Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
The effects of anxiety on the alpha rhythm of the electroencephalogram
(USC Thesis Other)
The effects of anxiety on the alpha rhythm of the electroencephalogram
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
THE EFFECTS OF ANXIETY ON THE
ALPHA RHYTHM OF THE ELECTROENCEPHALOGRAM
by
Waldo Dunbar
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA ·
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Psychology)
January 1956
. .
UNIVERSITY OF SOUTHERN CALIFORNIA
GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES 7
This dissertation, written by
· ------~---------------------· al ___ o __ Dun Pa r ----------- ----- -- -- -------
d h d
· · f h i s G "d C .
un er t e irection o ________ ui ance ommittee,
and approved by all its members, has been pre
sented to and accepted by the Faculty of the
Graduate School, in partial fulfillment of the
requirements for ~he degree of
DOCTOR OF PHILOSOPHY
-- -------------------------1?[:_·-~~--
} 1 s lo
Guidance 9or.ittCJJ
· -· ~ -- ·---✓~ ----·
Chairman
·--- ······---- --
/ .. -••'')
··ff,~'4~ ----~r--.... ----- -------
/ I I '
..,,, ' t" ;/
•• ••• : •••••••••• .J.: ••••• J..... .. •• .-• •
..
TABLE OF CONTENTS
HAPTE
I. THE P_OBLEM .
• • • • • • • • • • • • • • • •
II.
The purpose and importance of the study ..
Statement of t he problem: hypotheses •••
Introduction
• • • • • • • • • • • • • • •
The nature of electroencephalography
The alpha rhythm .......•
• •
• • •
The beta rhythm.
• • • • • • • • • • •
Related studies ..
• • • • • • • • • • • •
Personality ...
• • • • • • • • • • • • •
Psychoneuroses ..
Emotiona l factors .
• • • • • • • • • • • •
• • • • • • • • • • • •
Anxiety ..... .
• • • • • • • • • • • •
The problem of stress .
METHODOLOGY ......•
Selection of subjects .
motional stability.
Emotional instability
Materials .
• • • • • •
• • • • • • • • • •
• • • • • • • • • •
• • • • • • • • • •
• • • • • • • • • •
• • • • • • • • • •
• • • • • • • • • •
The Guilford-Zimmerman Temperament
Survey . .
• • • • • • • • • • • • • •
The stimulus
• • • • • • • • • • • • • •
PAGE
1
1
5
7
7
10
1
11
13
19
22
25
29
32
32
33
33
35
35
38
CHAPTER
The EEG and recording equipment .
• • • •
The procedure
• • • • • • • • • • • • • • •
Orientation of the subjects .
• • • • • •
Electrode placement ....
• • • • • • •
The three experimental conditions
• • • •
Treatment of the data .
• • • • • • • • • •
III. RESULTS • . .
• • • • • • • • • • • • • • • •
Inter-group comparisons
• • • • • • • • • •
Experimental condition resting 0 • • • •
Experimental condition stimulus
• • • • •
Expertmental condition recovery
• • • • •
Intra-group comparisons
• • • • • • • • • •
Condition resting vs. condition stimulus.
Condition resting vs. condition recovery.
iv
PAGE
40
41
41
43
44
47
50
54
54
54
57
59
59
61
Condition stimulus vs. condition recovery 63
IV. DISCUSSION
• • • • • • • • • • • • • • • • •
v.
Need for further research.
• • • • • • • •
...
SUMMARY AND CONCLUSIONS ..
• • • • • • • • •
Problem.
• • • • • • • • • • • • • • • • •
Method
. . - . . . . . . . . . . . . . . .
Results .
Discussion
• • • • • • • • • • • • • • • • •
• • • • • • • • • • • • • • • •
65
77
80
80
81
83
85
CHAPTER
Conclusions .
• • • • • • • • • • • • • • •
REFERENCES
• • • • • • • • • • • • • • • • • • • •
V
PAGE
86
88
LIST OF TABLES
TABLE PAGE
1. Distribution of Source of Available Subjects . 36
2. Distribution, Means, and Standard Deviations
of Alpha Wave Count for Each of Three
Experimental Conditions of Stable Group
• •
3. Distribution, Means, and Standard Deviations
of Alpha Wave Count for Each of Three
Experimental Conditions of Unstable Group
4. Significant Differences in Alpha Wave Count
During Condition Resting: Stable vs.
•
Unstable .•.•
• • • • • • • • • • • • • •
5. Significant Differences in Alpha Wave Count
During Condition Stimulus: Stable vs.
Unstable .•.......
• • • • • • • • •
6. Significant Differences in Alpha Wave Count
During Condition Recove1'Y: Stable vs.
52
53
55
Unstable . . • . . . • . • . . • . . . . . . 58
7. Intra-Group Comparisons of Stable and Un
stable: Condition Resting vs. Condition
Stimulus
• • • • • • • • • • • • •
60
TABLE p G
8.
Intra-Group Comparisons of Stable a d
Unstable: Condition esting vs .
Condition Recovery
• • • • • • • • • • • • •
62
9. Intra-Group Comparisons of Stabl e a nd
Unstable: Condition Stimulus V
•
Condition Recovery
• • • • • • • • • • • • •
64
CHAPTER I
The Problem
~he Purpose and Importance of the Study
One of the problems in the field of psychology is
to de t ermine the phyaiological mechanism underlying behav
i or . The role various bodily organs and systems play in
emotional reactions has been intensely investigated for
m any years. For the most part, however, these investiga
t ions have concerned themselves with autonomic nervous
system changes accompanying emotional activity. Since the
br a i n or particularly the cerebral cortex ii the most
important organ controlling psychological activity, the
i mport ance of studies relative to its function is obvious.
W ith the r elatively recent development of electro
physiology, and particularly electroencephalography, it
be came possible to study more directly the possible rela
tions that m i ght exist between emotional reactions a nd
electr i cal changes in the brain. The justification of the
ele ctroe ncephalogram as an indicator of the physiological
mechanisms underlying emotional activity is based on the
assumption that both the electroencephalogram and psycho
logi ca J phenomena are functions of events in the brain.
2
Among those primarily interested in the dynamics
of personality, Freud was probably the first to point out
the fundamental significance of the problem of anxiety.
Years ago, he wrote that anxiety is the ''fundamental
phenomenon and the central problem of neurosis." (7) The
significance he attached to it has continue d to be empha
sized by most psychotherapists, and the importance of its
evaluation for diagnostic purposes, as well as for trP-at
ment, is widely recognized.
More recently, May (26), in reviewing and evaluat
ing the various theories regarding the nature of anxiety,
stresses the important role anxiety plays
••. not only in psychopathology but in
the behavior of normal people. It is much
more prevalent than was formerly believed for
we are living in an 'age of anxiety' ...•
If one penetrates below the surface of poli
tical, economic, business, professional, or
domestic crises to discover their psychologi
cal causes, or if one seeks to understand
modern art, or poetry or philosophy or reli
gion, one runs athwart the problem of anxiety
at almost every turn ••.. Anxiety has
gradually come tc be seen as a central problem
in learning theory, in dynamic psychology, and
specifically in PSfChoanalysis and other forms
of psychotherapy. (26)
With the introduction of electroencephalography,
interest in the effects of anxiety upon the electrical
discharges from the brain immediately developed and has
continued to the present time. Berger, the originator of
3
electroencephalography, reported that according to
Lindsley (24), fear and anxiety produced faster alpha
waves. The predominanve of beta waves in anxiety states
has been observed by Cohn (4). Finley (6) found rapid
frequency wave patterns more prevalent in all types of
neuropsychiatric disorders than in normal records, but no
type of pattern which was characteristic of any particular
disorder. Brazier et al. (3) found the mean alpha wave
--
frequency of anxiety neurotics to be slightly higher than
the mean frequency of a normal g1oup, but still within the
normally expected range. Lindsley (23) reported no rela
tionship between emotional instability and the electro
encephalogram, but Lemere (21) noted a correspondence
between instability and a prominent alpha rhythm. None
of these studies made use of experimentally controlled
variables. In view of the inconclusive findings of these
empirical studies, additional investigations, preferably
of an experimental nature, seem warranted.
Under sufficiently intense emotional stress most
people are likely to react in a disorganized and ineffec
tive manner, and the greater the stress, the more dis
organized the behavior. Also, the greater the degree of
emotional disor
6
anization, the longer is the period re
quired for the recovery of psychological equilibrium.
4
Unstable individuals have been reported (43) to react
more strongly to a smaller amount of frustration than
stable persons and the reactions to frustrating situations
have been likened to "emotional disorganization." (32)
It is reasonable to suppose that a mildly stressful situa
tion will be experienced as more anxiety-provoking by an
emotionally unstable person than by a well-adjusted indi
vidual and is, therefore, ~xpected to create greater
disturbance in the former.
It is a well-established fact t hat the maximal
regularity of the alpha rhythm of the electroencephalogram
"' .,.. w •• ,~ .... t\ .. , ... ~..., ~ - If"~( " ' ., .. . ~ ....... . .. _ _......, ,-.u,,.. ,I'>-• .......... .. - .. _ . .. _ - · ., --· ,.., .. ' _ .._
is obtained under optimal conditions of relaxat fon an"cf ·- :--~~ ... . :
freedom from peripheral stimulation. The various condi
tions modifying, disrupting, or suppressing the alpha wave
pattern have been extensively reviewed by others (15, 22,
37). Although convincing experimental data are lacking,
leaders in electroencephalography are in general agreement
that emotional excitation tends to abolish the rhythmical
pattern of the cortical potentials, and that the alpha
wave is frequently replaced by the smaller and more rapid
beta wave (23).
The purpose of the present study was to investigate
whether a significant correlation exists between the emo
tional state of anxiety and some property of the
5
electroencephalogram. It was asked: Is the disorganized
behavior of persons under emotional stress and the disrup
tion of the rhythmical pattern of the electroencephalogram
under psychological excitation merely an interesting
coincidence, or are they both manifestations of some com
mon underlying neurological process? Conversely, does
the persistent regularity of the electroencephalogram
pattern of some persons, despite disrupting and annoying
stimuli, have its parallel in the organized behavior of
the well-integrated person in emotionally disturbing
situations?
Statement of the Problem: Hypotheses
The problem was to compare the electroencephalo
graph readings of two groups of subjects selected from a
normal population but differentiated on the basis of emo
tional stability, and to determine whether a significant
difference existed between the two sets of records.
Electroencephalograms were to be obtained under three
experimental conditions: (a) a rest period, (b) presenta
tion of the stimulus, and (c) a recovery period following
cessation of the stimulus. Thi e stimulus was a complex,
auditory problem intended to produce mild stress and
threat.
