Differences in induced gamma and upper alpha oscillations in the human brain related to...

www.elsevier.com/locate/ijpsycho

International Journal of Psychoph

Differences in induced gamma and upper alpha oscillations

in the human brain related to verbal/performance and

emotional intelligence

Norbert Jausovec*, Ksenija Jausovec

Univerza v Mariboru, Pedagoska fakulteta, Koroska 160, 2000 Maribor, Slovenia

Received 24 August 2004; received in revised form 12 September 2004; accepted 9 December 2004

Available online 19 January 2005

Abstract

Participating in the study were 30 respondents, who could be clustered as high-average verbal/performance intelligent (HIQ/

AIQ), or emotionally intelligent (HEIQ/AEIQ). The EEG was recorded while students were performing two tasks: the Raven’s

advanced progressive matrices (RAPM), and identifying emotions in pictures (IDEM). Significant differences in event-related

desynchronization/synchronization (ERD/ERS) related to verbal/performance intelligence were only observed while

respondents solved the RAPM. The HIQ and AIQ groups displayed temporal and spatial differently induced gamma band

activity. Significant differences in ERD/ERS related to emotional intelligence were only observed for the IDEM task. HEIQ

individuals displayed more gamma band ERS and less upper alpha band ERD than did AEIQ individuals. It can be concluded

that HIQ and HEIQ individuals employed more adequate strategies for solving the problems at hand. The results further suggest

that emotional intelligence and verbal/performance intelligence represent distinct components of the cognitive architecture.

D 2005 Elsevier B.V. All rights reserved.

Keywords: Intelligence; Emotional intelligence; Event-related coherence; Event-related desynchronization

Intelligence represents the individual’s overall

level of intellectual ability. It serves as a general

concept that includes several groups of mental

abilities. One of the most influential divisions of

intelligence splits it into verbal, performance and

0167-8760/$ - see front matter D 2005 Elsevier B.V. All rights reserved.

doi:10.1016/j.ijpsycho.2004.12.005

* Corresponding author. Tel.: +362 2 2293 606; fax: +386 2

258 180.

E-mail address: [email protected] (N. Jausovec).

social intelligence (Thorndike, 1920). In recent years,

the term social intelligence has been replaced by

emotional intelligence—the ability to recognize emo-

tion, reason with emotion and emotion-related infor-

mation, and process emotional information as part of

general problem-solving (Mayer et al., 2000). Neuro-

physiological research has been mainly interested in

the verbal and performance components of intelli-

gence. Most of these studies have demonstrated a

negative correlation between brain activity under

ysiology 56 (2005) 223–235

N. Jausovec, K. Jausovec / International Journal of Psychophysiology 56 (2005) 223–235224

cognitive load and intelligence (Anokhin et al., 1999;

O’Boyle et al., 1995; Haier and Benbow, 1995; Haier

et al., 1988, 1992; Jausovec, 1996, 1998, 2000;

Jausovec and Jausovec, 2000a,b, 2001; Lutzenberger

et al., 1992; Neubauer et al., 1995, 1999, 2002;

Neubauer and Fink, 2003—for a detailed overview

see Appendix A). The problems studied were similar

to those used in tests of intelligence, as well as more

elementary cognitive tasks, requiring respondents to

retrieve information from long-term memory. The

explanation of these findings was an efficiency theory.

This efficiency may derive from the non-use of many

brain areas irrelevant for good task performance as

well as the more focused use of specific task-relevant

areas in high intelligent individuals.

Studies which have used memory and learning

tasks, requiring encoding of information have pro-

duced some inconsistent results, opposite to what

would be predicted by the neural efficiency hypoth-

esis (Klimesch, 1999; Klimesch and Doppelmayr,

2001; Doppelmayr et al., 2002; Jausovec and Jauso-

vec, 2004a). Some studies have shown a specific

topographic pattern of differences related to the level

of intelligence. High-ability subjects made relatively

greater use of parietal regions, whereas low-ability

subjects relied more exclusively on frontal regions.

(Gevins and Smith, 2000; Jausovec and Jausovec,

2004a). More generally, these results suggest that

higher-ability subjects tend to better identify strategies

needed for the solution of the task at hand. It was

further reported that high intelligent subjects dis-

played more brain activity in the early stages of task

performance, while average individuals showed a

reverse pattern. This temporal distribution of brain

activity suggests that cognitive processes in high

intelligent individuals are faster than in average

intelligent individuals (Jausovec and Jausovec,

2004b).

A second characteristic of the reported studies

employing the EEG methodology was that they have

almost exclusively based their findings on analyzing

the alpha (7–12 Hz) and theta (4–6 Hz) bands.

Probably because the relationship of activity in these

bands with mental effort is well documented. Alpha

amplitude tends to decrease (desynchronization) with

increases in mental effort, while theta band amplitude

tends to increase (synchronization) (Nunez et al.,

2001; Klimesch, 1996, 1997, 1999).

Recent research has revealed that the gamma band

(N30 Hz) may be of particular relevance to cogni-

tion—attention and arousal, basic acoustic and visual

perception, perception of gestalt and language, music

perception (for review, see Pulvermuller et al., 1997;

Tallon-Baudry and Bertrand, 1999; BaYar et al.,

2001; Bhattacharya et al., 2001). Of special rele-

vance for high level cognitive processes is the

induced gamma band activity. In contrast to evoked

gamma responses which are strictly phase-locked,

induced gamma activity consists of oscillatory bursts

whose latency jitters from trial to trial and its

temporal relationship with stimulus onset is fairly

loose (Tallon-Baudry and Bertrand, 1999). It has

been suggested that induced gamma activity reflects

a binding mechanism which is enhanced when a

coherent percept is created in response to a given

stimulus. Neuronal activity is expressed in spatially

separate areas of the cortex, which requires processes

for linking the separate nodes of activity, thereby

allowing identification of the object as a whole. The

linking mechanism is provided by the oscillations in

the gamma band (Singer and Gray, 1995). To our

knowledge, there are no studies relating gamma band

activity to the level of intelligence. Some indirect

conclusions on the relationship between gamma band

oscillations and intelligence can be made based on

the research of Struber et al. (2000) and Bhatta-

charya et al. (2001). Using visual motion tasks,

Strqber showed that better task performance was

associated with higher gamma band activity. Bhatta-

charya showed that while listening to music, degrees

of the gamma band synchrony over distributed

cortical areas were found to be significantly higher

in musicians than non-musicians.

