latencies from the time of stimulus presentation. We identified early and late peaks of the VEP according to established convention and labeled them N50, P100, N100, P200, N200, P300 and The effects of gender on early and late components of N400, respectively. In addition, we identified one additional landmark on the VEP were studied in the object recognition task. Reac- the VEP waveform that we labeled late positive (LP). While the VEP tion times to the Relevant stimulus were not significantly P300 amplitude elicited by the Irrelevant stimulus was considerably different in males vs females (P = 0.38, F = 0.77; mean smaller than that produced by the Relevant stimulus, the N400/LP (1,29) component produced by the presentation of the Irrelevant stimulus was male RT = 488 ± 11.7 ms vs mean female greater than that produced by Relevant or Standard stimuli. Mean RT RT = 509 ± 20.5 ms). Grand averaged VEP waveforms for detection of the Relevant stimulus is indicated by the black arrows generated in males and females at the Oz site by presenta- on the graph. (B) This graph summarizes the effects of stimulus tion of Relevant and Irrelevant stimuli are shown in Fig. 3. presentation on the amplitude of discrete components of the VEP. The The early components (i.e., N50, P100, N100, P200 and mean values represent measurements taken from each subjects averaged VEP, not from the cumulated averaged VEPs of all subjects (as shown in N200) of the averaged Relevant VEP waveform, as well A). There were no significant differences between Relevant, Irrelevant, as the latencies of all the VEP components, were not signif- and Standard stimuli for any of the early components (i.e., N50, P100, icantly different in males vs females (Table 1; measured N100, P200 or N200) of the cumulated VEP. However, the Relevant within-subject and averaged—see Section 2). However, VEP P300 amplitude was significantly greater than that produced by averaged P300 and N400 VEP Relevant stimulus wave- Irrelevant or Standard stimuli, and the Irrelevant VEP LP amplitude was significantly greater than that produced by the Relevant or Standard forms indicated significant gender differences (Fig. 3A). stimuli. Values are expressed as means ± SEM. Asterisks represent Fig. 3B summarizes the effects of gender on each of the late significance levels P < 0.05. amplitude components of the Relevant VEP. The ampli- S.C. Steffensen et al. / Vision Research 48 (2008) 917–925 921

A RELEVANT C IRRELEVANT STIMULUS STIMULUS MALE 120 MALE FEMALE FEMALE 100 80

40 0 0

-40 -100 -80 5 µV -200 5 µV -120

-160

0 400 800 1200 0 400 800 1200 TIME (ms) TIME (ms)

BD 16 12 MALE MALE ** FEMALE FEMALE 10 ** V) V)

µ 12 µ 8

8 * 6

4 4

AMPLITUDE ( AMPLITUDE ( 2

0 0 P300 N400 LP P300 N400 LP

Fig. 3. The late components of the visual evoked potential are sensitive to gender. (A) This graph shows superimposed grand averaged VEPs generated at the Oz site by presentation of the Relevant stimulus and collapsed by gender. The amplitude of the P300 and N400 components of the Relevant VEP waveform appear greater in females than males. (B) This graph summarizes the effects of gender on the amplitude of the VEP elicited by presentation of the Relevant stimulus. Female P300 and N400 amplitudes were significantly greater than males. (C) This graph shows superimposed averaged VEPs produced by presentation of the Irrelevant stimulus and collapsed by gender. The amplitude of the P300 and N400 components of the Irrelevant VEP appear greater in females than males. (D) This graph summarizes the effects of gender on the amplitude of the VEP elicited by presentation of the Irrelevant stimulus. Only female N400 amplitudes were significantly greater in females than males in association with the Irrelevant stimulus. Values are expressed as means ± SEM. Asterisks and represent significance levels P < 0.05 and 0.001 < P < 0.05, respectively.