6
It has been postulated that both disorganized be
havior and the length of the period required for recovery
of integrated activity are functions of anxiety. In
addition, it has also been postulated that, when placed
in similar situations of emotional stress, emotionally
unstable persons react with a greater amcunt of anxiety
than do persons considered to be more stable. If varia
tions in the regularity of the electrical tracings from
the cortex, measured in terms of the persistence of the
alpha waves, are similarly dependent upon anxiety, it
follows that the greater the amount of anxiety created,
the greater will be the amount of disruption of the alpha
waves and the longer the period required f or the recovery
of synchronized cortical activity.
The present experiment was designed to test three
hypotheses: (a) since the two experimental groups were
representative of a normal population, there would be no
significant difference between their respective electro
encephalograms secured under a resting condition with
regard to the amount of alpha waves present; (b) the less
emotionally stable group would differ significantly from
the stable group in response to an anxiety-provoking
auditory stimulus measured in terms of the amount of alpha
waves abolished; (c) the unstable group, similarly meas
ured, woui differ significantly r ~om the stable in the
7
extent of recovery during the post-stimulation period of
the wave pattern previously obtained under a condition of
rest and minimal external stimulation.
Negative results would not necessarily indicate
that there is no relationship between the electroencephal
ogram and the em0tional state of anxiety. They would,
however, suggest that whatever correspondence exists, it
is not manifested by modifications of the alpha wave pat
tern obtained from the occipital area of the cortex and
requires a different method of investigation and analysis.
In addition, it is also possible that the effectiveness of
the criteria for distinguishing the two groups might be
inadequate.
Introduction
The Nature of Electroencephalography
Despite the fact that electrical activity has been
known since 1875 to be one of the properti~s of the living
animal brain, it was not until 1929 that the recording of
electrical potentials in a human brain was first described.
I t w a s discovered by Hans Berger, a German neuropsychia
trist, who called it the "elektrenkephalogram." its
English translation is the "electroencephalogram" and the
8
abbreviation "EEG" is customarily used to stand for both
electroencephalogram and electroencephalography depending
upon the context. Berger's initial reports were received
with widespread skepticism, but immediately following the
confirmation of his findings by Adrian and Matthews (1)
electroencephalography began to be used enthusiastically
as a clinical diagnostic aid in neurology and psychiatry
and as a research instrument in physiology and psychology.
The volume of published articles has increased to such
proportions that one source estimated it has reached the
rate of approximately one publication per day since
1946. (40)
The electrical potentials from the brain are seen
as continuous and rhythmic variations of different fre
quencies, amplitudes, and patterns of sinusoidal wave
forms. Although the precise origin of brain waves is not
known, it is generally agreed by EEG workers that they are
spontaneous and autonomous in character and are not ini
tiated by sensory stimulation. On the contrary, sensory
excitation tends to distort the rhythmic quality. Since
the electrical discharge from a single cortic~l neurone is
too small to be recorded by means of the EEG from the sur
face of the scalp, the obtained rhythmical oscillations
are widely accepted as representing the synchronized
activity of many neurones discharging in unison.
In this connection, Travis (34) ha suggested
that large, regular brain waves are indicative of a
state of cortical equilibrium represented by general
psychic activity, while a breakdown of this collective
-
action into more rapid and irregular waves of lower
amplitude represents a high degree of specific i ty in
psychic activity.
9
Summarizing the current pr e vailing views , Lindsley
states:
Some current conceptions are t hat the
potentials arise from cell bodies, that many
adjacent cortical cells must form a field of
essentially synchronous activity in order for
the potentials to summate suf ficiently to be
recorded; that the rhythmic alterations of
potential are relatively autonomous or self
generating in that they do not require or are
not positive manifestations of a sensory in
flux, but on the contrary are disrupted and
desynchronized by sensory impulses arriving
at the thalamus or cortex. Furthermore, that
the most prominent component of the electro
encephalogram, the •.. alpha rhythm, is a
manifestation of a resting state, physiologi
cally, an inattentive, waking state, psychologi
cally, and a relaxed and sensory free peripheral
state. (25)
In his initial report, Berger identified two types
of brain waves in the EEG of the normal adult which he
called "alpha" and "beta" waves. They were distinguished
by differences in frequency, amplitude, and reaction to
sensory stimulation. Although other wave patterns were
10
subsequently described by other investigators, these two,
but especially the alpha rhythm, have continued to receive
the most attention in EEG studies of emotion. In the
present investigation, interest was centered on the modifi
cation of the alpha wave under the experimental conditions
to be described later.
The alpha rhythm. By definition, the alpha wave
has a frequency of eight to twelve cycles per second with
an amplitude ranging from five to 100 microvolts. It
originates in all parts of the cortex, but in most normal
persons it is strongest and most persistent in the occipi
tal area. The regularity of the alpha rhythm in terms of
frequency and amplitude is best obtained where the subject
is in a relaxed, wakeful state, free from peripheral
stimulation. Sensory stimulation readily suppresses or
modifies the alpha pattern, varying in accordance with
the sensory area stimulated and the intensity, duration,
and frequency of the stimulus. The various conditions
affecting the alpha rhythm have been comprehensively
treated elsewhere (16, 24, 39).
The beta rhythm. Although beta waves are most
prominent over the precentral and frontal areas of the
brain, they are found to some extent over all areas.
11
They are fairly rapid, low-voltage waves, varying in fre
quency from fifteen to thirty cycles per second, and un
like the alpha, are not markedly affected by visual,
auditory, or tactual stimuli. Under conditions of excita
tion or sensory stimulation, they tend to replace the
suppressed alpha waves.
Related Studies
The considerable range of individual variability
in frequency, amplitude, and wave pattern in EEG records
has been repeatedly confirmed by numerous workers who have
studied groups of normal adult subjects. While individual
differences are pronounced, intraindividual variability
tends to be small.
Studying the consistency and individuality of
EEG's, Travis and Gottleber conclude that
.•• not only can an individual be dis
tinguished from other individuals by his
brain potentials, but under relatively con
stant experimental conditions an individual's
brain potentials are highly consistent from
day to day. (36)
It is probably this recognition of interindividual
variability and intraindividual consistency that has been
mainly responsible for prompting so many attempts to seek
personality and EEG correlates. This approach has offered
12
"plenty lure, but so far little substantial bounty" (25).
The same limitation can be applied to those studies more
directly concerned with the investigation of relationships
between emotions and the EEG.
The inconclusive, and often contradictory results
found in the literature are mainly due to the inadequacies
of the methods of investigation. In attempting to isolate
personality traits there has been too great a tendency to
rely upon subjective judgments and ratings. Results often
lack appropriate analysis and quantification. But prob
ably a more serious criticism is the common failure to
establish experimental conditions and to attempt to con
trol the variable being investigated. The preponderant
majority of studies lack experimental controls, thus mak
ing it difficult to compare and evaluate the results.
Often the results reported are merely clinical impressions
observed in the examination of a single subject.
Despite the inadequacies of the reported studies
and the lack of experimental evidence, leaders in the
field of electroencephalography remain convinced that some
personality and emotional differe nces must be reflected in
the tracings of the electrical discharges from the brain
(16, 25). There are few references in the literature
which are directly pertinent to the problem under
13
consideration. However, there is a large number of
studies having a bearing in a more general way and a
brief review of some of these should clarify the back
ground and bring the status of the present problem into
sharper form.
Personality
Gottlober (9) selected 67 subjects, 18 women and
49 men enrolled in graduate and undergraduate courses in
speech and psychology who were judged to be either intro
verted or extroverted on the basis of a subjective rating
and the results of the Nebraska Personality Inventory.
He found 82 per cent of the extroverted subjects to have
a "dominant-subdomina11t" type of wave pattern while only
18 per cent had a "mixed-rare" or low alpha pattern; 57
per cent of the introverts had high alphas and 43 per
cent low alphas. He concluded that individuals showing
a high degree of extroversion tend to show a
11
dominant
subdominant" alpha rhythm but indicated that it could not
yet be stated that introverts show a predominantly "mixed
rare" type of alpha pattern. Gottlober's study can be
criticized on a least two counts. His use of subjective
ratings makes it impossible to repeat precisely his inves
tigation. Also, his particular method of statistical
14
analysis seems not to be the most appropriate. Instead
of comparing the two groups in terms of deviations from
the normally expected alpha count, he based his conclu
sions on the differences between the percentage of high
and low alphas within each group.
Henry and Knott (13), attempting to repeat the
Gottleber study and using as subjects 40 men and 40
women classified according to degree of introversion and
extroversion as shown by the Nebraska Personality Inven
tory, not only did not obtain the same results, but their
results showed a tendency to be the opposite of Gott
lober's. Using a more appropriate statistical method of
analyzing their data, they found that only 43 per cent of
their extroverts, whereas 60 per cent of the introverts
had high alphas. In their study they found no statistical
ly significant relation between introversion and a high
alpha index.
While pursuing another study, Lindsley (23) ob
tained data on 88 normal children, ages ten to eighteen
years, on three aspects of personality; dominance or as
sertiveness measured by the Allport's Ascendance-Submis
sion est; intelligence based on the Stanford-Binet and
Otis tests; emotional stability determined by the
Woodworth-Matthew's Personal Data Sheet. These were
15
correlated with per cent time alpha and a qual i tative
rating of the pattern of the EEG. He found no significant
relationships.
Using a psychoanalytic approach, Saul, et al. (29)
searched for correlates of the individualit y of the EEG
pattern. They obtained data from 31 i ndividuals under going psychoanalysis who were retested after an interval
of seven months, using the alpha ind ex as a m ea sure of the
EEG. They found that 25 showed alpha indices on retest
agreeing with original indices within ten poi nts and on]y
two deviating more than se venteen points. All devi ations
greater than ten points were in the direction of a lowered
alpha index on retest, but there was no relationship
between these changes and the clinical changes repor ted
by the analysts. They also computed alpha indice s for 66
persons who were being analyzed or had just comple t ed
analysis. The distribution of alpha indices was c ompared
with that of a group of 200 normals and was f ound to be
similar, ranging from Oto 100 per cent.
Attempting to correlate alpha indices with t he emo tional and instinctive trends revealed by psychoanalysis ,
they found two opposing trends correlated with low an
high alpha indices. A low alpha count was relate d with a
trend toward activity and was found among individuals
16
havin "consistent, well-directed, freely indulged" activ
ity dries, and who were independent and socially aggres
sive "doers." High alpha indices were associated with a
pass i ve -dependent, receptive attitude oward other persons
and wer e found among persons who were seen as "solid
citizens , and patient workers and planners." However,
they point out, that these relationships are more in the
natur e of a t endency or trend. The "activen and "passive"
t ypes are not entirely unmixed or rigidly restricted to
high ad low alpha indices.
It is difficult to accept these conclusions with
any d gree of confidence. In addition to their failure
to defin their groups learly and objectively, Saul,
et al. ( 29 ) fail to treat their data for statistical
ignif i cance.