The present study followed two major goals: The

first one was to investigate differences in verbal/

performance and emotional intelligence by analyzing

event-related desynchronization/synchronization

(ERD/ERS) and event-related coherence (ERCoh)

in the induced gamma band. The neural efficiency

theory would predict less induced gamma activity

(ERD) in high verbal/performance intelligent indi-

viduals, while solving problems taken from an IQ

test, and no differences while identifying emotions

in pictures, which is one aspect of emotional

intelligence. A reverse pattern would be expected

for the high emotional intelligent group of individ-

N. Jausovec, K. Jausovec / International Journal of Psychophysiology 56 (2005) 223–235 225

uals. A second issue of the study was to investigate

intelligence-related differences in ERD/ERS in the

low frequency band. It was expected that the

comparison of ERD/ERS in the gamma band with

ERD/ERS in the low frequency band could provide

additional insight into differences related to intelli-

gence. The upper alpha band (10–12 Hz) was

chosen because it is most sensitive to semantic, or

task specific effects (Klimesch, 1999), and because

it allows for comparison with other reported studies.

1. Method

1.1. Subjects

The sample included 30 right-handed individuals

(18 females and 12 males). The participants were

student-teachers taking a course in psychology. The

mean age of the sample was 20.8 years (SD=0.7;

range 19–21). The individuals were selected from a

sample of 825 individuals who were tested with

emotional intelligence tests MSCEIT (Mayer–Salo-

vey–Caruso Emotional Intelligence Test, Mayer et

al., 2002), and with 9 WAIS subtests (6 verbal:

Information, Digit Span, Vocabulary, Arithmetic,

Comprehension, and Similarities; and 3 performance:

Picture Completion, Picture Arrangement, and Digit

Symbol). First were selected those students who

showed high or average scores on the WAIS and

MSCEIT tests. Next were selected students who

showed a low relationship between verbal/perform-

ance IQ and emotional IQ. (e.g., high IQ–average

emotional IQ and reverse—the Pearson’s correlation

between verbal/performance IQ and emotional IQ

Table 1

Means and standard deviations of verbal/performance intelligence and em

Emotional inte

AEIQ

Verbal and performance Male AIQ n=4 IQ

intelligence EI

HIQ n=3 IQ

EI

Female AIQ n=5 IQ

EI

HIQ n=5 IQ

EI

was not significant, r=0.26; pb0.16). In that way two

groups were formed: high (n=13)/average (n=17)

emotional intelligent groups (HEIQ/AEIQ–MSCEIT

emotional intelligence test), and high (n=15)/average

(n=15) verbal/performance intelligent groups (HIQ/

AIQ-WAIS full-scale IQ scores). The test results are

summarized in Table 1.

1.2. Procedure and materials

The EEG was recorded while students were

performing two tasks. The first one was based on

a modified version of Raven’s progressive matrices

(RAPM), and measured the analytical–figural

aspects of intelligence (Sternberg et al., 1996).

Students saw a figural matrix with the lower right

entry missing. They had to say which of the 4

options fitted into the missing space. The figures

were based on the advanced progressive matrices set

I and II (Raven, 1995). Forty matrices were

presented on a computer screen positioned about

100 cm in front of the respondent. The matrices

were presented at fixed 10 s interstimulus intervals.

They were exposed for 5 s following a 2-s interval,

when a cross was presented. During this time the

students were instructed to press a button (1–4)

which indicated their answer. The second task

measured the ability to identify emotions (IDEM),

which is one aspect of emotional intelligence

(Mayer et al., 2000). Forty pictures taken from

MEIS—the experimental version of the MSCEIT

test (Mayer et al., 2000), were presented at fixed 8 s

interstimulus intervals. Each picture was exposed for

5 s, followed by a 3-s interval with no stimulus

presentation. The respondents were instructed to

otional intelligence in relation to gender

lligence

HEIQ

=111 SD=5.5 n=2 IQ=108 SD=4.9

Q=66 SD=7.2 EIQ=119 SD=2.8

=127 SD=3.1 n=3 IQ=126 SD=1.7

Q=97 SD=3.0 EIQ=120 SD=5.1

=108 SD=6.6 n=4 IQ=106 SD=10.7

Q=72 SD=8.8 EIQ=123 SD=2.1

=127 SD=4.1 n=4 IQ=127 SD=3.7

Q=97 SD=5.1 EIQ=121 SD=4.5


mentally determine how much is each feeling

(happiness, sadness, fear, surprise, etc.), expressed

by the presented face or picture. No overt response

was requested because the scoring of the emotional

tasks was not based on a dcorrectT answer, but on

general consensus. A respondent’s general consensus

score on each task compares that individual’s

performance to the more than 1000 people in the

Slovene normative database who have taken the test.

To ensure that participants did emotion processing

during the recorded intervals, they were instructed

that upon finishing the test they would have to

report about the identified emotions. Even though

the RAPM and IDEM tasks differed with regard to

the way of responding, this had no major influence

on the ERD/ERS analysis because only time seg-

ments were analyzed where no manual response was

required. All pictures were generated by the STIM

stimulator.

1.3. EEG recording and quantification

EEG was recorded using a Quick-Cap with

sintered (Silver/Silver Chloride; 8 mm diameter)

electrodes. Using the Ten-twenty Electrode Place-

ment System of the International Federation, the

EEG activity was monitored over 19 scalp locations

(Fp1, Fp2, F3, F4, F7, F8, T3,T4, T5, T6, C3, C4,

P3, P4, O1, O2, Fz, Cz and Pz). All leads were

referenced to linked mastoids (A1 and A2), and a

ground electrode was applied to the forehead. Addi-

tionally, vertical eye movements were recorded with

electrodes placed above and below the left eye.

