Comparison of the EEG Alpha Power of Males and Females During an Attention Task

power spectrum analysis was performed on EEG's from 22 male and 22 female adult subjects under three conditions: 1. Resting; 2. During the first two minutes of the performance of an audit~ry continuous performance task (CPT); 3. During minutes eight through ten of the performance of an auditory CPT. Studies previously cited in the literature have reported finding electrophysiological gender differences using cognitively complex tasks (e.g. visual and spatial). The successful completion of such complex tasks, however, in no way insures the use of a single cognitive strategy by all subjects. In fact, many different cognitive strategies may conceivably enable a subject to _,/"' successfully complete a task with complex cognitive dimensions. In the present study a CPT was chosen so as to minimize strategy variation. A mixed ANOVA was performed on the absolute alpha power scores from eight bipolar recording sites. Males and females exhibited comparable lateralization patterns of brain activation during the resting condition and both time periods during the CPT. There was a significant decrease in absolute alpha power in the right temporal-occipital leads and the left temporal-occipital leads for both time periods during the CPT. These data provide evidence that previous observations of gender

The successful completion of such complex tasks, however, in no way insures the use of a single cognitive strategy by all subjects.
In fact, many different cognitive strategies may conceivably enable a subject to _,/"' successfully complete a task with complex cognitive dimensions.
In the present study a CPT was chosen so as to minimize strategy variation. A mixed ANOVA was performed on the absolute alpha power scores from eight bipolar recording sites. Males and females exhibited comparable lateralization patterns of brain activation during the resting condition and both time periods during the CPT. There was a significant decrease in absolute alpha power in the right temporal-occipital leads and the left temporal-occipital leads for both time periods during the CPT. These data provide evidence that previous observations of gender ii ~ differences during the performance of complex tasks (verbal and spatial tasks) reflect distinct cognitive strategies rather than hard-wiring brain differences.
rn addition, the data do not support the hypothesis that the right frontal lobe mediates the attention mechanism responsible for maintenance of vigilance.
iii Abstract Table   List List of of of Contents Tables   Figures   Literature Review   Table of  Lab, is to establish a set of encephalographic (EEG) norms for "normal", right handed individuals, ages 18 to 25 years, while they perform a simple cognitive tas k .
These norms can then be used to assist physicians and clinicians in their diagnosis of patients with organically based pathologies and may provide a baseline to study the effects of variations in cognitive processes on the EEG. However, the construction of norms is difficult at best and requires that the sample from which the norms were derived be as large and as homogeneous as possible.
Similarly, because the problem of increasing a sample's size is a function of time, the task of making certain that the sample is as homogeneous as possible with respect to all attribute factors (gender, age, and handedness) commands a high priority. Indeed, past research has indicated that the age (Duffy, Albert, McAnulty, & Garvey, 1984, p. 430) and the handedness (Galin, Ornstein, Herron & Johnstone, 1982, p.45) of an individual affect that individual's brain pnysiology and, perhaps, neuroanatomic wiring. Therefore, norms need to be established for each specific sub-populations (i.e. 18 to 25 year old, right handers). The effects of attribute factors other than age and handedness, such as gender, are less understood.
Gender, in particular, may have significant influences on an individual's brain structure and physiology . There has been an accumulation of research that has lent support to the theory that the human female brain is more functionally and anatomically symmetric than the human male brain. Consequently, the purpose of this experiment is to investigate whether males and females exhibit dissimilar lateralization of EEG activity while performing a simple attention task.
Because the attention task used in this present experiment will be the same task used in the lab's future construction of EEG norms, the results of this experiment will contribute as well to beginning the process of establishing reliable norms.
Sex differences in EEG and in task performance (a behavioral measure of subject vigilance) were analyzed.
Because the detection of brain activation was paramount to this investigation, and because "alpha suppression is greater over the active cerebral hemisphere" (Marquis, Glass & Corlett, 1984, p. 205), the level of alpha power (brain wave ranging from 7.5 to 12.5 hz) was utilized as an indicator of hemispheric activation/innactivation.