Following a similar approach, Rubin and Bowman (28)
studie d a group of 100 peptic ulcer patients. Data per
taining to work attitude, parental identification, attitude
toward mate and friends, and childhood history were
ob aine d by interview from some but not all patients.
They found that about 70 per cent of their patients fell
in the "dominant" EEG category of the Davis classification
system, whereas among a normal population only about 20
per cent are thus classified. This means that the alpha
17
waves obtained from the occipital area are present more
than 75 per cent of the time. Twenty per cent of their
patients were classified in the low alpha group. Since
the majority of their patients manifested a "dominant"
pattern, they concluded that there is a relationship
between peptic ulcer and a passive, receptive personality
structure. In support of their conclusions they offered
their personality data which indicated to them that peptic
ulcer patients are in general dependent, and as children
and adolescents, closely identified with their mothers.
As they reach maturity, t he high alpha group continues t.o
be dependent on a mother surrogate, but the low alpha
group achieves independence, through a strong "reaction
formation" to an underlying passivity. The "dominant"
alpha group remain attached to their jobs over a prolonged
period and usually are more passive toward their mates.
This study has several obvious shortcomings. The
differences between the obtained and the expected percent
ages are not treated for statistical significance. The
personality data was obtained only from an unspecified
limited number of the patients and no attempt was made to
classify it quantitatively. The subjects were a highly
selected group. They were all males, chosen in succession,
from a clinic in the eastern section of New York City
18
serving a marginal and sub-marginal economic group .
Saul and his collaborators (30) continued to collect
data over a three-year period and more recently reported a
modification of their earlier conclusions. They now felt
that ot all types of EEG's and not all types of individ
uals were included in their original study and moreover,
that they had so far demonstrated only a one-way relation
shi p. "e have predicted the EEG's from the make-up of
the ersonalit es, but how accurately the corresponding
personality tr nds can be answered from the EEG is a ques
tio we cannot yet a swer.'' However: it is their impres
s ion, "that the relationship is reciprocal and that it may
be possible to draw in erences from the EEG as to the
probable strength of certain emotional trends in an in
dividual.u
W lliams (42) reported data gathered from 900 sub
jects in t he British military service during World War II,
who were being studied in connect on with a selective pro cess for airplane pilots. They were divided into three
groups and the number of abnormal EEG's in each group was
determined. The results were as follows: (1) in the
highly selected group of R.A. F . pilots, 5 per ,ent of the
records were abnormal· (2) the group selected from the
medical ~orps of the air force sho ed 10 per c nt abnormal
19
records; and (3) in- the group of soldiers of the same units
who had developed psychoneurotic symptoms, 26 per cent of
the EEG' s were abnormal. W illiams concluded that there is
a surprising correlation between stability of personality
and stability of E ,G . The statistical significance of the
data is not present ed and the types of a bnormal records
are not specified .
Psychoneuroses
ecause anxiety plays such an important role in
emotiona l disorders, it is important to consider investi
gations of EEG's of psychoneurotics. The findings re
ported in the literature are varied and present no char
acteristi c picture except for a tendency toward suppres
sion of the alpha r.hythm and an increase in the low
amplitude, more rapid beta waves.
Finley (6) examined the records of over 4,500
neuropsychiat ric patients and those of 300 controls. He
found rapid frequency patterns widely distributed through
out all the neuropsychiatric disorders, but no type of
pattern which was characteristic of any particular dis
order . Finley's lack of interest in the statistical
analysis of his data is shown by his conclusion that
fas activity was present in three to eight per cent of
20
the control group, but appeared in 20 to #o per cent of
the psychoneuroses.
Brazier, Finesinger, and Cobb (3) compared the
EEG's of 100 psychoneurotic patients and 500 normal
adults. The diagnoses were based on "symptoms elicited
by examination" and four neurotic sub-groups were estab
lished: (a) thirty-one anxiety neurosis, (b) twenty
three hysteria, (c) twenty-seven reactive depression, and
(d) nineteen mixed neurosis. The percentage of different
wave frequencies in each record was determined and the
prevailing frequency was selected as the dominant fre
quency of that record. Twenty-five per cent of the nor
mals had a frequency of 10.0 cycles per second with a
J
normal distribution around this mode. The patients had
a bimodal distribution, one peak at 9.0 cycles and the
second at 10.5. A significant difference between the
mean frequency of the two groups was found. The authors
felt that the anxiety neurosis group, which had a mean of
11.2 cycles per second as against 10.6 for the normals,
was primarily responsible for the difference. A signifi
cant difference was also found between the patients and
normals in the amount of time the beta wave was present.
The reactive depressive group was held to be primarily
responsible for this difference. The authors concluded:
21
"The EG's of psychoneurotics show no gross abnormalities
of the wave patter and, in general, fall with·n the nor
mal ran e . "
Heppenstall Hill, and Slater (14) recorded EEG's
from 300 p ychoneurotic patients i a mil tary neuro
psych atri rehab lita ion cen er . The authors were
mainly in erested in invest·gating the use of the EEG n
prognosis and treatment. After discarding ten records as
unsatisfactory, they fond 121 of the re aining 290
records to be normal, 70 doubtful, and 99 had "nonspeci
fi abnormal ty.' Fast rhythms predominated in 18 per
cent and slow in 46 per ce t of the records . Less than
10 p et of the patients sho ed fast act vity i the
abnormal records . The authors commented on the fact that
oth r attributes o~ widely different kinds, such as physi
cal a ility and sensory acuity, have been found to show
greater variability in the neurotic than in the normal .
The impossibil ty of omparing studies based on
vague psychiatric classifications is so obvious that it
hardly merits mentioning . However, it is interesting to
note that in the one study of psychoneurotics reported
above which makes a serious effort to treat its data
statistically, the conclusion is drawn that the EEG's of
psychoneuro ics show no abnormalities and, in general ,
22
fall within the normal range.
Emotional Factors
The close relationship between emotion and auto
nomic nervous system factors is well known. Several
studies using the EEG and autonomic measures in investi
gations of emotional states have been reported. These
studies are also noteworthy because they are among the
very few which have introduced experimental conditions
into their investigations of emotional states and the EEG.
Assuming that emotional disturbances accompany
rustration, Sherman and Jost (33), studied the emotional
changes following different situations, including one of
frustration. Two groups of 18 children each were used as
subjects. The normal group was used as a control "because
t hey were from a private school and were known to be
extremely well adjusted." Fifteen of the 18 children in
the experimental group had been diagnosed as being
"neurotic or having behavior disturbances with neurotic
s g s . ' The remaining three were considered to be
schizophrenic . Motor and verbal behavior was observed
an au anomi c measure s , inclu ing the EEG, were taken
uri g he different experimental conditions. Frustra
t on ~as induced by means of a learning situation. The
23
subjects learned a simple series of four numbers. After
two successes, a complicated sequence using up to nine
numbers was presented which produced failure. The sub
ject was given constant assurance by the experimenter
that the series was not difficult and that others had
learned it very quickly. The alpha wave was found to be
dominant 85 per cent of the time in the control group and
62 per cent of the time in the experimental group in the
frustrating situation. In concluding, the authors stated
that the EEG and the autonomic measures were better indi
cators of emotional reactivity than overt behavior or
analysis of verbal expression. However, dominance of the
alpha rhythm in the condition of rest also differentiated
the groups. This count was not given and, consequently,
cannot be compared with the count obtained during frustra
tion. The statistical significance of the obtained
difference was not reported. The small number of subjects
and the lack of precision in defining the experimental
groups are additional reasons for caution in accept ing
the authors' conclusions.
Darrow, et al. (5) using the EEG and various
autonomic measures studied the reactions of ten individ
uals in response to unspecified "emotional stimulation."
They reported a decrease in the alpha waves and an increase
24
in the faster beta waves but the statistical significance
of the observed changes was not given. Analyzing the EEG
and the relationships among the autonomic measures, they
inferred that emotional stimulation produces two main
effects: (a) the effect of excitation processes and (b)
the homeostatic effects of a rise in blood pressure.
These results were interpreted as establishing a close
correspondence between the alpha and beta frequencies and
the frequencies of sympathetic and parasympathetic fibers
controlling the blood supply of the brain.
Following a method developed by Grinker (10) for
recording electrical potentials from the hypothalamus,
Hoagland, et al. (15) obtained simultaneous recordings
from the occipital area of the cortex and, presumably,
from the hypothalamus. Similar though not identical
activity from both regions was observed for a short period
following emotional stimulation of the subject by probing
with emotionally significant questions. Increased activ
ity of the delta index, described as waves of less than
eight cycles per second, in the two recordings was roughly
synchronized, but the delta waves from the hypothalamic
region preceded slightly those from the cortex. It was
concluded that the waves probably originated in the hypo
thalamus as a result of the autonomic activation and were
25
reflected in the cortical record as the conscious corre-
late of the emotional response. Other studies have failed
to show any correlation between delta waves and emotional
states and it now appears that the reported waves were
either artifacts or due to electrical changes in the skin
associated with autonomic effect s (25).
Anxiety
At the time the present research was initiated
there were no references in the literature to investiga
tions of anxiety and the EEG which made use of experi
mentally controlled variables. Recently, however, two
similar studies using an experimental method have appeared
in the literature.
Ulett and Gleser (37) compared the EEG's of anxious
and non-anxious subjects obtained under varying experiment
ally controlled conditions. A total of 96 subjects was
divided into three groups: (a) thirty subjects, screened
by psychiatric interview, who were considered to be least
likely to develop anxiety symptoms under stress, (b)
twenty-five neuropsychiatric patients who had been diag
nosed psychoneurotic or character disorder with anxiety as
the predominant symptom, and (c) forty-one subjects, used
as a control group made up of "sophisticated medical
26
students and hospital staff." Leads were attached to the
occipital area and the records were analyzed by an auto
matic analyser. Groups (a) and (b) were subjected to
experimental anxiety-producing situation to determine
whether such stress could affect the photic driven EEG
and whether the two groups might react differently. The
control group was sub jected to the intermittent flashing
light only. The anxiety-provoking situation was a threat
of electric shock to the hand. The subjects were told
they would experience a variety of subjective sensations
during photic stimulation which were known to the experi
menter. However, this was not true. They were asked to
report their experiences at specified intervals and they
were informed each time of the correctness or incorrect
ness of their response. They were also told that as an
incentive to correct reporting they would be given an
electric shock if they gave ten incorrect responses. The
EEG's recorded under resting condition were compared to
photically induced records obtained during and following
stress. Although the EEG's of both the anxious and non
anxious groups taken under stress differed from the ori
ginal records at the five per cent level, there was no
difference between the anxious and non-anxious groups.
The authors concluded that the occipital rhythm
induced by photic stimulation is altered by emotional
27
stress in a way similar to the blocking of the alpha wave.