Electrode impedance was maintained below 5 kV.

The digital EEG data acquisition and analysis system

(SynAmps) had a bandpass of 0.15–50.0 Hz. At

cutoff frequencies the voltage gain was approxi-

mately �6 dB. The 19 EEG traces were digitized

online at 1000 Hz with a gain of 1000 (resolution of

0.084 AV/bit in a 16-bit A to D conversion), and

stored on a hard disk. Epochs were comprised from

the 3000 ms preceding and 5000 ms following the

stimulus presentation and automatically screened for

artifacts. Excluded were all epochs showing ampli-

tudes above F50 AV (less than 3%). The analysis

was performed with the Scan 4.3 software.

The upper alpha frequency band was individually

determined based on the mean alpha peak frequency

(IAF=10.2 Hz, SD=0.80) (Klimesch, 1999; Burgess

and Gruzelier, 1999). On average, this method

resulted in a band of 10.2–12.2 Hz for the upper

alpha band. The broad gamma band had a range

from 31 to 49 Hz. The ERD/ERS and ERCoh were

determined using the method of complex demodu-

lation with a simultaneous signal envelope compu-

tation (Andrev, 1999; Otnes and Enochson, 1978;

Thatcher et al., 1994). In this method the raw data

for each channel are multiplied, point by point, by a

pure cosine based on the selected center frequency,

as well as by a pure sine having the same center

frequency. Both time series are then low-pass

filtered (zero-phase digital filter �48 dB/octave

roll-off) by the half-bandwidth (1 Hz for the upper

alpha frequency band, and 9 Hz for the gamma

band).

The quantification of induced ERD was done using

the intertrial variance method (induced, non-phase-

locked activity). The formulas used were as follows

(Pfurtscheller, 1999):

IV jð Þ ¼1

N � 1

XNi¼1

�yf i; jð Þ � yf jð Þ

�2

ð1Þ

In Eq. (1) N is the total number of trials, yf(i,j) is the

jth sample of ith trial data, and y f( j) is the mean of the

jth sample over all trials. The ERD (IV) data were

used to calculate the ERD/ERS values which were

defined as the percentage change of the power at each

sample point (Aj), relative to the average power in the

resting 1000 ms reference interval (R) preceding the

stimulus onset (�1500 ms to �500 ms):

ERD jð Þ% ¼ R� Aj

Rð2Þ

A positive ERD indicates a power decrease, and a

negative ERD a power increase (Pfurtscheller, 1999).

The ERD/ERS values were determined for four

1000-ms time windows (from stimulus onset till

4000 ms). The ERD/ERS values were collapsed for

the frontal (Fp1, Fp2, F3, F4, F7, F8, Fz) and

parieto-occipital (T5, T6, P3, P4, O1, O2, Pz)

electrode locations. This aggregation of electrodes

was used because several studies have reported a

frontal/parieto-occipital topographic pattern of brain

activity related to high/low levels of intelligence


(e.g., Gevins and Smith, 2000; Haier et al., 2003;

Jausovec and Jausovec, 2004a).

The induced ERCoh was determined from the

same time series: yi,n( j). The time-dependent complex

correlation between each pairwise combination of

channels, i and k, was then computed as:

qq ¼

XJj¼1

hy

jð Þi;n � yi;n

ihy

jð Þk;n � yk;n

i4ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiXJj¼1

��y jð Þi;n � yi;n

��2 XJj¼1

��y jð Þk;n � yk;n

��2vuut

ð3Þ

In Eq. (3) * denotes complex conjugation and y i,nand yk,n are the complex average potentials (i.e., event

related—calculated by averaging the complex time

series for channel i across all trials), which are

removed prior to forming the cross-products. The

induced ERCoh method utilizes the normal coherence

formula, however, with explicit subtraction of the

mean (Thatcher et al., 1994).

To reduce the large data set the ERCoh measures

were averaged for 1000 ms time windows (from

stimulus onset till 4000 ms). They were Fisher-z-

transformed and collapsed with respect to distances

and location into frontal, parieto-occipital, and long

distance (connecting frontal with parietal, or occipi-

tal electrode locations) Z-ERCoh. This aggregation

of electrodes was used to allow comparison with the

patterns obtained with the ERD/ERS analysis. The

long distance aggregation of electrodes was per-

formed, because of reported differences in long

distance coherence in relation to music perception

(Bhattacharya et al., 2001).

1.4. Statistical analysis

The data were analyzed using the statistical

package SPSS for Windows 11. All univariate

repeated measure analyses of variance were corrected

for violation of the sphericity assumption Huynh-

Feldt (Jennings, 1987). The behavioral data (RAPM

scores) were analyzed by an ANOVA with the

between-subject factor of verbal/performance and

emotional intelligence. The ERD/ERS data were

analyzed by separate general linear models (GLM)

for repeated measures. Between-task differences

were analyzed for each frequency band—4 (Time:

1000 ms segments from stimulus onset till 4000

ms)�2 (Location: frontal, parieto-occipital).

Between-group differences were analyzed for each

task (RAPM and IDEM), frequency band (induced

gamma and upper alpha), and intelligence (verbal/

performance and emotional intelligence)—4 (Time:

1000 ms segments from stimulus onset till 4000

ms)�2 (Location: frontal, parieto-occipital)�2 (Intel-

ligence level: high, average)�2 (gender). The

ERCoh data were analyzed in the same way, except

that the factor Location had three levels: long, frontal

and parieto-occipital. All between-group differences

were analyzed for all tasks—correctly and incorrectly

solved.

2. Results

2.1. Behavioral data

The ANOVA conducted for the between-group

differences of the RAPM task showed a significant

between-group difference for the HIQ/AIQ groups

( F (1 ,26)=26.16, pb2.5e�05; MHIQ=23.13,

MAIQ=17.39); and no difference for the HEIQ/AEIQ

groups (MHEIQ=20.33, MAEIQ=20.18).