Support £or Brain I Gender Hypothesis
Past neuropsychological research has indicated that an~tomical and functional asymmetry differences may exist between the male and the female cerebral hemispheres. Anatomical, clinical, and normative studies have suggested that the female cortex is both functionally and anatomically more symmetrical than the male cortex. Hence, it has been postulated that "men and women differ in the degree to which the cerebral hemispheres are specialized for processing different types of information" (Berfield, Ray & Newcombe, 1986, p. 731).

Anatomical Studies:
Recent studies of human and primate brains, using methods ranging from post-mortem examination to magnetic resonance imaging (MRI) and computerized axial tomography (CAT), have provided conflicting results regarding anatomical, gender brain differences. Lacoste-Utamsing and Holloway (1982), after measuring the posterior fifth of nine male and five female corpus callossa post-mortem, reported finding that the splenium was much larger in the female brain than in the male brain (p.1431).
In light of the fact that the corpus callosum allows the two hemispheres to communicate, and because females exhibit relatively larger corpus callossa than males, there may exist a greater capacity, and possible need, for the two hemispheres of the female brain to interact. However, Witelson (1985), in an attempt to directly replicate In addition to the corpus callosum, researchers have also been interested in possible cortical differences between males and females. The perceived role of the cerebral hemispheres in human cognition, coupled with the popular belief that the genders were singularly proficient at performing various cognitive tasks (e.g. math and language), fostered this scientific interest. Consequently, the cerebral hemispheres were targeted as areas of the brain that would provide the greatest differentiation of males and females. Wada, Clarke and Hamm (1975), after the post-mortem examination of 100 temporal planums (planum length), found no significant gender differences in planum size.
(However, they concluded that there was a trend for the left planum in the male to be larger than the left planum in the female (i.e. p > .1) (p.243)). In addition, many studies that have examined the length and width of male and female hemispheres by CAT have reported finding no significant gender differences (Chui & oamasio, 1980;Koff, Naeser, Pieniadz, Foundas & Levine, 1986 ;Yeo, Turkheimer, Raz & Bigler, 1987).
on the other hand, Bear, Schiff, Saver, Greenberg and Freeman (1986), also using the CAT technique, examined the cerebral hemispheres of 66 subjects and reported finding the male brain to be more asymmetrical with enhanced right-frontal and left-occipital predominance (p. 602).
Lacking consensus, the results of these anatomical investigations serve to bolster the need for more quality, anatomical research. As a result, no amount of confidence can be had in either of the two hypotheses (i.e. difference vs. no difference).