Table 1 Summary of the effects of gender on the amplitude and latency of distinct components of the VEP obtained at the Oz site under Relevant and Irrelevant conditions Amplitude (lv) Latency (ms) Relevant Irrelevant Relevant Irrelevant Male Female Male Female Male Female Male Female N50 2.67 ± 0.35 1.90 ± 0.34 1.97 ± 0.39 1.87 ± 0.29 47.73 ± 3.59 45.33 ± 4.01 50.13 ± 2.92 51.47 ± 3.26 P100 7.45 ± 1.04 8.19 ± 0.92 7.74 ± 0.91 8.45 ± 1.31 85.87 ± 2.76 82.40 ± 2.89 85.60 ± 2.40 81.872.40 N100 16.00 ± 1.29 18.92 ± 1.37 16.61 ± 1.53 19.68 ± 1.64 133.07 ± 2.75 128.53 ± 4.53 132.27 ± 2.60 127.73 ± 3.09 P200 13.11 ± 1.23 13.52 ± 1.32 13.96 ± 1.21 14.19 ± 1.41 196.8 ± 3.71 193.33 ± 5.13 206.13 ± 5.18 196.27 ± 2.88 N200 4.86 ± 0.76 5.43 ± 1.07 2.93 ± 0.45 4.92 ± 1.03 244.27 ± 8.60 238.13 ± 8.39 257.60 ± 11.05 239.47 ± 9.05 P300 6.18 ± 0.89 12.32 ± 2.0** 3.66 ± 1.24 6.47 ± 0.86 307.20 ± 7.59 301.60 ± 8.47 311.73 ± 9.78 296.0 ± 5.27 N400 3.43 ± 0.61 6.65 ± 1.31* 1.57 ± 1.3 8.84 ± 1.24** 350.40 ± 5.67 358.67 ± 15.34 374.40 ± 9.98 383.73 ± 6.55 LP 5.58 ± 0.70 7.59 ± 1.14 8.67 ± 2.2 7.73 ± 1.17 428.0 ± 9.14 454.67 ± 14.85 462.13 ± 16.88 468.0 ± 19.69 Values are expressed as means ± SEM. While there were no significant differences in latencies for any VEP component for the Relevant or Irrelevant stimulus there were significant differences between males and females for both P300 and N400 amplitudes. Asterisks and represent significance levels P < 0.05 and 0.001 < P < 0.05, respectively, for comparisons between gender.