This conclusion does not seem entirely warranted for sever
al reasons. There is no evidence to indicate that the
physiological stress used in the experiment is the equiva
lent of "emotional" stress. Since the two groups had been
selected on the basis of responsiveness to stress, the
failure to obtain a differential reactivity following the
experimental threat would suggest either (a) the criteria
for distinguishing the two groups were inadequate, or (b)
the stimulus was not effective in creating stress and
anxiety. They suggested that the lack of differentiation
between the anxious and non-anxious subjects might be due
to the possibility that the experimental stress increased
the attention or "vigilance" in the non-anxious to a degree
which equalled the anxiety in the neurotic patients. How
ever, it does not seem reasonable to equate attention in
a normal to anxiety in a neurotic.
In another experiment, Ulett, et al. (38) investi
gated the EEG's of anxiety-proned normals, non-anxiety
proned normals, and neurotic patients whose major symptom
.
was anxiety. The normals were carefully examined by psy
chiatric interview and psychological testing and rated
according to anxiety-proneness. Brain wave records were
taken on all subjects under the following conditions:
28
(1) at rest with the eyes closed and with the eyes opened,
(2) under stress as described in the preceding study above,
and (J) under photic stimulation. The intermittent light
stimulus varied in frequency from two to thirty flashes
per second and each frequency used was applied for a period
of forty seconds with an equal period of rest between pre
sentation of stimuli. The EEG recordings were obtained
from the occipital area and an autcmatic analyser was used
in the analysis of the data. Differences between the non
anxious and the two anxious groups were observed and a
scale for anxiety-proneness was developed using the follow
ing factors: (a) differences in the resting EEG, which
showed the anxious groups having less alpha wave activity
and a relatively high incidence of the slow and fast fre
quencies; (b) differences in the EEG response to photic
stimulation, the anxious groups reacting with a low photic
driving response in the frequencies between eight to
eighteen flashes per second, but with a higher response in
the ranges of two to seven and twenty to thirty flashes
per second; {c) differences in the amount of discomfort
and dysphoria produced during photic stimulation as re
ported by the subjects. They found the scale correlated
0.51 with the validating psychological and psychiatric
criterion for anxiety-proneness used in the initial rat
ings of the subjects.
29
This study off.ers evidence supporting the possibil
ity of measuring anxiety by means of the EEG. However,
there were several aspects that were omitted or not treated
adequately. The procedure in obtaining the resting EEG is
not described clearly. It is not known whether or not any
attempt was made to allay the anxiety of the subjects dur
ing this period. The differences noted in the resting EEG
are not subjected to proper statistical treatment. There
fore, it is not known how meaningful the obtained differ
ences are. While mention is made of a condition of stress,
the EEG responses during this period are not reported. An
attempt is made to quantify the sub jects' reports of dis
comfort ex~erience d during the photic stimulation, but this
necessarily remains sub jective and, to that extent, un
reliable. Although the anxiety-proneness scale correlated
moderately with the original validating psychological and
psychiatric criterion for anxiety-proneness, its effective
ness as a selective instrument needs to be further tested
by cross-validation.
ThA Problem of Stress
In reviewing the effects of psychological stress
upon performance, Lazarus, Deese, and Osler (20) under
score that stress cannot be adequately defined alone in
30
terms of stimulus or response operations. It is, there
fore, necessary to conceptualize it in terms of an inter
vening variable. In addition, it is also necessary to
consider the concept of motivation and the situation giv
ing rise to motivated behavior. Stress arises when the
achievement of a goal is threatened by a particular situa
tion. The use of situations which are believed to thwart
the motives of most people is considered an adequate
approach as long as one is studying the reactions of
groups and not individuals alone. They note that despite
the great variations of conditions and tasks employed in
experiment to produce stress, all "have been motivational
in frame of reference."
The great majority of studies reviewed show impair
ment of performance under stress. This is true of both
verbal and perceptual-motor responses. In addition, they
observe that many qualitative changes in behavior under
stress have been reported. Some of these characteristics
are: stereotyped and maladaptive responses, "inattention,"
"disorganized activity," and "increased overt activity."
Also frequently reported are indications of emotional
reactivity such as pulse changes, sweating, subjective
anxiety, and tremors. Studies of relationships between
performance under stress and personality correlates
31
consistently show a small but statistically significant
correlation.
The techniques used in inducing stress are grouped
into two classes: failure-stress and task-induced stress.
Unsolvable tasks, interruptions before completion, and the
introduction of false norms are methods of producing
failure-stress. Situations which make excessive demands
upon the subject, such as various types of distractions,
give rise to task-induced stress. Disparagement of the
subject's performance by the experimenter is seen as both
a threat of failure and a distraction. The technique used
in the present experiment, like disparagement, contains
elements of both classes.
CHAPTER II
Methodology
Selection of Subjects
Initially, the plan was to use neuropsychiatric
patients whose predominant symtpom was anxiety as one
group of subjects and compare their EEG's with those taken
from a representative sample of a normal population. This
idea was rejected because, in addition to the problem of
determining the degree of anxiety, it was felt that the
use of such patients would present other difficulties.
For example, the anxiety level might be so chronically
high that it would be impossible to raise it higher in the
experimental situation. Also, a chronic anxiety state may
have resulted in biochemical or neurological changes that
would affect the EEG in some unknown fashion. A third
difficulty would be the problem of designing an experi
mental condition that would have comparable threat value
to an otherwise heterogeneous group.
To overcome these objections, it was then decided
to select from a homogeneous population, two groups which
differed in the one variable of emotional stability. The
members of both groups were assumed to be relatively norrral
33
since they were functioning productively as students and,
apparently, were not incapacitated to any marked degree.
By using this sou1ce for the selection of subjects, it
was believed that the effect, if any, of the anxiety vari
able on the alpha rhythm would be highlighted. The meas
uring device was a group personality test which was con
sidered particularly appropriate for the differentiation
of the two groups. It offered the advantages of having
I
isolated the variable emotional stability by factor-
analytic procedures and of having used for its normative
data college students in the same area from which the
present data were obtained.
Emotional Stability
Those persons obtaining a C-score ranging from
seven to ten, inclusive, on the E scale of the Guilford
Zimmerman Temperament Survey are described, for the purpose
of this experiment, as being emotionally stable.
Emotional Instability
Those persons obtaining a C-score ranging from zero
to three, inclusive, on the E scale of the Guilford
Zimmerman Temperament Survey are considered, for the pur
pose of this study, to be emotionally unstable.
34
Approximately 400 students enrolled in introductcry
courses in psychology at the University of Southern Cali
fornia were given the Guilford-Zimmerman Temperament Survey
(12) during regular class hours. The students were not
necessarily psychology majors since this course is part of
the General Studies program in the university and every
liberal arts student is required to take four of the six
courses included in the General Studies program. The test
was administered in routine fashion as part of the course
work in learning about psychological tests. Exact figures
are not available, but it was estimated that approximately
one-fourth of the students fell in the higher C-score
range and about o e-fifth in the lower range. All students
who were asked to take pa~t in the experiment agreed to do
so. However, not all students falling in the two cate
gories were asked. An undetermined number was not con
tacted due to absences or "drop-outs·" from class. A
maximum age limit of twenty years was arbitrarily set. It
was believed that older students, being more sophisticated,
would be less motivated and less likely to become ego
involved in the experimental problem and, would therefore,
experience little or no frustration and anxiety.
A total of 102 persons were scheduled to serve as
subjects, fifty-five of whom were classified stable and
35
forty-seven unstable. Of these, seventeen failed to keep
their appointments for unknown reasons; thirty EEG records
were not usable due to evidence of pathology, or to arti
facts produced mainly from interference from the university
broadcasting station; five EEG's were composed almost
entirely of rapid, low amplitude, waves and could not be
used; the remaining fifty EEG's, twenty-five from the
stable group and twenty-five from the unstable, were used
as the source of the experimental data. The age range of
the two groups was seventeen to twenty years. There were
fourteen males and eleven females in the high-scoring
group and fifteen males and ten females in the low-scoring.
A breakdown of the total number of persons contacted is
shown in Table 1.
Materials
The Guilford-Zimmerman Temperament Survey
This test is a revision and consolidation of the
previously published Guilford series of tactored personal
ity inventories. It consists of 300 items which are
"y "
11
">" "N" ft h
answered by encircling the es, ., or o a er eac
statement. Scores are obtainable on ten relatively inde
pendent traits which have been isolated by factor-analytic
Table 1
Distribution of Source of
Available Subjects
Stable Unstable
Males Females Males Females Totals
Appointments not kept
9
2 4 2
17
EEG's not used due to:
Artifacts and
pathology 6 8
7 9
30
Absence of alpha
waves 2
3
0 0
5
Experimental subjects 14 11
15 10 50
Totals 31
24 26 21 102
37
procedures. The ten factors are: General Activity,
Restraint, Ascendance, Sociability, Emotional Stability,
Objectivity, Friendliness, Thoughtfulness, Personal Rela
tions, and Masculinity.
Although the entire test was administered to the
population sample, for the purpose of this experiment,
interest was centered only in the scores on trait E, Emo
tional Stability, since this was the basis for the selec
tion of the experimental subjects. Therefore, only this
scale will be described more fully. Factors D, Depression,
and C, Cycloid Disposition, from the Inventory of Factors
STDCR were combined for form factor E in the present survey.
Scores on this scale are interpreted as indicating in a
general way the degree of emotional stability or "neurotic
tendency.n A high score is supposed to reflect emotional
stability with a cheerful and optimistic disposition. A
low score, on the other hand, indicates opposite attitudes,
instability, and a predisposition toward neuroticism. The
following items are representative of those included in
the Survey. "You often find it difficult to go to sleep at
night because you keep thinking of what happened during the
day." "It takes a lot to get you emotionally stirred up or
excited." "You often feel guilty without a very good
t " reason for 1.
"Your mood often changes from happiness
38
to sadness, or vice versa, without your knowing why."
"You sometimes feel 'just miserable' for no good reason
at all."
The Stimulus
The stimulus was a twenty-eight second, high fidel
ity recording of a. male and female voice reading different
passages concurrently and at approximately equal intensi
ties. Using this technique, Murphy (27) successfully
differentiated the EEG responses of stutterers and non
stutterers. The content presented by the female voice was
considered relatively abstract while the passage read by
the male was evaluated as more concrete in meaning. Follow
ing is a reproduction of the two passages:
(Female) "Truth about matter-of-fact 1s never an
absolute certainty. It is not an imitation or copy of what
happens. The word 'truth' stands for the fact that a
given human hypothesis can illuminate the relationships
between certain events more than any other. A particular
truth is a relationship between a certain human statement
and a specific set of events. It enables the individual
to meet the least disappointment as he acts in accordance
with his statement. We say that all material truth is
probable as opposed to certain."
39
(Male) "I splashed through the slimey swamp,
shivering. The cold wind and rain beat against me, my
lungs ached, and my muscles felt like steel knots, but I
ran faster as I heard the barking of the wild dogs in the
distance. Streaks of dark, red blood oozed down my arms,
and my hands felt numb. Red and yellow flashes were
bursting across the sky when a sharp pain pierced my head.