2.2. Between-task differences

Differences related to the type of tasks (RAPM/

IDEM) were only observed in the upper alpha band

ERD/ERS. Significant task by time (F(3,87)=3.68,

pb0.031), and task by location (F(1,29)=16.90,

pb3.0e�04), interaction effects were observed. The

magnitude of upper alpha desynchronization was

about equal for both tasks. The RAPM task in

comparison to the IDEM task showed less frontal

and more parieto-occipital desynchronization. For the

RAPM task this desynchronization was more pro-

nounced in the later phases of problem solving,

whereas for the IDEM task upper alpha desynchro-

nization was higher in the first 1000 ms of

identifying emotions in faces and pictures.

2.3. Verbal/performance intelligence—RAPM

The GLM conducted for the ERD/ERS in the

induced gamma band showed significant interaction

ERD/ERS %8 .7

.6

.5

.4

.3

.2

.1

.7

.6

.5

.4

.3

.2

.1

.7

.6

.5

.4

.3

.2

.1

4

0

-4

-8

-12

8

4

0

-4

-8

-12

8

4

0

-4

-8

-12

.7

.6

.5

.4

.3

.2

.1

8

4

0

-4

-8

-12

0 - 1000 (ms)

1000 - 2000 (ms)

2000 - 3000 (ms)

3000 - 4000 (ms)

frontal frontalparieto-occipital

A|Q H|Q

parieto-occipitallong

ERCoh

Fig. 1. Mean percentages of ERD/ERS (left column) and z-transformed ERCoh (right column) of induced gamma band activity of AIQ and HIQ

individuals while solving the RAPM. The ERD/ERS data are collapsed for the frontal and parieto-occipital brain areas, while the ERCoh data

are collapsed for long (connecting frontal areas with parietal or occipital areas), frontal and parieto-occipital electrode locations.


LOCATION

parieto-occipitalfrontallong

Z-E

RC

oh

.7

.6

.5

.4

.3

.2

.1

GROUP

AIQ

HIQ

Fig. 2. Z-transformed ERCoh values of induced gamma band

activity of AIQ and HIQ individuals while identifying emotions in

pictures (IDEM).

TIME (s)

3-42-31-20-1

Syn

chro

niza

tion

- D

esyn

chro

niza

tion

%

20

10

0

-10

GROUP

AEIQ

HEIQ

Fig. 3. Mean percentages of ERD/ERS of induced gamma band

activity of AEIQ and HEIQ individuals while identifying emotions

in pictures (IDEM).


effects between the factors time and group

(F(3,78)=3.14, pb0.048), as well as between the

factors time, location and group (F(3,78)=3.68,

pb0.019). As can be seen in the left column of

Fig. 1, the HIQ group displayed increased gamma

band ERS from stimulus onset till 3000 ms,

whereas the AIQ group displayed more induced

gamma band ERS in the 3000–4000 ms time

segment. The two groups also differed with respect

to the brain areas they activated. In the HIQ group,

ERS in the induced gamma band was mainly

located in the parieto-occipital areas accompanied

by ERD in the frontal areas, while in the AIQ group

ERS in the induced gamma band was also observed

in frontal brain areas. A similar pattern of the

induced gamma band activity was also present for

the ERCoh data. The GLM revealed a significant

interaction effect between the factors time and group

(F(3,78)=3.07, pb0.038), and between the factors

time, location and group ( F (6,156)=7.00,

pb1.32e�6). As can be seen in the right column

of Fig. 1, the HIQ group showed a time-related

increase in induced gamma band coherence over the

parieto-occipital areas, whereas the AIQ group

showed an increase in induced gamma band

coherence over the frontal areas which was accom-

panied by a coherence decrease over the parieto-

occipital areas.

The GLM conducted for the ERD/ERS data in the

induced upper alpha band showed no significant

intelligence-related differences. Both groups showed

an equal time-related increase of ERD from stimulus

onset (12% ERD) till 4000 ms (22% ERD).

2.4. Verbal/performance intelligence—IDEM

In the induced gamma band for the ERD/ERS

data, no significant group-related differences were

observed. The GLM for the ERCoh data showed a

significant interaction effect between the factors

location and group (F(2,52)=4.01, pb0.033). As

can be seen in Fig. 2, the HIQ group showed higher

gamma band coherence over the parieto-occipital

areas, whereas the AIQ group showed more gamma

band coherence over the frontal areas and long

distance associations—connecting frontal and pari-

eto-occipital areas. This pattern is similar to the one

observed in HIQ individuals while solving the

RAPM task.

The analysis of ERD/ERS in the induced upper

alpha band showed no significant group-related

differences.


2.5. Emotional intelligence—RAPM

The GLMs conducted for the ERD/ERS and

ERCoh in the induced gamma band, as well as the

ERD/ERS in the induced upper alpha band, showed

no significant differences related to the level of

emotional intelligence.

2.6. Emotional intelligence—IDEM


induced gamma band showed a significant main

effect for the factor group (F(1,26)=6.37, pb0.018),

and a significant time by group interaction effect

(F(3,78)=3.78, pb0.014). The HEIQ individuals

while solving the IDEM task showed induced

gamma band ERS, whereas the AEIQ individuals

showed induced gamma band ERD. As can be seen

in Fig. 3, this difference increased from stimulus

onset till 4000 ms.

The GLM conducted for the ERCoh in the

induced gamma band revealed a significant time,

location, group and gender interaction effect

(F(6,156)=2.81, pb0.014). Subsequent GLMs con-

ducted for each location separately revealed a

significant group by gender interaction effect for

TIME (s)3-42-31-20-1

Des

ynch

roni

zatio

n %

30

20

10

0

GROUP

AEIQ

HEIQ

Fig. 4. Mean percentages of ERD/ERS of induced upper alpha band

activity of AEIQ and HEIQ individuals while identifying emotions

in pictures (IDEM).

the frontal ERCoh (F(1,26)=4.89, pb0.036). The

HEIQ group showed no significant gender-related

differences of ERCoh values, whereas the AEIQ

males displayed much higher ERCoh values than the

AEIQ females.


induced upper alpha band showed a significant time

by group interaction effect F(3,78)=6.15, pb0.001).