Clinical Studies:
Clinical studies, examining lesion sequelae, have resulted in an even more inconsistent picture of gender brain differences.
Lansdell, in the early 1960s, conducted a series of experiments that has lent support to the theory that males were less functionally symmetric than females. As an example, males with left temporal lobe ablation were found to exhibit greater impairment in proverb interpretation (Lansdell, 1961) ·, in word association (Lansdell, 1973), and in performance on the verbal scale of the Wechsler Adult Intelligence Scale (WAIS) (Lansdell, 1968b) than did females. However, Lansdell did not find any significant gender differences in the .
vocabulary performance of subjects with left temporal lobe ablations (Lansdell, 1968a) . Lansdell also reported that males with right temporal lobe ablations exhibited decrements in performance on spatial tasks such as the Graves Design Judgment Test (Lansdell, 1962) and the nonverbal (i.e. performance) subtest of the WAIS (Lansdell, 1968b), whereas females did not. Also reported was the fact that males and females performed comparably on the Mooney's Closure Faces Test (Lansdell, 1968a).
From the late 1970s to the late 1980s researchers had begun to take seriously the theory that the male brain and the female brain were functionally different.  Thompson, 1976;Friedland & Kershner, 1986;Inglis & Lawson, 1981;McGlone, 1984;Sundet, 1986;Yeo, Turkheimer & Bigler, 1984). Lewis and Kamptner (1987), after examining the performance of 66 unilaterally brain  (1979), in a similar experiment, examined the alpha asymmetry of 11 right handed subjects (6 males and 5 females) while they performed six separate cognitive tasks (resting, vigilance task, mental letter task, block design task, embedded figures task, rod-frame task) and concluded that females were not as consistently lateralized as were males (p.222).
Similar results were reported in a study conducted by Trotman and Hammond (1979). After examining the alpha asymmetry of 10 right handed subjects (5 male and 5 female) while performing three verbal and three spatial tasks, Trotman and Hammond reported that only males exhibited task-related alpha asymmetries and concluded that such results suggested "a stricter hemispheric lateralization of underlying function in the male brain than in the female brain" (p.430  Earle & Pikus, 1982). However, other researchers have reported results that do not support the gender-brain difference hypothesis. Davidson et al. (1976, p.130), in a builtin replication of their previously cited EEG study, reported finding no significant differences between male and female alpha asymmetry scores (n=20). Similarly, Galin, Ornstein, Herron and Johnstone (1982), after examining the alpha asymmetries of 90 subjects (45 male and 45 female) while they performed left and right hemisphere tasks, also reported finding no significant alpha asymmetry differences between males and females (p.49).
Hence, the normative (EEG)· literature has suffered in much the same way as has the anatomical and the clinical literature. The effects of conflicting results has rendered any clear statement of relationship between gender and the brain (i.e. brain physiology, brain anatomy} virtually impossible. Nevertheless, just as the gender-brain difference theory has not enjoyed consistent empirical support, its alternative hypothesis has not fared much better.

Attention and Brain Function
In addition to examining the impact that the spatial and the verbal components of tasks have on the male and female brain, a number of researchers have also . studies of attention have also been conducted with the assistance of EEG. For instance, Ray and Cole (1985), after examining the frontal and the parietal alpha power of 18 right handed subjects (9 male and 9 female) while they performed two separate visual attention tasks (i.e. a rejecti.on task and an intake task), reported that the left parietal site exhibited significantly less alpha power activity than did the right parietal site during both attention tasks (p.751).
Gender was initially included as a factor within the design of this experiment, however, the effects of gender were not reported. Heilman and Van Den Abell In summation, the results of numerous clinical, behavioral, ERP, CBF, PET and EEG studies have served only to sketch a picture of the attending, human brain.
Nevertheless, it is apparent from the results of these studies that the frontal and the parietal lobes, as well as the right hemisphere in general, play an important role in the cognitive process of attention both for males and for females.

Hypothesis and Prediction
In the present investigation the effects of gender and attention on cerebral activation (i.e. alpha suppression) was being examined. It was hypothesized that: 1) males and females have differently organized brains; the different functions are shared more by the hemispheres in the female; 2) 3)

4)
lb alpha power is inversely related to brain activation, and thus, level of attention; there is an attention mechanism, specific to vigilance, in the right, frontal area; the longer a subject is forced to be attentive, the greater the loss of attention becomes.
corresponding predictions: there will be a task x hemisphere x gender interaction, with only males exhibiting significant differences in alpha power between the left and right hemispheres; for all subjects: alpha power resting > alpha power intake 2 > alpha power intake l; all subjects alpha power at the right, frontal leads will be significantly less than at the left, frontal leads during both intake 1 and intake 2; all subjects will perform significantly better (a greater percent correct) during intake 1 than in intake 2.

Apparatus
-An Axon Systems data acquisition system was used to amplify, digitize, and measure subject's brain waves.
The recorded measures were subjected to a spectral analysis by Fast Fourier Analysis on an IBM compatible, AT style computer. The CPT task was presented by tape recorder through a speaker directly in front of and above the subject. A paper response recorder (model P2C), manufactured by Ralph Ger-brands Co., was used to record targets and subject responses on response paper.