I cried out; then everything went black."
The experimental subjects were instructed to pay
attention only to the ~emale voice and to attempt to
ignore any other sounds. At the end of the recording they
would be asked to reproduce as much of the female passage
as possible. In order to increase its effectiveness as a
distracting or frustrating agent, the male passage was
calculated to be more stimulating in terms of imagery and
feeling on the basis of its content, slightly greater
intensity, and wider pitch variation, than the female
excerpt which was intended to be more theoretical, less
stimulating, and more monotonous in pitch range . During
the recording of the stimulus record, both speakers spoke
into the same microphone and maintained the volume of the
voices at a relatively constant level by using monitoring
device. Murphy also points out that because" ••• low
frequency sounds are much more disruptive than those of
40
high frequency ..• the male voice served to enhance
further a masking, distracting, and frustrating effect
upon the subjects." In developing the stimulus record
Murphy tried out several other passages on a number of
people. These sub jects found it extremely difficult to
identify more than ten per cent of the total of 87 words
in the female passage and reported that they felt emo
tionally uncomfortable while attempting to concentrate.
In describing their feelings, they used such words as
"irritated," "bothered," "upset, '
1
"confused," "exaspera
ted," "frustrated," and "anxious."
The EEG and Recording Equipment
The EEG records were made on an Offner, type D,
six channel electroencephalograph with an attached ink
writing dynograph. The recording speed of the dynograph
was 2.5 centimeters per second. Silver composition elec
trodes, about 0.5 mm. in diameter, were used as leads.
The stimulus record was a reproduction of the one
originally developed by Murphy (27), and was re-recorded
by a professional recording studio, The Studio and Artists
Recorders, Hollywood, California. In order to maintain
the experimental stimulus as constant as possible, a suffi
cient number of records were prepared to insure that no
41
one would be played more than 15 times. They were played
on a California Model 6A Transcription Player phonograph
and reached the subjects by a loudspeaker placed approxi
mately five feet above and behind the subjects' head.
All EEG and stimulus equipment were in one room,
while the subject was in an adjoining room which was
electrically shielded and grounded. After the electrode
placements, the door between the two rooms was closed
leaving the subject's room in total darkness. Necessary
communication was conducted through a sliding panel in the
door separating the two rooms.
During the experiment, the subject was seated in a
reclining position on a cloth-covered easy-chair with his
arms placed on the armrests and his feet on a hassock.
Procedure
_ Q.rientation of the Subjects
One of the requirements of the experimental design
was to take EEG readings while the subject was in a re
laxed state, free from peripheral stimulation. Consequent
ly, it was considered highly important to keep at a mini
mum any anxiety or apprehension the subject might have
concerning the role he was to play in the experiment. As
42
one attempt to accomplish this, the scheduled appearances
of the students at the Speech and Hearing Clinic, where
the experiment was being conducted, were set sufficiently
far apart in time so that no one student would be kept
waiting. At least forty-five minutes were allowed for
each subject and this proved to be more than enough since
the usual amount of time required for the electrode connec
tions and the experiment proper was about twenty minutes.
Allowing plenty of time between appointments offered other
advantages in that it permitted the experimenter to work
unhurriedly, to be more casual, and to take as much time
as seemed necessary to make the student feel comfortable.
As the student reported to the clinic, he was met
by the experimenter and escorted to the room containing
the EEG equipment. The electroencephalograph was turned
on and he was given a ''dry run" demonstration while the
experimenter explained the various operations. He was
then taken into the adjoining room where he was seated on
the chair he was to occupy durin the experiment. Addi
tional information regarding the experimental procedures
was given, but no mention was made of the stimulus record
ing. Each subject was encouraged to express any anxiety
he migh~ be feeling with respect to the experiment. Where
no apprehension was evident, casual conversation was
43
carried on during the placement of the electrodes. A
frequently expressed fear was one of being shocked by
electricity or of being exposed to some uncomfortable ex
perience. The subject was reassured that nothing trauma
tic or embarrassing would happen to him, and was instructed
to sit back comfortably on the chair in a semi-reclining
position with his eyes closed gently as if preparing to
sleep. His hands were placed on the armrests of the chair
and the necessity of avoiding movement as much as possible
due to the effect of muscle movement on the record was
explained. The door between the two rooms was closed,
leaving the subject in darkness. This was the condition
of rest during which the first part of the EEG record was
then taken.
Electrode Placement
Recordings from four cortical areas were obtained,
using silver composition electrodes, 0.5 mm. in diameter,
which were held in place by a head band. Two electrodes
were placed about 6.o centimeters above each supraorbital
ridge and approximately the same distance apart. These
readings were not part of the experimental design and no
predictions regarding them were made. The two significant
electrode placements were connected to the left and right
44
occipital areas, approximately 2.5 centimeters above and
3.0 centimeters laterally from the inion. In each place
ment area the hair was clipped at the scalp and cleaned
with 2lcohol. Bentonite paste was used to increase the
efficiency of the contact between the electrode and the
scalp. Two "indifferent" electrodes were place on each
ear lobe to produce a monopolar recording of the left and
right occiput. Alcohol and Bentonite paste were also used
on the ear lobe connections. Each electrode connection
was tested for efficiency by measuring the electrical re
sistance which was kept between five and ten thousand
ohms.
The Three Experimental Conditions
Following the placement of the electrodes, the
door between the two rooms was closed, leaving the subject
alone, and the recording of the electrical potentials
began. As described above, the subject remained on a
chair in a reclining position with his feet on a hassock
and his eyes closed during this period. The recording in
the resting state continued until the experimenter had
ascertained by inspection that an EEG containing the maxi
mum number of alpha waves for that subject had probably
been secured. This period varied between two and ten
45
minutes. The stimulus period was then introduced.
As soon ~s the resting EEG was obtained, the slid
ing panel on the door between the two rooms was opened
and additional instructions were read to the subject.
These instructions followed closely those used by Murphy
(27) in his investigation but were slightly modified to
create a stronger feeling of threat, thus increasing the
anxiety-provoking value of the stimulus. The instructiora
were as follows:
"This is a test. You are about to listen to a
recording. Please listen carefully to what the woman is
saying. Try to disregard any other sounds you might hear
and listen to the woman only. You will be asked to repeat
all you can remember of the meaning of what the woman says.
Other students have taken this test and have done very
well. Your performance will be compared with theirs. Do
you understand? Do as well as you can."
When the subject made it clear he understood, he
was again requested to keep his eyes closed until notified
to open them. The stimulus was then presented and the
experimenter indicated on the dynograph tracings the exact
moment of the presentation. The duration of the stimulus
was twenty-eight seconds.
46
There were several elements in the set of instruc
tions calculated to create anxiety. In the first place,
each subject was led to believe, without directly being
told, that he was required to do nothing except sit quietly
while an electrical record of his brain waves was being
secured. The sudden confrontation with a statement that
he is going to be tested, in view of the long process of
conditioning in school, can reasonably be expected to evoke
some reaction of anxiety from beginning college students.
Being told that his performance would be compared with
others was also considered to be anxiety-provoking in most
instances. As explained above, it is extremely difficult
to understand more than a very few words in the passage
read by the woman. However, the student obviously had no
prior knowledge of this. Being misled to believe that
others had done very well was designed to evoke in the
student, while attending to the almost impossible task,
feelings of anxiety and loss of self-esteem under the
erroneous belief that he had failed where others had done
so well.
Upon the termination of the stimulus, the experi
menter again marked the EEG indicating the end of the
stimulus period and the beginning of the recovery period.
As noted previously, the brain waves continued to be
47
recorded for a period equal to the twenty-eight second
interval required for the presentation of the stimulus.
Durin the period of recovery the subject continued to
sit quietly with his eyes closed.
After the EEG for the three periods was taken, each
student was asked whether he had had previous information
regarding the details of the experiment. All denied prior
knowledge . Some admitted having asked other students who
had served as subjects about the experimental procedures
but indicated no mention had been made of any stimulus or
problem presentation. Each subject was then informed of
the importance of keeping the experimental conditions
constant and his cooperation was solicited. He was urged
not t o reveal to other students the fact that a stimulus
was used in the experiment.
Treatment of the Data
Electroencephalograms on fifty subjects were taken
under conditions of rest, stimulation, and recovery. The
duration of each condition was twenty-eight seconds. The
number of alpha waves was determined for each duration,
but only those alphas having a frequency of eight to
twelve per second and appearing in sequences of two or
more were counted . Amplitude was disregarded, but each
48
wave had to be clearly visible and sinusoidal in form.
Thus, the baste data available for statistical analysis
were the alpha wave counts secured from each of the fifty
subjects during the three experimental conditions, making
a total of 150, twenty-eight second intervals of recorded
cortical activity.
The twenty-eight second interval of the resting
EEG selected for wave count analysis customarily was the
interval immediately preceding the reading of the instruc
tions. However, where earlier tracings appeared by
inspection to have a greater predominance of alphas, an
earlier interval was selected. Although recordings from
both occipital lobes were obtained, it had been planne d
to use only the record from one hemisphere for quantita
tive evaluation. The record from the left lobe was
arbitrarily chosen, but exceptions were made where it was
marred by pathology or artifacts and the recording from
the right hemisphere was clear.
The wave count data were treated statistically to
determine whether significant difference s existed between
the stable and the unstable groups in each of the three
experimental conditions. The a ppropriate means, standard
deviations, and standard errors of the means were
calculated, using the alpha wave count data, and the
49
statistically significant differences were determined by
means of !-ratios (11).
The statis t ical results will be presented briefly
in Chapter III, ·but will be discussed more fully in
Chapter IV.
-
CHAPTER III
Results
It was postulated that emotional integration was
related to the synchronized, rhythmical pattern of the
alpha wave and that the greater the emotional disorganiza
tion, the greater would be the disruption of the alpha
wave. Predictions were made regarding the cortical be
havior of two groups of subjects under three different
experimental conditions. The experimental hypotheses
were to be investigated by comparing the mean alpha wave
count of the two groups under the specified conditions.
To eliminate the possibility of bias in the analy
sis of the EEG records, all identifying marks from both
sets of records were removed and replaced with the numbers
from one to fifty by a disinterested person. The records
were then mixed and subsequently analyzed in random
fashion. Although the process of counting individual
waves naturally requires subjective judgment at times, it
was presumed that such subjectivity would be consistently
applied to the analysis of all records thus cancelling out
its cumulative effects in any one direction.
The results are presented to show the basic alpha
wave count of the two groups during each of the three
-
. 51
experimental conditions. The group means are then com-
-- -
pared to determine the differences between the two groups .
. I
The significance of the differences was determined by the
use oft-ratio and was interpreted in terms of the prob-
-
ability of a difference of the obtained magnitude occurring
by chance.