As can be seen in Fig. 4, the AEIQ individuals

displayed higher induced upper alpha ERD than the

HEIQ group, who showed a time-related decrease in

induced upper alpha ERD.

3. Discussion

The central goal of the study was to investigate

differences in induced gamma band ERD/ERS and

ERCoh related to the level of verbal/performance and

emotional intelligence.

In relation to the level of verbal/performance IQ

two differences were observed. First, the temporal

distribution of ERD/ERS in the gamma band

suggests that the integration of visual information

in HIQ individuals was faster than in AIQ individ-

uals. In HIQ individuals induced gamma ERS

increased from stimulus onset till 3000 ms and then

started to decrease, whereas in AIQ individuals

induced gamma ERS started to increase after 3000

ms from stimulus onset. A second characteristic of

HIQ brain activity was that induced gamma band

ERS and ERCoh were more intense over the parieto-

occipital brain areas, while in AIQ individuals they

were more intense over frontal brain areas. Several

PET and fMRI studies have shown that parieto-

occipital brain areas play a central role in spatial

encoding and retrieval, as well as spatial perception

and imagery, while frontal areas are more involved

in monitoring and manipulating information held in

the working memory (Cabeza and Nyberg, 2000).

From this viewpoint, the greater involvement of the

parieto-occipital brain areas in HIQ individuals could

point to a more adequate strategy use which to a

greater extent focused on the figural information

provided by the RAPM. Support for this hypothesis

is also provided by a recent study showing that

individuals who describe themselves in terms of

concrete cognitive processes exhibited lower frontal


EEG coherence and higher spectral power in the

gamma band (Travis et al., 2004). It could be further

speculated that AIQ individuals were to a greater

extent preoccupied with information held in short-

and long-term memory, which could also point to an

inferior capacity of working memory. The temporal

and spatial differences observed in induced gamma

band oscillations between the HIQ and AIQ groups

of individuals are in line with theories of intelligence

which regard general mental speed and working

memory capacity as key components of the cognitive

architecture (Kyllonen and Christal, 1990; Kail and

Salthouse, 1994; Demetriou et al., 2002).

The analysis of ERD/ERS in relation to the level

of emotional intelligence revealed a clear cut

difference in brain oscillations between the induced

upper alpha and gamma band. This difference was

only present for the emotional intelligence task of

identifying emotions in pictures—IDEM. The pat-

tern of ERD/ERS in the induced upper alpha band

was in line with the neural efficiency theory—HEIQ

individuals displayed a time-related decrease in

ERD, whereas the AEIQ individuals displayed

increased ERD. This pattern of induced upper alpha

oscillation replicates the findings obtained by a

previous study also investigating differences

between high and average emotional intelligent

individuals (Jausovec et al., 2001). On the other

hand, the pattern of ERD/ERS in the induced

gamma band was contrary to what would be

predicted by the neural efficiency theory—the HEIQ

group displayed induced gamma band ERS, while

the AEIQ group displayed induced gamma band

ERD. The difference increased from stimulus onset

till 4000 ms. A similar explanation as for the

differences observed in verbal/performance intelli-

gence could also apply for the differences in

emotional intelligence, namely that the HEIQ

individuals solved the IDEM task by relying more

on figural and less on semantic information pro-

vided by the displayed pictures. This would explain

the increased ERS in the induced gamma band and

the decreased ERD in the induced upper alpha band

shown by the HEIQ group. A reverse strategy–more

semantically and less figural oriented–could be

hypothesized for the AEIQ group of individuals.

The ERD/ERS in the induced upper alpha as

well as induced gamma band in HIQ/AIQ individ-

uals while solving the RAPM were opposite to

what would be predicted by the neural efficiency

hypothesis and contrary to previous studies which

used the same or similarly complex problem

solving tasks while recording brain activity (e.g.,

Haier et al., 1988, 2003; Jausovec, 2000). Two

reasons may have accounted for this difference.

First, the studies reporting results which were in

line with the neural efficiency theory (NET) have

used PET methodology or recorded the ongoing

EEG (see Appendix A). In comparison with the

ERD/ERS approach, both methods (PET and ongoing

EEG) have a rather poor temporal resolution—about

30 min for PET and 10–15 min for the ongoing

EEG. Hence, such methods may not be sensitive

enough to measure subtle changes in brain activity,

especially when such changes occur over short

times. Intense and short lasting brain activity of high

IQ subjects could in the averaged (10–15 min)

power spectrogram provoke a false impression of

less brain activity supporting the NET. This masking

effect could be more pronounced for complex and

difficult tasks like the RAPM, which require a longer

solution time. It is in this direction that the results

of the present study point, suggesting faster visual

integration processes in HIQ individuals. A second

reason for the discrepancy between the findings of

the present study and other studies using similar

tasks could be that studies which reported results in

line with the NET based their findings exclusively on

the analysis of the theta and alpha bands. In the

present study significant verbal/performance intelli-

gence-related differences were only observed in the

induced gamma band. A possible reason could be

the figural content of the RAPM requiring respond-

ents to perceive the matrix as a whole. Oscillations

in the induced gamma band are assumed to serve as

a mechanism for the visual representation of objects

and as a means of constructing coherent representa-

tions based on the integration of visual information

(Tallon-Baudry and Bertrand, 1999), whereas oscil-

lations in the upper alpha band are related to semantic

processes (Klimesch, 1999).

The analysis of ERD/ERS in the induced gamma

band showed a clear cut relation to the level of

emotional intelligence on one hand, and the level of

verbal/performance intelligence, on the other. Differ-

ences related to the level of verbal performance/


intelligence were only reflected in induced gamma

band oscillations while respondents were solving the

RAPM task, and vice versa, differences related to

the level of emotional intelligence were only

reflected in induced gamma band oscillations while

respondents were solving the IDEM task. Therefore,

it could be speculated that the verbal/performance

and emotional intelligences represent distinct com-

ponents of the cognitive architecture. Further sup-

port for this speculation was provided by the

analysis of task-related differences in brain activity.