Procedure
Participants sat in a comfortable lounge chair while electrodes were placed, according to the International 10-20 system (Jasper, 1958) Participants were given the following instructions: For the first phase of this experiment we ask that you sit quietly with your eyes closed; your arms in your lap and your legs extended outward. After resting for a period of approximately three minutes we will ask that you perform a task which will constitute phase two: The task will require that you listen to a tape recording in which the letters of the alphabet are spoken randomly, one right after the other. When you hear the same letter spoken twice, (e.g. a b d d k} consider it a target (i.e. d d) and press the button (the subject holds a button in his/her right hand) . Please keep your eyes closed throughout the procedure (both phases) and try not to move in the seat. We will verbally signal you when we are about to begin phase two. (Ax (Bx S)).

Resuits EEG Data
The EEG data were first evaluated by a four way ANOVA with GENDER, TASK, HEMISPHERE, and REGION as factors. Mean alpha power and standard deviations for males and females at each of the eight bipolar sites during resting, intake 1, and intake 2 can be seen in Table 1.   = 4.18, p < .OS, and the right, F(l,86) = 9.7, p < .OS, hemispheres while resting than while performing the attention tasks, with no significant differences between intake 1 and intake 2.
Prediction #3: Simple interaction effects tests at the frontal region failed to show any significant differences between the cerebral hemispheres during resting, intake 1, or intake 2. Table 3 shows the mean performance (i.e. % correct) and standard deviations for males and females during intake 1 and intake 2. The performance data were analyzed with a two-way, mixed ANOVA with Gender and Task as factors.

performance Data
Tab l e 3 Performance ( % correct) means and standard deviations for males and females during intake l and intake 2

~ICTION ONE
The results of this study do not support the hypothesis that males and females exhibit dissimilar activation of their cerebral hemispheres while performing a vigilance task. These results are consistent with those of Shepherd (1982). Nevertheless, before any conclusions can be put forth, other possible explanations for the observed r.esul t must be systematically explored.
One possible explanation may be that there are actual structural brain differences (Trotman & Hammond, 1979) and/or processing strategy differences between males and females, but that the measure used in this present experiment was too insensitive to detect these differences. Despite the plausibility of this explanation, previous researchers have reported finding significantly different patterns of activation between males and females using EEG (Davidson et al., 1976;Earle & Pikus, 1982;Glass et al., 1984;Ray et al., 1976;Trotman & Hammond, 1979;Wogan et al., 1979).
Hence, this explanation appears quite unlikely.
Another explanation might be that the measures were taken from the wrong locations -0n the head (i.e. regions of the cortex that are not gender specific) . Indeed, this explanation might be an appropriate one if it were not for the fact that of those EEG studies that did report finding significant gender differences (Davidson et al., 1976;Earle & Pikus, 1982;Glass et al., 1984;RaY et al., 1976;Trotman & Hammond, 1979;Wogan et al., l979), only two included head locations not represented in the present study (Earle & Pikus, 1982;Glass et al., 1984). Similarly, of the EEG studies that had reported finding no significant gender differences, all included head locations not included in the present study (Davidson et al., 1976;Galin et al., 1982;Harter et al., 1990;Shepherd, 1982). Consequently, this explanation appears insufficien·t also. However, of those studies that did report finding gender differences, five of six studies used sample sizes either at or below 9 (i.e. nine males and eight females) (Davidson et al., 1976;Earle & Pikus, 1982;Ray et al., 1976;Trotman & Hammond, 1979;Wogan et al., 1979). As a result, either the error variances of these studies were extremely small, or their between groups effect sizes were extremely large.
Unfortunately, such statistics were not made available by the authors.
A fourth, more plausible explanation might be that males and females possess similar brain structures, but that they utilize different cognitive strategies when faced with an elaborate task.
In previous studies, subjects showed EEG differences only when they were engaged in complex spatial or verbal tasks. In those studies alpha asymmetries were greater for males than for females. However, when tasks used are complex they conceivably contain so many different cognitive components that any one strategy, from an array of many, may enable a subject to satisfactorily complete a given task.
In the present investigation a continuous performance task (vigilance tas.k) was chosen in order to control for this source of variability. It is being assumed that there is a restricted range of cognitive strategies that can be employed during an accurate performance of this task. A previous study (Shepherd, 1982) in which EEG was observed during a continuous performance task also failed to find significant gender differences, though that study only involved two bilateral electrode sites (Ol-P3 and 02-P4).
If we were to assume that any hard-wiring difference between the male and the female brain would result in different patterns of activation for males and females, even when subjects performed a fundamental cognitive task such as a continuous performance task, it is reasonable to assume that previously found gender differences may be due to processing strategy differences only. Perhaps males and females possess the same hard wiring, but they utilize different, genderstereotyped processing strategies while performing a complex task (verbal and spatial tasks), but not when they are required to perform a simpler cognitive task (vigilance task) . This difference in processing strategy might very well produce a corresponding difference in measured EEG. In fact, Wogan et al. (1979) suggested that subjective reports of the strategies used by subjects might help to clarify the relationship between EEG and behavior, especially when subjects perform some of the more cognitively complex verbal and spatial tasks (p.223).
To conclude, these results do not lend support to the hypothesis that previous observations of EEG gender differences during task performance are related to differences in brain organization. The alternative hypothesis which states that observed gender differences are related to differences in the processing strategies of males and females appears to enjoy greater support.