Table 2 shows the distribution, means, and standard
deviations of the alpha wave count of all emotionally
stable subjects during each of the three experimental
conditions. Similarly, Table 3 shows the distribution,
means, and standard deviations for the unstable group. It
will be noted that in all conditions considerable variabil
ity in alpha wave count exists among the subjects within
both groups. Wide individual differences in the number of
alpha waves in the EEG's of normal persons is a character
istic that has been well established. Inter-individual
comparison betwAen groups also shows a great deal of over
lapping of scores. 1he scores of the subjects within each
.
group are neither consistently higher nor consistently
lower than the scores of the second group.
• f
52
Table 2
.
Distrtbution,. Means, and Standard Deviations of
' ~
Subjects
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Means
S.D.
Alpha wave ~bunt .f6r Each of Three Experi-
mental Conditions of Stable Group
Resting
263
260
256
208
214
282
219
225
287
222
226
223
263
186
227
274
174
159
200
205
259
166
172
232
240
225 .68
35 .91
Stimulus
276
137
225
233
263
268
180
179
273
102
92
218
274
234
161
268
82
45
189
111
256
131
126
235
147
188 . 20
21 . 97
Recovery
282
154
270
201
241
276
215
203
271
175
125
202
268
202
218
264
70
87
171
126
270
132
158
230
183
199 . 76
19.00
Table 3
Distribution, Means, and Standard Deviations of
Alpha Wave Count for Each of Three Experi
mental Conditions of Unstable Group
53
Subjects Resting Stimulus Recovery
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Means
S.D.
243
185
281
181
197
234
246
214
178
203
192
136
217
219
250
190
256
213
241
244
268
293
224
168
298
222.84
39.36
109
140
231
113
170
104
90
161
lo6
123
97
28
76
72
239
134
269
59
218
107
130
258
149
38
238
210
165
227
132
201
192
178
142
83
179
119
45
187
78
245
142
271
73
221
146
232
293
215
35
227
169.52
21.67
Inter-Group Comparisons
Experimental Condition Resting
The significant data used in comparing the two
groups during the condition of rest are presented in
54
Table 4. Although the two groups had been differentiated
on the basis of the variable, emotional stability, they
were both considered to be representative of a normal
population. It was hypothesized that the two groups would
behave cortically in similar fashion during a resting
state conducive to securing a maximal wave count. Inter
group comparison of the means discloses a small difference,
but it i~ not statistically significant. This finding
supports the prediction that the stable would not be dif
ferentiated from the unstable in terms of the number of
alpha waves secured during a relaxed, stimulus-free condi
tion.
Experimental Condition Stimulus
Table 5 shows the data used in comparing the stable
and unstable during the condition of stimulus presentation .
The unstable, it will be noted, reacted to the stimulus
with a mean alpha wave count lower than the mean of the
stable. Comparison of the scores shows a difference
55
Table 4
Significant Differences in Alpha Wave Count During
Condition Resting : Stable vs. Unstable
Group
liable
Unstable
Mean
225.68
222 . 84
S.D.
35.91
39.36
2 . 84
t
0.26
p
NS
56
Table 5
Significant Differences in Alpha Wave Count During
Condition Stimulus: Stable vs. Unstable
Group
Stable
Unstable
Mean
188.20
138.36
S.D.
21. 97
21.34
t
p
49.84
7. 99
. 01
57
between the means of the two groups which is significant
beyond the one per cent level of confidence. This result
bears out the prediction made regarding the difference in
response of the two groups to a disturbing stimulus, and
tends to support the second hypothesis formulated. In
view of the extent of the difference found, it can be
said with a great deal of confidence that the stable are
highly differentiated from the unstable in terms of the
number of alpha waves present during the presentation of
the disturbing auditory stimulus.
Experimental Condition Recovery
The significance of the difference between the
stable and unstable subjects during the recovery period
are found in Table 6. It will be seen that a difference
exists between the two groups with the unstable having a
lower mean alpha wave count. Determination of the signi
ficance of the difference gives a !-ratio exceeding the
one per cent level of significance. The unstable thus
continue to show to a statistically significant degree a
greater disruption of the alpha wave pattern. This result
tends to support the third hypothesis which predicted a
difference between the two groups with the stable mani
festing less disturbance as represented by the higher
58
Table 6
Significant Differences i Alpha vave Count During
Condition Recovery: Sta le vs . Unstable
Group
Stable
Unstable
Mean
199 . 76
169 . 52
S.D .
19 . 0
21 .67
t
p
30 . 24 5 . 14 . 01
59
.
mean alph~ wave count .
Intra-Group Comparisons
Although the three hypotheses the experiment pro posed to test were supported by the results of the om parison between th~ two groups in th tree experi ental
conditiois, it was decided to explore more fully th
effects of the three conditions within each group . It
also occurred to he experimenter that the difference
foun between the two groups during Condition ecovery
could be due primarily to a larger difference found dur ing Condition Stimulus and not to a faster rate of recov ery . While no predictions were made regarding the recov ery rate of the two groups , it was believed hat intra
group comparison in the different experimental conditions
would make the differences found between the two groups
more meaningful.
Co dition Resting vs . Co dition Stimulus
The data for comparing the d ff rence between the
means of Condit on esting and Condition tih~lus for
oth he stabl and
pari on o he stale
stable s found in Table 7. Com mea wave count in the two
60
Table 7
Intra-Group Comparisons of Sta ble and Unstable:
Condition Resting vs. Condition Stimulus
Cond. Rest . Cond. Stirn .
Group Mean S.D. Mean S .D. M1-M
2
Stable 225 . 68 35 .91 188.20 21.97 37.48
Unstable 222 .84 39 . 36 138 . 36 21.34 84 .48
t p
4.35 .01
9. 25 . 01
61
conditions indicates a very significant difference between
the two. Similar comparison of the unstable also indicates
a difference between the means of the two conditions beyond
-
the one per cent level of confidence . Accordingly, it can
be stated with considerable confidence that the stimulus is
an effective agent in disrupting the alpha wave pattern,
not only of the unstable, but of the stable as well.
Condition Resting vs. Condition Recover~
Table 8 shows the means, standard deviations, dif
ference between the means, and significance of the differ
ence between the resting condition and the recovery period
for the stable and the unstable . For both groups the mean
alpha wave count in Condition Resting is greater than the
mean in Condition Recovery to a degree that is statisti
cally very significant. In other words, the cortical
activity taking place during recovery from stimulation was
markedly different from the behavior prior to the presenta
tion of the stimulus. Suppression of the alpha waves
continued to occur during the period of recovery to a de
gree that differentiated it from the condition of optimal
alpha activity. Neither group recovered its cortical
equilibrium during the recovery period to an extent com
parable to the synchronization found in the resting
62
Table 8
Intra-Group Comparisons of Stable and Unstable:
Condition Resting vs. Condition Recovery
Cond. Rest. Cond. Recov.
Group Mean S.D. Mean S.D. M1-M
2
Stable 225.68 35 . 91 1 9.76 19.0 25 . 92
Unstable 222.84 39.36 169.52 21 .67 53 . 32
t p
3 .13 . 01
5.81 .01
condition.
Qondition Stimulus vs. Co dition Recovery
Table 9 provides the means, standard deviations,
difference between the means, and the significance of the
difference between the means of Condition Stimulus and
Condition Recovery for both the stable and unstable groups.
While comparison of the mean scores of the stable for the
two conditions reveals a difference significant at the ten
per cent level, the difference for the unstable reaches the
one per cent level. Another way of stating this is, that
in comparison with the amount of alpha wave breakdown dur
ing the stimulus period, the unstable made a very signifi
cant gain in alpha wave count during Condition Recovery.
The stable, on the other hand, also made some gain, but
not to the same extent as the unstable. The significance
of this finding will be discussed later.
64
Table 9
Intra-Group Comparisons of Stable and Unstablo:
Condition Stimulus vs. Condition Recovery
Group
Cond. Stirn.
Mean S.D.
Cond. Recov.
Mean S.D. M1-M2
Stable 188.20 21. 97 199.76 1 .0 11.56
Unstable 138.36 21.34 169 .52 21.67 31.16
t
p
1. 96 .10
5.03 .01
CHAPTER IV
Discussion
On the assumption that resistance of the alpha wave
pattern to disturbing stimuli is related to emotional in
tegration, as previously stated, three hypotheses were
formulated regarding the effect of experimentally induced
anxiety on the alpha rhythms of the stable and unstable
subjects. The results obtained support the three hypothe
ses and indicate to a very significant degree that anxiety
has a very pronounced disrupting effect on the alpha wave
pattern. The predictions made regarding the cortical
behavior of the two groups during each of the three experi
mental conditions are consistent with the findings.
Although the EEG of a given individual is relatively
constant, it is modifiable by internal and external stimuli.
At any given time, therefore, it represents a record of a
temporary cortical responsiveness to the surrounding condi
tions rather than a record of a permanent state of cortical
organization. With this consideration in mind, the inves
tigation was planned to determine whether experimentally
induced anxiety influenced the organization of the electri
cal discharges from the brain typically obtained during a
66
relaxed condition with stimulation held to a minimum.
Recognizing that the process of taking the electrical
recording itself could be anxiety provoking, especially
to the emotionally unstable, special effort was made to
allay anxiety and discomfort in order to hold constant,
as much as possible, the conditions under which the first
series of recordings was secured. Reference to Table 4,
page 55, indicates that this objective was achieved. W ith
respect to t he amount of alpha activity during a resting
condition, there is no significant difference between the
stable and unstable subjects. In other words, while re
laxed and undisturbed, the cortical behavior of the two
groups was similar.
The results shown in Table 5, page 56, clearly
differentiate the two groups in their reactions to the
disturbing stimulus. These results are in keeping with
the prediction that the unstable would experience, during
Condition Stimulus, a greater threat and loss of emotional
equilibrium which would be reflected in a significantly
greater decrease of cortical synchrony as measured by the
alpha wave count. It thus appears that the readiness of
the emotionally unstable to react to anxiety with a loss
of integrated behavior probably has a clos e correspondence
to the susceptibility of their rhythmical brain wave
67
patterns to break down during stress into more rapid and
irregular waves of lower amplitude. Conversely, analysis
of the electroencephalograms of the stable revealed a
pattern of synchronous activity which in comparison was
much more resistive to the disrupting impact of the
stimulus. Again, the greater stability of the cortical
organization can be compared to the greater capacity of
the more stable indivi dual to withstand the disorganizing
effects of anxiety.
The difference found be tween the t wo groups dur ing
Condition Stimulus is made more meaningful by inspection
of Table 7, page 60. The results show that the alpha
waves of both groups are markedly suppressed by the stim
ulus and demonstrate t he effectiveness of the stimulus as
a disturbing agent. The difference found in Condition
Stimulus, therefore, is not based on the greater cortical
reactivity of the unstable alone. The finding that both
the stable and unstable are affected by the stimulus to a
statistically significant degree, emphasizes the greater
disturbance found in the unstable. A similarity is seen
here to life situations where a well-adjusted person might
experience some fear and anxiety but an anxious person
would react with a much greater amount of anxiety and
disorganized behavior.