Significant differences were only observed in the

upper alpha band oscillations. The comparison of

upper alpha ERD between the RAPM and IDEM

tasks for the latter showed more upper alpha ERD

in frontal brain areas which were extremely high in

the first second of the process and declined in the

latter phases. This spatial and temporal distribution

of upper alpha ERD is in line with studies stressing

the importance of the frontal cortex in emotional

behavior. With respect to this process, the decoding

and rapid readjusting of the reinforcement value of

visual signals is likely to be crucial for emotions

(for a review, see Roberts et al., 1998). As stressed

by Zajonc (1980), emotions are formed as first

impressions and do not need a prolonged elabo-

ration as reasoning does. On the other hand, no

task-related differences in the gamma band oscil-

lations (ERD and ERCoh) were observed. As

mentioned, gamma band oscillations are assumed

to relate to the integration of visual information

(Tallon-Baudry and Bertrand, 1999). The RAPM

and IDEM tasks were both figural, and probably

required the same amount of visual information

processing, therefore no task-related differences in

the gamma band could be observed. Task-related

differences were only present in the upper alpha

band which is related to semantic processes (Kli-

mesch, 1999). These results provide additional

evidence for the functional specificity of induced

gamma activity with respect to induced upper alpha

activity (Tallon-Baudry and Bertrand, 1999).

The functional specificity of the analyzed fre-

quency bands and tasks could further explain the

obtained ability-related patterns, which in the

gamma band were opposite to the NET, whereas

in the upper alpha band for the IDEM task they

were congruent with the NET, and opposite for the

RAPM task. The RAPM and IDEM tasks are both

figural and therefore a more figural and less

semantic strategy was probably an adequate solution

approach. Such a solution approach could explain

the obtained temporal and topographic patterns of

induced gamma band oscillations in HIQ and HEIQ

individuals while solving the RAPM and IDEM

tasks. Based on the between-task analysis one could

further speculate that the reasoning processes

involved in solving the RAPM task, and the

emotional judgments needed for the solution of

the IDEM task to some extent differed with respect

to the level of semantic processes involved, and

therefore provoked induced upper alpha oscillations

opposite and congruent with the NET.

Some studies have shown that intelligence-related

differences are also related to gender and task

modality (e.g., Neubauer et al., 2002; Neubauer

and Fink, 2003). The present study has revealed

only one gender- and ability-related difference

which occurred in relation to the level of emotional

intelligence. This could be expected, because several

studies have shown that females are superior to

males with regard to the level of emotional

intelligence, even if scores are based just on male

consensus (Mayer et al., 2000, 2002). A possible

explanation for the lack of displayed gender–ability-

related differences could be a methodological one.

Most of the studies which reported such interactions

were correlation studies, using figural and verbal

tasks, whereas the present study was based on

extreme groups using only figural tasks.

In conclusion, the results of the present study

suggest that high verbal/performance and emotional

intelligent individuals employed more adequate

strategies for the tasks which are related to their

intellectual excellence. This strategy was probably

more figural and less semantic. The question

remains if this predominantly figural strategy was

employed because of the figural nature of the

RAPM and IDEM tasks, or, as some theories

suggest, that spatial information processing strat-

egies (spatial-regrouping) are superior to verbal and

other types of strategies—at least in some areas of

endeavor (Cranberg and Albert, 1988). Further

research will be needed to clarify if such a strategy

of spatial-regrouping is superior also for tasks with

no figural content.


Appendix A

Overview of studies investigating the relationship between brain activity and intelligence.

Author Method Tasks used NET

Anokhin et al., 1999 EEG coherence, and dimensional complexity,

broad band: theta, alpha and beta

Verbal (naming of categories); and

visuo-spatial (mental rotation) tasks

Y

Doppelmayr et al., 2002 Log-transformed EEG power values in 3

narrow alpha frequency bands

Resting (correlation with different

intelligence tasks—LGT-3 and IST-70)

N

Gevins and Smith, 2000 Event-related potentials (ERP), EEG power

values, theta, lower alpha band

Working memory tasks (dn-backT) Y/N

Haier et al., 1988 Positron emission tomography (PET)—

fluor-2-deoxyglucose

Raven’s advanced progressive matrices

(RAPM)

Y

Haier et al., 1992 PET—fluor-2-deoxyglucose Computer game bTetrisQ Y

Haier and Benbow, 1995 PET—deoxyglucose Mathematical reasoning Y

Haier et al., 2003 PET—fluor-2-deoxyglucose RAPM, viewing emotional video tapes Y

Jausovec, 1996 EEG spectral power, broad alpha band Resting, solving of closed and open

problems, free-recall tasks

Y

Jausovec, 1998 EEG spectral power, and entropy measures,

broad alpha band

Resting, Stroop tasks, reasoning tasks,

numerical tasks

Y

Jausovec, 2000 EEG spectral power and coherence, lower

(7.9–10.0 Hz), and upper alpha band

(10.1–12.9 Hz)

Solving of closed and open problems Y

Jausovec and Jausovec, 2000a ERP—approximated entropy measures Visual and auditory oddball tasks Y

Jausovec and Jausovec, 2000b Induced and event-related desynchronization

(ERD), theta (4–7 Hz) and upper alpha

(10–13)

Auditory oddball task Y

Jausovec and Jausovec, 2001 EEG current density—low resolution brain

electromagnetic tomography (LORETA)

Auditory oddball task Y

Jausovec et al., 2001 Induced and event-related ERD, theta

(4.4–6.4 Hz), lower-2 alpha (8.4–10.4 Hz),

upper alpha (10.4–12.4)

Emotional intelligence tasks Y

Jausovec and Jausovec, 2003 EEG current density—LORETA RAPM, emotional intelligence tasks Y

Jausovec and Jausovec,

2004a

EEG current density—LORETA, induced

ERD, theta (4.4–6.3 Hz), lower-1 alpha

(6.4–8.3), lower-2 alpha (8.4–10.3 Hz),

upper alpha (10.4–12.3)