tude of the Relevant P300 VEP component was signifi- males (P = 0.03, F(1,29) = 4.95). With regard to the Irrele- cantly greater in females than males (P = 0.009, vant stimulus, the late components of the averaged P300 F(1,29) = 7.89). The amplitude of the Relevant N400 VEP and N400 VEP waveforms indicated significant gender dif- component was also significantly greater in females than ferences (Fig. 3C). Fig. 3D summarizes the effects of gender 922 S.C. Steffensen et al. / Vision Research 48 (2008) 917–925 on each of the late components of the Irrelevant VEP. tem. There was no significant difference in P300 or N400 While differences between males and females for the Irrele- amplitudes between O1 vs O2 or P3 vs P4 in males or vant P300 VEP were not significant, the amplitude of the females for either the Relevant or Irrelevant stimuli Irrelevant N400 VEP component was significantly greater (P > 0.05). Notwithstanding the fact that N400 amplitudes in females than males (P = 0.001, F(1,29) = 14.1). To rule were slightly lower at O1 and O2 and P3 and P4 vs the Oz out any inter-day variability in the amplitude of the late site for both Relevant and Irrelevant stimuli (compare with VEP components we ran all subjects in a second session. Table 1), N400 amplitudes remained significantly higher in While there was considerable inter-subject variability in females than males for the Irrelevant stimulus at both the the late components of the VEP waveform, there was less O1 (P = 0.02, F(1,29) = 6.46) and O2 (P = 0.01, than 5% inter-day variability (e.g., 1st session Relevant F(1,29) = 7.16) sites. male P300 amplitude = 6.18 ± 0.89 lV vs 2nd session Rel- evant male P300 amplitude = 6.4 ± 0.81 lV; 1st session Relevant female P300 amplitude = 12.32 ± 2.0 lV vs 2nd 4. Discussion session Relevant female P300 amplitude = 11.75 ± 2.2 lV). To rule out any specific object stimulus effect (i.e., is there There are two main types of visual object search: paral- something particular about the diamond or the diamond lel and serial (Luck & Hillyard, 1994). Parallel, or ‘‘pre- with lines that might produce gender-specific N400 differ- attentive,” processing occurs when an object contains one ences or eye-movement effects), we switched the objects or more features that are absent from the distractors in (i.e., diamond vs diamond with lines; see Fig. 1) for the the scene, causing an object to ‘‘pop-out” from a back- Relevant and Irrelevant stimulus and ran the same subjects ground of homogeneous distractors (Bravo & Nakayama, through the paradigm. Reaction times did not differ signif- 1992; Dehaene, 1989; Egeth, Jonides, & Wall, 1972; icantly between objects when used as the Relevant stimulus Nakayama & Silverman, 1986; Saarinen, 1997; Theeuwes, (P = 0.3, F = 0.82), nor between males vs females 