PREDICTION TWO
The results of this study support the hypothesis that subjects exhibit an overall decrease in alpha power during the continuous performance task, relative to the resting condition. However, this phenomenon was limited to the left and right temporal-occipital regions. The Prediction, which went a step further by predicting significant alpha-power differences between intakes 1 and 2, was not borne out by these results . These results are consistent with the previously documented and conventionally accepted relationship between mental effort and alpha power suppression (Marquis et al., 1984;Pollen & Trachtenberg, 1972). Pollen and Trachtenberg (1972) reported that when subjects perform progressively more difficult mental tasks that there was a corresponding_ decrease of alpha power.
In the present investigation, alpha power in the left and the right temporal-occipital regions were the greatest during the resting condition and the lowest during the continuous performance task (i. provide a intelligible understanding of the relationship between attention, the brain, and performance.

PREDICTION THREE
The results of the present experiment do not support the hypothesis that the right, frontal region contains an attention mechanism specific to the maintenance of vigilance. There were no significant differences in alpha power between the left and the right frontal regions during the continuous performance task. These results are consistent with those of Heilman and Van Den Abell (1980) and Ray and Cole (1985) (EEG studies).

In the experiment conducted by Heilman and Van Den
Abell, subjects performed a visual attention task which required them to signal the presence of a target. In the Ray and Cole study, subjects performed several intake and rejection tasks possessing both verbal and spatial components. Though vastly different tasks were used by the two studies, no significant alpha power differences between the frontal regions were reported.
The remaining EEG studies failed to sample EEG from the frontal leads.
The two remaining brain imaging studies (CBF and PET), however, did implicate the right, frontal region in attention (Haier et al., 1988;Roland, 1982 is actually the summation of electrical activity throughout the brain, this explanation seems quite reasonable. Regardless of this explanation's plausibility, more PET and CBF research, using larger sample sizes, will be required to conclude that EEG is too insensitive for the invest~gation of the attention phenomenon.

PREDICTION FOUR
The results of the present experiment lend support to the hypothesis that the longer a subject is forced to be attentive, the greater the loss of attention.
In the present study, subject performance was significantly better during intake 1 than it was in intake 2. The primary importance of this finding is simply to demonstrate that the task manipulation did work. Hence, the attention level of subjects, as operationally defined in the present experiment, did decline.
However, there were no corresponding significant alpha power differences in any of the eight recording regions.