Memory tasks N

Jausovec and Jausovec,

2004b

Induced ERD, theta (4.17–6.16 Hz), lower-1

alpha (6.17–8.16) lower-2 alpha (8.17–10.16

Hz), upper alpha (10.17–12.16)

Working memory and learning tasks Y/N

Klimesch, 1999 ERD, theta, upper alpha band Semantic and episodic memory tasks N

Klimesch and Doppelmayr,

2001

Log-transformed EEG power values,

in 3 narrow alpha frequency bands

Resting (correlation with different

intelligence tasks—LGT-3 and IST-70)

N

Lutzenberger et al., 1992 EEG dimensional complexity, broad band:

2–35 Hz

Resting, emotional imagery N

Neubauer et al., 1995 ERD, upper alpha Sentence verification test Y

Neubauer et al., 1999 ERD, narrow alpha bands (8–10 Hz,

10–12 Hz)

Posner’s letter matching task Y

Neubauer et al., 2002 ERD, upper alpha (10.70–12.69 Hz) Posner’s matching tasks (verbal,

numerical, figural)

Y

Neubauer and Fink, 2003 ERD, upper alpha (10.10–12.09 Hz) Stankov’s Triplet Number test Y

O’Boyle et al., 1995 EEG spectral power, broad alpha band. Chimerical face processing, word Y

processing

NET—Neural Efficiency Theory (Y=supporting NET; N=not supporting NET).


References

Andrev, C., 1999. Quantification of event-related coherence

(ERCoh). In: Pfurtscheller, G., Lopes da Silva, F.H. (Eds.),

Handbook of Electroencephalography and Clinical Neuropsy-

chology, Event-Related Desynchronization, vol. 6. Elsevier,

Amsterdam, pp. 119–137.

Anokhin, A.P., Lutzenberger, W., Birbaumer, N., 1999. Spatiotem-

poral organization of brain dynamics and intelligence: an EEG

study in adolescents. Int. J. Psychophysiol. 33, 259–273.

BaYar, E., BaYar-Eroglu, C., KrakaY, S., Schqrmann, M., 2001.

Gamma, alpha delta, and theta oscillations govern cognitive

processes. Int. J. Psychophysiol. 39, 241–248.

Bhattacharya, J., Petsche, H., Pereda, E., 2001. Long-range

synchrony in the g band: role in music perception. J. Neurosci.

21, 6329–6337.

Burgess, A.P., Gruzelier, J.H., 1999. Methodological advances in

the analysis of event-related desynchronization data: reliability

and robust analysis. In: Pfurtscheller, G., Lopes da Silva, F.H.

(Eds.), Handbook of electroencephalography and Clinical

Neuropsychology, Event-Related Desynchronization, vol. 6.

Elsevier, Amsterdam, pp. 139–158.

Cabeza, R., Nyberg, L., 2000. Imaging cognition II: an empirical

review of 275 PET and fMRI studies. J. Cogn. Neurosci. 12,

1–47.

Cranberg, L.D., Albert, M.L., 1988. The chess mind. In: Obler,

L.K., Fein, D. (Eds.), The Exceptional Brain. The Guilford

Press, New York, pp. 156–190.

Demetriou, A., Christou, C., Spanoudis, G., Platsidou, M., 2002.

The Development of Mental Processing: Efficiency, Working

Memory, and Thinking. Blackwell Publishing, Oxford.

Doppelmayr, M., Klimesch, W., Stadler, W., Pfllhuber, D., Heine,C., 2002. EEG alpha power and intelligence. Intelligence 30,

289–302.

Gevins, A., Smith, M.E., 2000. Neurophysiological measures of

working memory and individual differences in cognitive ability

and cognitive style. Cereb. Cortex 10, 830–839.

Haier, R.J., Benbow, C.P., 1995. Sex differences and lateralization

in temporal lobe glucose metabolism during mathematical

reasoning. Dev. Neuropsychol. 4, 405–414.

Haier, R.J., Neuchterlein, K.H., Hazlett, E., Wu, J.C., Paek, J.,

Browning, H.L., Buchsbaum, M.S., 1988. Cortical glucose

metabolic rate correlates of abstract reasoning and attention

studied with positron emission tomography. Intelligence 12,

199–217.

Haier, R.J., Siegel, B., Tang, C., Abel, L., Buchsbaum, M.S., 1992.

Intelligence and changes in regional cerebral glucose metabolic

rate following learning. Intelligence 16, 415–426.

Haier, R.J., White, N.S., Alkire, M.T., 2003. Individual differences

in general intelligence correlate with brain function during

nonreasoning tasks. Intelligence 31, 429–441.

Jausovec, N., 1996. Differences in EEG alpha activity related to

giftedness. Intelligence 3, 159–173.

Jausovec, N., 1998. Are gifted individuals less chaotic thinkers?

Pers. Individ. Differ. 25, 253–267.

Jausovec, N., 2000. Differences in cognitive processes between

gifted, intelligent, creative and average individuals while

solving complex problems: an EEG study. Intelligence 28,

213–237.

Jausovec, N., Jausovec, K., 2000a. Correlations between ERP

parameters and intelligence: a reconsideration. Biol. Psychol.

50, 137–154.

Jausovec, N., Jausovec, K., 2000b. Differences in event-related and

induced brain oscillations in the theta and alpha frequency bands

related to human intelligence. Neurosci. Lett. 293, 191–194.

Jausovec, N., Jausovec, K., 2001. Differences in EEG current

density related to intelligence. Cogn. Brain Res. 12, 55–60.

Jausovec, N., Jausovec, K., 2003. Spatiotemporal brain activity

related to intelligence: a low resolution brain electromagnetic

tomography study. Cogn. Brain Res. 16, 267–272.

Jausovec, N., Jausovec, K., 2004a. Intelligence related differences

in induced brain activity during the performance of memory

tasks. Pers. Individ. Differ. 36, 597–612.