1993; Treisman & Gelade, 1980; Treisman & Gormican, (1,29) 1988; Verghese & Nakayama, 1994; Wolfe, 1994). Parallel (P = 0.24, F(1,29) = 0.88). The amplitude of the Irrelevant N400 VEP component remained significantly greater in processing is distinguished by a relatively short RT latency when compared to the longer latencies of serial processing females than males (P = 0.018, F(1,29) = 8.74; mean male Irrelevant N400 amplitude = 3.43 ± 0.61 lV vs mean due to distractor stimuli (Duncan & Humphreys, 1989; female Irrelevant N400 amplitude = 6.64 ± 1.3 lV) when Saarinen, 1997; Salyer, 2001; Theeuwes, 1993; Treisman the target objects were switched as the Relevant stimulus. & Gelade, 1980; Wolfe, 1994). In the object recognition task we adopted, the late components of the VEP waveform were differentially altered by the behavioral task. 3.3. Laterality The P300 amplitude of the VEP was enhanced in association with the Relevant stimulus, regardless of the pop-out As P300 and N400 amplitudes at the Oz site were greater object (i.e., similar RT) that was presented as the Relevant in females than males, especially in association with the stimulus. This result was expected, partly based on our pre- Irrelevant stimulus, we evaluated the potential for laterality vious studies (Salyer, 2001), and the fact that the P300 effects that might contribute to the gender differences. amplitude is known to be dependent on the allocation of Table 2 shows the mean amplitudes of the P300 and attentional resources, as well as target salience, or the N400 components of the VEP at the O1, O2, P3, and P4 degree to which an object pops-out from a background electrode locations of the standard International 10–20 sys- of distractor stimuli (Coull, 1998; Katayama & Polich,

Table 2 Summary of the effects of gender on the amplitude and latency of distinct components of the VEP obtained at O1, O2, P3 and P4 sites under Relevant and Irrelevant conditions Amplitude (lV) Relevant Irrelevant Male Female Male Female P300 O1 9.54 ± 1.25 10.56 ± 1.23 7.23 ± 1.1 7.57 ± 0.90 O2 8.82 ± 1.17 9.83 ± 1.37 6.56 ± 0.93 7.53 ± 1.21 P3 9.47 ± 1.19 9.1 ± 1.6 6.27 ± 0.98 5.81 ± 1.2 P4 8.17 ± 1.08 8.96 ± 1.58 5.93 ± 0.82 6.07 ± 1.16 N400 O1 3.08 ± 0.61 4.54 ± 0.62 4.47 ± 0.83 7.69 ± 0.95* O2 3.36 ± 0.55 4.98 ± 0.67 4.32 ± 0.89 8.39 ± 1.22* P3 2.25 ± 0.59 2.5 ± 0.68 3.36 ± 0.65 4.62 ± 0.83 P4 2.12 ± 0.64 2.41 ± 0.68 3.60 ± 0.63 4.53 ± 0.82 Values are expressed as means ± SEM. While there were no significant differences between O1 and O2 or P3 and P4 within gender or across gender, there were significant differences between males and females for N400 amplitudes at these sites. Asterisks represent significance levels P < 0.05, for comparisons between gender. S.C. Steffensen et al. / Vision Research 48 (2008) 917–925 923