68
Considerations of characteristic modifications of
the EEG by sensory stimulation shows that the obtained
difference is not likely to be due to a difference in
reactivity to the auditory stimulus per se. While light,
briefly applied, e f fectively blocks the alpha waves
temporarily, continuous application of visual stimulation
for a.period of ten seconds or longer usually results in
adaptation and a return of the alpha rhythm (24). Sound
also blocks alpha but is less effective than light and
adaptation takes place more rapidly (24). An auditory
stimulus loses its effectiveness after a number of repeti
tions and adaptation to continuous auditory stimulation is
more rapid than to intermittent stimulation (2). Meaning
ful verbal material presented over a period of time not
only quickly produces adaptation, but results in an in
crease in alpha waves (35).
The results (Table 6, page 58) support the third
hypothesis which stated that during Condition Recovery the
stable would approach to a significantly greater degree
the optimal number of alpha waves obtained during Condition
Resting than would the unstable. This more highly
synchronized brain wave pattern of the stable is inter
preted as the greater capacity of the non-anxious person
in comparison with the anxious individual to recover his
69
optimal emotional equilibrium following a period of stress.
The more sustained cortical reactivity of the unstable
follow~ng termination of the disturbing stimulus is seen
as manifesting a greater carry-over effect of the anxiety
provoked by the stimulus. However, further analysis of
he results shows that this interpretation does not tell
the whole story. It will be noted that the difference
between the two groups is an absolute difference and not
a difference in rate of recovery . This difference could
well be primarily a function of the greater difference
between the stable and unstable found in Condition Stimu lus . Since the unstable were much more highly disturbed,
they naturally would have to travel a longer distance to
recover the opt imal wave count of Condition Resting .
Consequently, i t could have been possible for the unstable
to have made a relatively greater recovery during Condi
tion Recovery and still show a greater disturbance in
terms of a lower alpha wave count . Consideration of the
absolute difference in wave count alone, while sustaining
the hypothesis, might well hide the possibility of greater
gains by the unstable. Comparing the groups' cortical
reaction during Condition Recovery with each of the two
other conditions shows this actually to be the case.
70
Following cessation of the stimulus, neither group
attained the optimal wave count of the first period
(Table 8, page 62). Although the stable achieved a closer
approximation to the original wave pattern, because of
their lower reactivity to the stimulus, the difference be
t ween the conditions of rest and recovery was not complete
ly obliterated. The comparative influences of Condition
Stimulus and Coridition Recovery on the groups are seen in
Table 9, page 64 . It will be noted that while the unstable
made a recovery of alpha wave count, significant at the one
per cent l evel of confidence, the stable did no t show as
significant a change. Though giving no information regard
ing he comparative rate of recovery, where one group shows
a statistically significant change at the one per cent
lev 1 a nd the second group a change at the ten per cent
level , it seems logical to conclude that t he former made
a greater change.
This discovery of apparently greater gains by the
unstable prevent s a simple interpretation of the cortical
behavior of the two groups during the period of recovery
a nd indica t es a limitation of the experimental design.
omparison of t he wave counts of the two groups during
Cond tion Recovery states nothing of the rate or pattern
--
of the recovery or of its relationship to the amount of
,.,, 1
,_
disruption during the preceding period. Using the wave
count for the entire period averages out the changes
occurring during shorter intervals within the period and
does not describe adequately the process of recovery. To
get a clearer picture of the changes taking place a dif
ferent statistical approach is necessary. However, the
present study was limited to offering experimental evi
dence regarding a possible connecticn between anxiety and
the synchronized electrical discharges of the brain. To
answer the questions raised, a separate study planned to
investigate t he relationships between the amount of
anxiety and the processes of suppression and recovery of
the rhythmical wave pattern would be necessary.
A secondary finding, while not stated as part of
the major hypotheses, lends additional support to the
interpretation linking the pattern of cortical activity to
emotional integration. Clinically, it is known that some
persons under mild stress react in a more productive and
more highly organized fashion. Since emotionally unstable
individuals tend to show emotional disorganization in the
face of threat, it follows that it is the more stable who,
under stress, is the one more likely to become more alert
and ready for a superior performance. Analysis of the
individual subjects' cortical response t~ the three
72
experimen al conditions revealed seven instances where
a n increase in the number of alpha waves occurre follow
ing resenta ion of the disturbing stimulus . Of the seven
per sons who responded with an increase in alpha activity,
s i x belonged to the stable group and one to the unstable .
h i l e the small number of cases prevents meaningful sta
tistical treatment, t his di tribution of the cases respond
ing to stimulation with ·ncreased alpha activity is a
r nd consistent with the interpretation that emotional
stability is positively related to the stability of the
brain wave pattern.
It is interesting to note that Travis and Eagan
( 35) and Williams (41), in studies very diss milar to the
present investiga ion, reached conclusion which are in
harmony with the interpretation suggested here. Accept ing the fact that the usual cortical reaction to stimuli
r e sults in suppression of the alpha waves, these investi gators were interes ed in studying conditions that might
i nduce an increase in alpha activity following stimula-
t o . They concluded that "faci itation" of alpha
'-
activi ty can be induced by "increased mental activity"
and by stimuli creating 'general states of readiness,
attentiveness or heightened awareness . " Williams con
clud d that the same stimulus under different conditio s
73
can increase or decrease the alpha waves and suggested
"that blocking and facilitation are symptoms of two dif
ferent types of reaction," a conclusion and suggestion
which can also apply to t he findings of this study.
Prior to the initiation of the current study,
studies had appeared in the literature suggesting that
rapid, low-amplitude waves accompanied a state of anxiety,
but experimental evidence had been lacking. These fre
quently reported observations had been based on clinical
studies of individual cases (4) or on comparisons of EEG
records of neuropsychiatric patients with records of nor
m al persons taken under resting conditions (3, 6, 14).
The varied conditions under which empirical studies are
conducted make ready comparison and evaluation of results
difficul t . Since then, however, two experimental investi
gations of the relationship between anxiety and the EEG,
with which this study can be compared, have been reported.
In the Ulett and Gleser study (35) the EEG records
of both the anxious and non-anxious groups secured under
conditions of stress differed from their respective resting
recor sat t he five per cent level but no difference was
found between the two experimental groups . In comparison,
the find ings of the present study differentiated the
effects of the two condition of rest and threat at the
74
I
one per cent level and, also, differentiated the experi
mental groups on the basis of their responses during
stress. Despite the absence of identical results, the
trends are similar and the interpretations of Ulett and
Gleser are in accord with the present findings. They
concluded that the occipital rhythm induced by photic
stimulation is altered by emotional stress in a way simi
lar to the blocking of the alpha wave.
In a later, very similar experiment, Ulett, et al.
(38) obtained results which differentiated the non-anxious
group from the two anxious groups in the following factors:
(a) resting EEG's, with the anxious subjects having less
alpha and greater beta activity, (b) EEG response to the
photic stimulation, and (c) discomfort and dysphoria pro
duced by the photic stimulation as reported by the subjects.
These results are consistent with the present results in
demonstrating a relationship between threat or anxiety and
changes in the EEG. One of the findings, however, is not
in agreement, namely, the difference between the groups in
the resting electroencephalograms. The reason for this is
not clear, but a possible explanation may be offered.
Since the electroencephalography process, itself, may be
anxiety-provoking, and more so to an anxious or unstable
person, this investigator actively tried to allay anxiety
75
in his subjects during the resting period. Ulett, et al.
--·-
make no mention of ft similar effort on their part. The
lack of agreement in this phase of the experiments may
well be due to the difference in procedure.
While Sherman and Jost (33) were primarily inter
ested in investigating the relationship between frustra
tion and the brain waves, a comparison with their results
may be made with some justification. Using a group of
well-adjusted children and a group of neurotic children,
they subjected both groups to a frustrating learning
situation. Electroencephalograms were taken during a rest
ing period and during frustration, with the neurotic chil
dren showing less alpha wave activity in both periods. If
it can be assumed that the neurotic children experience
greater anxiety during both periods, and this does not seem
unreasonable, then it can be stated that the results of
both investigations are in agreemeut. The explanation
offered in the preceding paragraphs account for the incon
sistent results in the resting EEG's is applicable in this
instance also.
Studying the reactions of stutterers and non
stutterers, Murphy ( 27) found: (a) no significant diffeP
ence between stutterers and non-stutterers in the resting
EEG, (b) significant difference between stutterers and
76
non-stutterers in the EEG secured during presentation of
the stimulus with the stutterers showing greater break
down in alpha activity, and (c) significant difference
between stutterers and non-stutterers during the period
of recovery with the non-stutterers showing greater re
covery of optimal alpha activity. It will be seen that
these results are highly consistent with the results ob
tained in the comparison of the reactions of the stable
subjects in the present experiment. Murphy's interpreta
tion of the observed differences between the stutterers
and non-stutterers could be equally applicable to the
differences between the stable and unstable. He attributes
the greater reactivity of the stutterers to a difference in
the ncortical excitatory state " of the two groups. This
state is described as "a condition of the organism char
acterized by reactivity of the cortex to stimuli impressed
upon it" and associated "with conditions of emotional ten
s:L on or arousal." In other word s , the stutterer is an
emotionally unstable person whos e anxiety is readily
aroused. This conclusion, supported by the similarity of
the results of t he two experiments, should have theoreti
cal and practical implications for the speech pathologist.
-
77
Need for Further Research
The interpretation of the results of the comparison
between the stable and unstable during the period of re
covery, as previou~ly indicated, was complicated by possi
ble differences between the two groups in the rates of
recovery. The failure to account clearly for this was
at·~ributed to a limitation of the study. W hile the dis
ruptive effect of anxiety on the alpha pattern has been
demonstrated, within the limits of this study, no informa
tion is available regarding the quantitative relationship
between the two factors. Additional research is neede d to
determine the ratio between the severity of anxiety and
the rate of suppression and recovery of the alpha rhythm.
It is suggested that, in addition to the usual wave count
analysis, changes in the ampli t ude of the alpha waves also
be considered.
It has been point ed out that t he present r esul t s of
no difference between t he t w o experim en tal groups in a r e laxed, stimulus- f ree s tate is in contras t to the m ore fre quently reported r esul t s. M ost studies in the l iterature
have found difference s in the resting EEG's of anxious and
non-anxious or stable and unsta bl e individuals. It has
been sugges t ed that thi s lac k of a greement may have been
due to differ e nces in experime ntal procedure s or to a
78
difference in the subjects used. Another possibility
exists. It has been demonstrated (17, 31) that the alpha
wave is subject to conditioning. If anxiety induces a
suppression of alpha, then, by a process of conditioning,
it could be possible for a prolonged state of anxiety to
effect more or less permanent neurological changes pr -
ductive of fewer alpha and more beta waves. It would be
worthwhile to conduct an investigation for the purpose
of determining whether the EEG!s of chronically anxious
persons could be differentiated from those of persons
experiencing a temporary condition of anxiety. It may
be that, cortically, there are at least two types of
anxieties.