Jausovec, N., Jausovec, K., 2004b. Differences in induced brain

activity during the performance of learning and working-

memory tasks related to intelligence. Brain Cogn. 54, 65–74.

Jausovec, N., Jausovec, K., Gerlic, I., 2001. Differences in event

related and induced EEG patterns in the theta and alpha

frequency bands related to human emotional intelligence.

Neurosci. Lett. 311, 93–96.

Jennings, J.R., 1987. Editorial policy on analyses of variance with

repeated measures. Psychophysiology 24, 474–475.

Kail, R., Salthouse, T.A., 1994. Processing speed as mental

capacity. Acta Psychol. 86, 199–225.

Klimesch, W., 1996. Memory processes, brain oscillations and EEG

synchronization. Int. J. Psychophysiol. 24, 61–100.

Klimesch, W., 1997. EEG-alpha rhythms and memory processes.

Int. J. Psychophysiol. 26, 319–340.

Klimesch, W., 1999. EEG alpha and theta oscillations reflect

cognitive and memory performance: a review and analysis.

Brain Res. Rev. 29, 169–195.

Klimesch, W., Doppelmayr, M., 2001. High frequency alpha and

intelligence. Paper presented at the Tenth Biennial Meeting of

ISSID, Edinburgh.

Kyllonen, P., Christal, R., 1990. Reasoning ability (little more than)

working memory capacity? Intelligence 14, 389–433.

Lutzenberger, W., Elbert, T., Birbaumer, N., Ray, W.J., 1992. The

scalp distribution of the fractal dimension of the EEG and its

variation with mental tasks. Brain Topogr. 5, 27–34.

Mayer, J.D., Caruso, D.R., Salovey, P., 2000. Emotional intelligence

meets traditional standards for an intelligence. Intelligence 27,

267–298.

Mayer, J.D., Salovey, P., Caruso, D.R., 2002. Mayer–Salovey–

Caruso Emotional Intelligence Test (MSCEIT). MHS, Toronto.

Neubauer, A.C., Fink, A., 2003. Fluid intelligence and neural

efficiency: effects of task complexity and sex. Pers. Individ.

Diff. 35, 811–827.

Neubauer, V., Freudenthaler, H.H., Pfurtscheller, G., 1995. Intelli-

gence and spatiotemporal patterns of event-related desynchro-

nization. Intelligence 3, 249–266.

Neubauer, A.C., Sange, G., Pfurtscheller, G., 1999. Psychometric

intelligence and event-related desynchronization during per-

formance of a letter matching task. In: Pfurtscheller, G., Lopes

da Silva, F.H. (Eds.), Handbook of Electroencephalography and


Clinical Neuropsychology, Event-Related Desynchronization,

vol. 6. Amsterdam, Elsevier, pp. 219–232.

Neubauer, A.C., Fink, A., Schrausser, D.G., 2002. Intelligence and

neural efficiency: the influence of task content and sex on the

brain–IQ relationship. Intelligence 30, 515–536.

Nunez, P.L., Wingeier, B.M., Silberstein, R.B., 2001. Spatial–

temporal structures of human alpha rhythms: theory, micro-

current sources, multiscale measurements, and global binding of

local networks. Hum. Brain Mapp. 13, 125–164.

O’Boyle, M.W., Benbow, C.P., Alexander, J.E., 1995. Sex differ-

ences, hemispheric laterality, and associated brain activity in the

intellectually gifted. Dev. Neuropsychol. 4, 415–443.

Otnes, R.K., Enochson, L., 1978. Applied Time Series Analysis.

John Wiley, New York.

Pfurtscheller, G., 1999. Quantification of ERD and ERS in the time

domain. In: Pfurtscheller, G., Lopes da Silva, F.H. (Eds.),

Handbook of Electroencephalography and Clinical Neuropsy-

chology, Event-Related Desynchronization, vol. 6. Elsevier,

Amsterdam, pp. 89–105.

Pulvermqller, F., Birbaumer, N., Lutzenberger, W., Mohr, B.,

1997. High frequency brain activity: its possible role in

attention, perception and language processing. Prog. Neurobiol.

52, 427–445.

Raven, J.C., 1995. Advanced Progressive Matrices Set I, Set II.

Oxford Psychologists Press, Oxford.

Roberts, A.C., Robbins, T.W., Weiskrantz, L., 1998. The Prefrontal

Cortex. Oxford University Press, Oxford.

Singer, W., Gray, C.M., 1995. Visual feature integration and the

temporal correlation hypothesis. Annu. Rev. Neurosci. 18,

555–586.

Sternberg, R.J., Ferrari, M., Clinkenbeard, P., Grigorenko, E.L.,

1996. Identification, instruction, and assessment of gifted

children: a construct validation of a triarchic model. Gift. Child

Q. 40, 129–137.

Strqber, D., Basar-Eroglu, C., Hoff, E., Stadler, M., 2000.

Reversal-rate dependent differences in the EEG gamma-band

during multistable visual perception. Int. J. Psychophysiol. 38,

243–252.

Tallon-Baudry, C., Bertrand, O., 1999. Oscillatory g activity in

humans and its role in object representation. Trends Neurosci.

19, 151–162.

Thatcher, R.W., Toro, C., Pflieger, M.E., Hallet, M., 1994. Human

neural network dynamics using multimodal registration of EEG,

PET and MRI. In: Thatcher, R.W., Hallet, M., Zeffiro, T. (Eds.),

Functional Neuroimaging: Technical Foundations. Academic

Press, Orlando, FL, pp. 259–267.

Thorndike, E.L., 1920. Intelligence and its use. Harper’s Mag. 140,

227–235.

Travis, F., Arenader, A., DuBois, D., 2004. Psychological and

physiological characteristics of a proposed object-referral/self-

referral continuum of self-awareness. Conscious. Cogn. 13,

401–420.

Zajonc, R.B., 1980. Feeling and thinking—preferences need no

inferences. Am. Psychol. 35, 151–175.

Differences in induced gamma and upper alpha oscillations in the human brain related to...

Documents

Transcript of Differences in induced gamma and upper alpha oscillations in the human brain related to...