1998; Picton, 1992). However, it was most unexpected that to increase with decision accuracy or decision confidence. this visual paradigm would result in the elicitation of an In our study, we used a simple paradigm wherein subjects appreciable N400/LP component produced by the Irrele- rarely responded incorrectly, suggesting that they were vant stimulus. confident of their responses and therefore the VEP showed Although the N400 is believed to be part of the brain’s an LP component. However, it is our opinion that interpre- typical response to stimuli that are potentially meaningful, tations regarding the LP are problematical as it occurs the N400 is often associated with semantic or complex pic- around the time that subjects respond to the Relevant stim- ture processing under non-congruent conditions (Hagoort, ulus or during the time they hesitate to not respond to the 2003; Kim et al., 2001; Koyama et al., 1992; Kutas & Irrelevant stimulus. Federmeier, 2000; Kutas & Iragui, 1998; Wicha, Bates, In our visual processing paradigm, P300 amplitudes Moreno, & Kutas, 2003). Simply, the N400 appears to be associated with the Relevant and Irrelevant stimuli were inversely related to the expectancy of a word or stimulus. greater in females than males, supporting previous studies The N400 component has been shown to be smaller for demonstrating that event-related potentials (ERPs) are sen- primed words than for unprimed words (Bentin et al., sitive to gender (Chu, 1987; Hoffman & Polich, 1999). 1985; Deacon et al., 1995; Kim et al., 2001), and Kutas However, they were not uniquely sensitive, as N400 ampli- and Iragui (1998) found that this effect happens regardless tudes were also greater in females than males. Indeed, this of age. Another method typically used to show that the may explain why a prominent N400 was not obvious in N400 is larger for unexpected stimuli involves using sen- averaged VEPs from combined male and female subjects, tence completion or matched (related or unrelated) pairs except for perhaps at the Oz site. In order to test that there of words. The N400 amplitude is greater for words that was not a specific effect of the target we chose as the Rele- are not expected to end the sentence or for unrelated pairs vant stimulus, and to rule out possible eye-movement of words. Similarly, the N400 has been shown (although effects, we switched the targets for the Relevant and Irrele- not strictly) to be reduced by factors that increase the vant stimuli. Under these conditions, the N400 continued object’s predictability within a given context (Koyama to be exclusively sensitive to gender. The relationship et al., 1992; Kutas & Federmeier, 2000; Kutas & Iragui, between gender and the P300 has been controversial as 1998; Tartter, Gomes, Dubrovsky, Molholm, & Stewart, some studies see no gender bias or larger amplitudes in 2002). In our study, the visual targets were all equally pre- males. For example, Oliver-Rodriguez, Guan, and John- dictable and parallel-processed, as the subjects were shown ston (1999) looked at facial attractiveness and the emo- the three visual stimuli over multiple iterations, both before tional component and found that P300 amplitudes were and during the recording session—yet there were differen- greater in male participants. Although, in separate studies, tial N400 effects. One possible explanation is that the females were found to have larger P300 components when N400 represents the recognition that the Irrelevant is not evaluating emotion presented in faces (Morita, Morita, the Relevant or the correct target for response, and thus Yamamoto, Waseda, & Maeda, 2001; Yamamoto et al., a decision must be made not to respond to the Irrelevant 2000). Considering these contradictory findings, research- stimulus—thus creating the N400 waveform. These find- ers have looked for factors that could help explain the ings suggest that the N400 may be related more to rele- dichotomy between gender VEP components. One hypoth- vancy than predictability, and that an N400 component esis that has been proposed explains that head size and can be elicited to non-congruent stimuli in a simple object geometry may account for more of the difference between recognition task, somewhat similar to the semantic process- gender VEPs than actual biological and physiological dif- ing/N400 theory of some authors. In this study, ‘non-con- ferences (Guthkelch, Bursick, & Sclabassi, 1987). Other gruent’ could mean that the Irrelevant stimuli are not possible explanations for the gender difference are seasonal congruent with, or relevant to, what the subject was asked variation (Deldin, Duncan, & Miller, 1994) and emotion to respond to, therefore producing a larger N400 (Morita et al., 2001; Yamamoto et al., 2000). Finally, it component. has recently been proposed that hemispheric asymmetry The LP (possibly analogous to the LPC or P600 dis- might give rise to greater P300 amplitudes in females than cussed by other authors) has been seen in a variety of stud- males (Roalf, Lowery, & Turetsky, 2006). If the brains of ies, usually involving linguistic processing in either visual men are typically more lateralized than those of women or auditory paradigms. For example, Misra and Holcomb (Kolb & Wilshaw, 1996), women should evince more sym- (2003) found that the LP was only present for repeated, metrical processing of visual stimuli than men. In our unmasked words, and Stuss, Picton, Cerri, Leech, and Ste- visual recognition task, we did not observe gender differ- them (1992) reported that a LP followed the N400 only ences in hemispheric asymmetry or lateralization that when the subject responded correctly. Moreover, it has might contribute to the differences between males and been shown that repeated words in incidental learning females. Mainly, neither P300 nor N400 amplitudes were and correctly recognized old words in recognition tasks eli- significantly different across lateral occipital or parietal cit larger late positive components than new words (Paller, locations in association with Relevant or Irrelevant stimuli. Kutas, & Mayes, 1987; Senkfor & Van Petten, 1998). In a Similar to Oz, N400 VEP amplitudes were significantly study by Finnigan et al. (2002), the LP was demonstrated greater in females than males at lateral occipital and pari- 924 S.C. Steffensen et al. / Vision Research 48 (2008) 917–925 etal locations. Thus, our studies indicate that the N400/LP Egeth, H., Jonides, J., & Wall, S. (1972). Parallel processing of multi- component of the VEP appears to be a more selective mea- element displays. Cognitive Psychology, 3, 674–698. sure of gender than the P300. P300 amplitude is generally Ehlers, C. L., Wall, T. L., Garcia-Andrade, C., & Phillips, E. (2001). Auditory P3 findings in mission Indian youth. Journal of Studies on associated with stimulus probability, which was the same Alcohol, 62(5), 562–570. across gender in this study, by task salience, which was also Finnigan, S., Humphreys, M. S., Dennis, S., & Geffen, G. (2002). ERP the same across gender in this study, and by attentional ‘old/new’ effects: Memory strength and decisional factor(s). Neuro- resources. Therefore, these findings suggest that women psychologia, 40(13), 2288–2304. allocated greater attentional resources towards, and/or Ganis, G., & Kutas, M. (2003). An electrophysiological study of scene effects on object identification. Brain Research. Cognitive Brain attributed greater task salience to the distracting (i.e., Irrel- Research, 16(2), 123–144. evant) stimuli than men. These findings support the pre- Gunter, T. C., Friederici, A. D., & Schriefers, H. (2000). Syntactic gender vailing hypothesis that the P300 is sensitive to gender; and semantic expectancy: ERPs reveal early autonomy and late however, the N400/LP component of the VEP appears to interaction. Journal of Cognitive Neuroscience, 12(4), 556–568. be more selective to gender. This obtains despite the lack Guthkelch, A. N., Bursick, D., & Sclabassi, R. J. (1987). The relationship of the latency of the visual P100 wave to gender and head size. of differences in the early components of the VEP and in Electroencephalographic Clinical Neurophysiology, 68(3), 219–222. RT between males and females, suggesting that sensory Hagoort, P. (2003). Interplay between syntax and semantics during processing and motor performance do not contribute to sentence comprehension: ERP effects of combining syntactic and the differences. While a few behavioral studies have demon- semantic violations. Journal of Cognitive Neuroscience, 15(6), 883–899. strated gender differences for the processing of distracting Hagoort, P., & Brown, C. M. (1999). Gender electrified: ERP evidence on the syntactic nature of gender processing. Journal of Psycholinguistic stimuli (Lesch & Hancock, 2004), this study appears to Research, 28(6), 715–728. be the first to demonstrate a physiological correlate for Hillyard, S. A., Mangun, G. R., Luck, S. J., & Heinze, H. J. (1990). gender differences in the processing of distracting stimuli. Electrophysiology of visual attention. In E. R. John, T. Harmony, L. The gender differences appear to be in the later stages of L. Prichep, & P. A. Valdes-Sosa (Eds.), Machinery of the mind visual processing, and do not appear to be attributed to (pp. 186–205). Boston: Birkhauser. Hoffman, L. D., & Polich, J. (1999). P300, handedness, and corpus callosal hemispheric asymmetry, motor performance, type of visual size: Gender, modality, and task. International Journal of Psychophys- stimulus, or expectancy. iology, 31(2), 163–174. Isreal, J. B., Wickens, C. D., & Donchin, E. (1980). The dynamics of P300 during dual-task performance. Progress in Brain Research, 54, References 416–421. Jemel, B., George, N., Olivares, E., Fiori, N., & Renault, B. (1999). Event- Acosta, V. W., & Nasman, V. T. (1992). Effect of task decision on P300. related potentials to structural familiar face incongruity processing. International Journal of Psychophysiology, 13(1), 37–44. Psychophysiology, 36(4), 437–452. Barrett, S. E., Rugg, M. D., & Perrett, D. I. (1988). Event-related Katayama, J., & Polich, J. (1998). Stimulus context determines P3a and potentials and the matching of familiar and unfamiliar faces. Neuro- P3b. Psychophysiology, 35(1), 23–33. psychologia, 26(1), 105–117. Katayama, J., & Polich, J. (1999). Auditory and visual P300 topography Bennington, J. Y., & Polich, J. (1999). Comparison of P300 from passive from a 3 stimulus paradigm. Clinical Neurophysiology, 110(3), 463–468. and active tasks for auditory and visual stimuli. International Journal Kim, M., Kim, J., & Kwon, J. S. (2001). The effect of immediate and of Psychophysiology, 34(2), 171–177. delayed word repetition on event-related potential in a continuous Bentin, S., McCarthy, G., & Wood, C. C. (1985). Event-related potentials, recognition task. Brain Research. Cognitive Brain Research, 11(3), lexical decision and semantic priming. Electroencephalographic Clinical 387–396. Neurophysiology, 60(4), 343–355. Kolb, B., & Wilshaw, I. Q. (1996). Fundamentals of human neurophys- Bravo, M. J., & Nakayama, K. (1992). The role of attention in different iology. New York, NY: W.H. Freeman and Company. visual-search tasks. Perception & Psychophysics, 51(5), 465–472. Koyama, S., Nageishi, Y., & Shimokochi, M. (1992). Effects of semantic Chu, N. S. (1987). Pattern-reversal visual evoked potentials: Latency context and event-related potentials: N400 correlates with inhibition changes with gender and age. Clinical Electroencephalography, 18(3), effect. Brain and Language, 43(4), 668–681. 159–162. Kramer, A. F., & Strayer, D. L. (1988). Assessing the development of Conill, J. (1998). P300 potentials evoked by visual stimulation. Reviews in automatic processing: An application of dual-task and event-related Neurology, 26(151), 448–451. brain potential methodologies. Biological Psychology, 26(1–3), Coull, J. T. (1998). Neural correlates of attention and arousal: Insights 231–267. from electrophysiology, functional neuroimaging and psychopharma- Kutas, M., & Federmeier, K. D. (2000). Electrophysiology reveals cology. Progress in Neurobiology, 55(4), 343–361. semantic memory use in language comprehension. Trends in Cognitive Deacon, D., Mehta, A., Tinsley, C., & Nousak, J. M. (1995). Variation in Sciences, 4(12), 463–470. the latencies and amplitudes of N400 and NA as a function of semantic Kutas, M., & Iragui, V. (1998). The N400 in a semantic categorization priming. Psychophysiology, 32(6), 560–570. task across 6 decades. Electroencephalographic Clinical Neurophysiol- Dehaene, S. (1989). Discriminability and dimensionality effects in visual ogy, 108(5), 456–471. search for featural conjunctions: A functional pop-out. Perception & Lesch, M. F., & Hancock, P. A. (2004). Driving performance during Psychophysics, 46(1), 72–80. concurrent cell-phone use: Are drivers aware of their performance Deldin, P. J., Duncan, C. C., & Miller, G. A. (1994). Season, gender, and decrements? Accident; Analysis and Prevention, 36(3), 471–480. P300. Biological Psychology, 39(1), 15–28. Luck, S. J., & Hillyard, S. A. (1994). Electrophysiological correlates of Duncan-Johnson, C. C., & Donchin, E. (1977). On quantifying surprise: feature analysis during visual search. Psychophysiology, 31(3), The variation of event-related potentials with subjective probability. 291–308. Psychophysiology, 14(5), 456–467. Mertens, R., & Polich, J. (1997). P300 from a single-stimulus paradigm: Duncan, J., & Humphreys, G. W. (1989). Visual search and stimulus Passive versus active tasks and stimulus modality. Electroencephalo- similarity. Psychological Review, 96(3), 433–458. graphic Clinical Neurophysiology, 104(6), 488–497. S.C. Steffensen et al. / Vision Research 48 (2008) 917–925 925

Misra, M., & Holcomb, P. J. (2003). Event-related potential indices of Saarinen, J. (1997). The speed of visual search in the absence of sensory masked repetition priming. Psychophysiology, 40(1), 115–130. effects. Perception, 26(1), 85–91. Mitchell, K. W., Howe, J. W., & Spencer, S. R. (1987). Visual evoked Salyer, D. L. (2001). Behavioral (reaction time) and electrophysiological potentials in the older population: Age and gender effects. Clinical (visual evoked potential) evidence for a parallel-to-serial continuum of Physics and Physiological Measurement, 8(4), 317–324. visual performance processing: Study I and study II. Psychology. Provo: Morita, Y., Morita, K., Yamamoto, M., Waseda, Y., & Maeda, H. (2001). Brigham Young University, pp. 99. Effects of facial affect recognition on the auditory P300 in healthy Sangal, B., & Sangal, J. M. (1996). Topography of auditory and visual subjects. Neuroscience Research, 41(1), 89–95. P300 in normal adults. Clinical Electroencephalography, 27(3), Nakayama, K., & Silverman, G. H. (1986). Serial and parallel processing 145–150. of visual feature conjunctions. Nature, 320(6059), 264–265. Senkfor, A. J., & Van Petten, C. (1998). Who said what? An event-related Naumann, E., Huber, C., Maier, S., Plihal, W., Wustmans, A., Diedrich, potential investigation of source and item memory. Journal of O., & Bartussek, D. (1992). The scalp topography of P300 in the visual Experimental Psychology. Learning, Memory, and Cognition, 24(4), and auditory modalities: A comparison of three normalization 1005–1025. methods and the control of statistical type II error. Electroencepha- Stuss, D. T., Picton, T. W., Cerri, A. M., Leech, E. E., & Stethem, L. L. lographic Clinical Neurophysiology, 83(4), 254–264. (1992). Perceptual closure and object identification: Electrophysiolog- Olivares, E. I., Iglesias, J., & Rodriguez-Holguin, S. (2003). Long-latency ical responses to incomplete pictures. Brain and Cognition, 19(2), ERPs and recognition of facial identity. Journal of Cognitive Neuro- 253–266. science, 15(1), 136–151. Sutton, S., Braren, M., Zubin, J., & John, E. R. (1965). Evoked-potential Oliver-Rodriguez, J. C., Guan, Z., & Johnston, V. S. (1999). Gender correlates of stimulus uncertainty. Science, 150(700), 1187–1188. differences in late positive components evoked by human faces. Tartter, V. C., Gomes, H., Dubrovsky, B., Molholm, S., & Stewart, R. V. Psychophysiology, 36(2), 176–185. (2002). Novel metaphors appear anomalous at least momentarily: Paller, K. A., Kutas, M., & Mayes, A. R. (1987). Neural correlates of Evidence from N400. Brain and Language, 80(3), 488–509. encoding in an incidental learning paradigm. Electroencephalographic Theeuwes, J. (1993). Visual selective attention: A theoretical analysis. Acta Clinical Neurophysiology, 67(4), 360–371. Psychologica, 83(2), 93–154. Picton, T. W. (1992). The P300 wave of the human event-related potential. Treisman, A., & Gormican, S. (1988). Feature analysis in early vision: Journal of Clinical Neurophysiology, 9(4), 456–479. Evidence from search asymmetries. Psychological Review, 95(1), 15–48. Picton, T. W., Bentin, S., Berg, P., Donchin, E., Hillyard, S. A., Johnson, Treisman, A. M., & Gelade, G. (1980). A feature-integration theory of R., Jr., Miller, G. A., Ritter, W., Ruchkin, D. S., Rugg, M. D., & attention. Cognitive Psychology, 12(1), 97–136. Taylor, M. J. (2000). Guidelines for using human event-related Verghese, P., & Nakayama, K. (1994). Stimulus discriminability in visual potentials to study cognition: Recording standards and publication search. Vision Research, 34(18), 2453–2467. criteria. Psychophysiology, 37(2), 127–152. West, W. C., & Holcomb, P. J. (2002). Event-related potentials during Polich, J. (1986). Normal variation of P300 from auditory stimuli. discourse-level semantic integration of complex pictures. Brain Electroencephalographic Clinical Neurophysiology, 65(3), 236–240. Research. Cognitive Brain Research, 13(3), 363–375. Polich, J. (1998). P300 clinical utility and control of variability. Journal of Wicha, N. Y., Bates, E. A., Moreno, E. M., & Kutas, M. (2003). Potato not Clinical Neurophysiology, 15(1), 14–33. Pope: Human brain potentials to gender expectation and agreement in Polich, J., & Bloom, F. (1988). Event-related brain potentials in Spanish spoken sentences. Neuroscience Letters, 346(3), 165–168. individuals at high and low risk for developing alcoholism: Failure Wickens, C., Kramer, A., Vanasse, L., & Donchin, E. (1983). Performance to replicate. Alcoholism: Clinical and Experimental Research, 12(3), of concurrent tasks: A psychophysiological analysis of the reciprocity 368–373. of information-processing resources. Science, 221(4615), 1080–1082. Pritchard, W. S. (1981). Psychophysiology of P300. Psychology Bulletin, Wolfe, J. M. (1994). Visual search in continuous, naturalistic stimuli. 89(3), 506–540. Vision Research, 34(9), 1187–1195. Roalf, D., Lowery, N., & Turetsky, B. I. (2006). Behavioral and Yamamoto, M., Morita, K., Tomita, Y., Tsuji, K., Kawamura, K., & physiological findings of gender differences in global–local visual Maeda, H. (2000). Effect of facial affect stimuli on auditory and visual processing. Brain and Cognition, 60(1), 32–42. P300 in healthy subjects. The Kurume Medical Journal, 47(4), 285–290.