There is evidence suggesting the importance of
hereditary factors (8, 18, 22) in determining the EEG
pattern. The families of epileptics , of children with
behavior disorders and of psychopathic adults have been
found to have a larger percentage of abnormal records
than is found in a normal population. The instability of
the EEG, then, may be partially due to genetic influences.
Information is needed to determine whether, and to what
extent, the readiness to react to stress with a breakdown
of rhythmical wave pattern is a familial characteristic.
79
The determination of the relative influence of
functional, neurological, and hereditary factors in
anxiety could have important implications for the diagno
sis, treatment and prognosis of mental illnesses in which
anxiety was a major symptom.
The assumption has been made that disor; anized
behavior under anxiety-provoking situations is related to
the breakdown of the alpha wave pattern under similar
conditions. Experimental evidence, supporting or reject
ing this assumption, could be provided by devising a study
designed to record concurrently task performance and the
EEG under conditions stimulating feelings of threat or
anxiety.
While it has been shown that anxiety, as defined
in this experiment, very strikingly suppresses the alpha
wave pattern, it does not rule out the possibility that
other emotions may similarly modify the EEG. Research is
needed to determine whether, and in what way, other emo
tions may influence the electrical discharges from the
brain.
CHAPTE V
Summary and Conclusions
Problem
The purpose of this study was to investigate
whether a significant correlation exists between the emo
tional state ot anxiety and some characteristics of the
electroencephalogram. It is well known that under suffi
ciently intense emotional stress most people are likely
to react in a disorganized and ineffective manner. Also,
the greater the degree of emotional disorganization, the
longer is the period required for the recovery of psycho
lo5ical equilibrium. Though convincing experimental data
are lacking, leading workers in electroencephalography are
in general agreement that emotional excitation tends to
abolish the rhythmical pattern of the cortical potentials,
and that the alpha wave is frequently replaced by the
smaller and more rapid beta wave.
On the presupposition that emotional integration is
related to the synchronized, rhythmical pattern of the
alpha wave, predictions were made relative to the amount
of disruption of the alpha pattern of emotionally stable
and emotionally unstable sub ·ects under different
81
experimental conditions. These predictions were formulated
into three hypotheses: (a) there would be no significant
difference between the electroencephalograms of the two
groups obtained during a relaxed, stimulus-free state; (b)
the unstable would differ significantly from the stable in
response to an anxiety-provoking auditory stimulus; (c) the
unstable would differ significantly from the stable in the
extent of recovery of the wave pattern previously obtained
in the resting condition.
Method
The Guilford- Zimmerman Temperament Survey, a fac
torially-derived test of personality, was used to select
the subjects. The population from which the subjects were
chosen consisted of university undergraduate students en
rolled in introductory psychology courses. Those students
obtaining a C-score of seven, or above, on the E-Scale,
Emotional Stability, of the Guilford-Zimmerman, were
classified as emotionally stable and those obtaining a
c-score of three, or below, were classified as unstable.
There were twenty-five subjects in each experimental group,
fourteen males and eleven females in the stable a:1d fifteen
males and ten females in the unstable. The age range was
seventeen to twenty years.
82
~he EEG records were made on an Offner, type D,
six channel electroencephalograph with a n attached ink writing dynograph. Electrodes were attached to the left
and right occiput and two 'indifferent" electrodes were
placed on each ear lobe to produce a monopolar recording .
EEG's were obtained during three experimental
conditions, each twenty-eight seconds in duration : (a)
a resting period with stimulation held to a minimum, (b)
presentation of the stimulu , and (c) a recovery period
following cessation of the stimulus .
The stimulus was a twenty-eight second recording
of two voices, male and female, reading different passages
concurrently. The reading by the female voice was intended
to be abstract and monotonous, while the males was calcu lated to be stimulating in terms of imagery and feeling
in order to increase its distracting value. Prior to the
presentation of the stimulus, the subjects were informed
they were to listen to a recording, and were instructed
to play attention to the female voice only and to disre-
gard _ other sounds. o mention was made of the male voice.
To increase motivation and feelings of anxiety, each per
son was told he would be tested, that others had done
well, an that his score would be compared with theirs.
82
The EEG records were made on an Offner, type D,
six channel electroencephalograph with an attached ink
writing dynograph. Electrodes were attached to the left
and right occiput and two "indifferent" electrodes were
placed on each ear lobe to produce a monopolar recording.
EEG' s were obtained during three experiment~· 1
conditions, each twenty-eight seconds in duration: (a)
a resting period with stimulation held to a minimum, (b)
presentation of the stimulus, and (c) a recovery period
following cessation of the stimulus.
The stimulus was a twenty-eight second recording
of two voices, male and female, reading different passages
concurrently. The reading by the female voice was intended
to be abstract and monotonous, while the males was calcu
lated to be stimulating in terms of imagery and feeling
in order to increase its distracting value. Prior to the
presentation of the stimulus, the subjects were informed
they were to listen to a recording, and were instructed
to play attention to the female voice only and to disre
gard other sounds. No mention was made of the male voice.
To increase motivation and feelings of anxiety, each per
son was told he would be tested, that others had done
well, and that his score would be compared with theirs.
I
83
esults
~he basis for statistical analyses was he alpha
wave count determined for each of the three experimen al
conditions on each subject . Only those alpha s having a
fre uency of eight to. twelve cycles per second and appear
ing in sequences of two or more were included in the w a ve
count . Appropriate means and standard dev ations were
calculated . Compartsons were made by computing the differ ences between the group means and the significance of the
differences as determined by the use oft -ratios .
-
Inter- group comparisons during the three experi mental conditions pro uced the followl g results :
1 . o difference between the stable and the
unstable in the resting EEG .
2 . Very significant differences between th
stable and the unstable during stimulation
with the unstable showing a lower wave count .
3. Very significant differences between the
stable and the unstable during recovery, with
the unstable showing a lower w a ve count .
Intra-group comparisons betwee the di erent
exper metal conditions obtained the fo l lowing results :
84
1. Condition Resting vs. Condition Stimulus
a. Within the stable group, very significant
differences with a lower wave count in
Condition Stimulus.
b. Within the unstable group, very significant
differences with a lower wave count in
Condition Stimulus.
2. Condition Resting vs. Condition ecovery
a. Within the stable group, very significant
differences with a lower wa e count in
Condition ecovery.
b. Within the unstable group, very significant
differences with a lower wave count in
Condition Recovery.
3. Condition Stimulus vs. Condition Recovery
a. Within the stable group, a difference
significant at the ten per cent level with
a lower wave count in Condition Stimulus .
b. r1thin the unstable group, a difference
significant at the one per cent level with
a lower wave count in Condition Stimulus.
85
Discussion
The findings of no difference between the stable
and the unstable during the resting condition are in
contrast to the frequently reported results of empirical
studies and to the results of similar, though not identi
cal, experiments of Sherman and Jost and Ulett, et al.
The conflicting results could be due to differences in
re~ording procedures or to differences in the basis for
selection of the subjects.
The results obtained during Condition Stimulus are
in agreement with those of Sherman and Jost, Ulett, et al.
and Murphy, who found significant differences between
their experimental groups with respect to the EEG's se
cured during controlled stimulation. Sherman and Jost
used "well-adjusted" and neurotic children in a frustrat
ing learning situation. Ulett, et al. compared the
reactions of anxiety-proned and non-anxiety-proned normals
and f'l.eurotic patients to photic stimulation. Using
stutterers and non-stutterers as subjects, Murphy obtained
results which are in very close agreement to those ob
tained in the present study.
The results of this experiment were interpreted as
supporting the assumption that emotional stability and
the stability of the EEG as manifested by the persistence
86
of the rhythmical alpha wave pattern despite disturbing
stimuli have a ommon n urological basis .
Conclusions
The general conclusio s, within the limits of this
study , were : (a) that emotionally unstable persons do
not differ, with respect to the resting EEG, from emotion
ally stable individuals, (b) that under anxiety-provoking
conditions, unstable individuals differ significantly
from stable persons in terms of alpha rhythm breakdown,
(c) that the unstable continue to show significantly
greater disturbance than the stable following cessation
of the stimulus, (d) that additional research is needed
to determine the relationship between the degree of emo tional disturbance and the rate of suppressi n and
recovery of the optimal alpha wave count .
Some suggestions for further research were offered :
1 . The quantitative relationship between anxiety
and the rate of suppr ession and recovery of
the alpha rhythm .
2. Neurological changes resulting in predominance
of beta waves caused by chronic states of
anxiety by a process of conditioning .
3. The influence of hereditary factors in the
EEG's of emotionally unstable persons .
4. The effects of other emotions on the EEG .
87
5. Concurrent recording of task performance and
the EEG under condition of threat to test the
assumption that disorganized behavior during
stress is related to t he alpha wave breakdow n
during similar conditions .
E F E R E C E S
References
1. Adrian, E . D. & M atthews, B H. C. The Berger
rhythm: Potential changes from the occi ital
lobes of man. Brain, 1934, 57, 355- 385 .
-
2 . gchi, . K. The adaptation an variability of
respons of the human brain rhythm . Psychol.
Bull., 1936, TI, 771-772 .
3 . Brazier, M . A .. , Finesinger, J . E . , Cobb, s .
contrast between the electroen~ephalograms of 100
psychoneurotic patients and those of 500 adults .
Amer . I · Psychiat., 1945, 101, 443-448 .
4 . Cohn, R. The in luence of emotions on the human
electroencephalogram . ~- nerv . ment. Dis., 1946,
104, 351 - 357 -
5 . Darrow, C. v . , Jost, H. Solomon, A. P., & Mergener,
J . C. Autonomic indicatiors of excitatory and
homeostatic effects o the electroencephalogram .
l· Psychol . , 1942, _ll, 115-1 30 .
6 . Finley, K. H. On the occurrence of rapid frequency
potential changes in the human electroencephalo
gram . Amer . ~ - Psychia~ . , 1944 , 101, 194-200 .
7.
8 .
9 .
10 .
r Pud , S . The problem of anxiet~.
orton, 1936 .
ew York: J. w.
Gottlieb, J . S . , Knott, J . . , Ashby, M . C. Studies
i n primary behavior disorder and psyc opathic
personality : II. The inheritance of electro cortical activity . Amer . J. P ychiat . , 1947, 103,
823- 827 . -
Gittlober, A. The relationship between brain ooten
tials a~d personality . J . exp . Psychol ., i938,
22, 67- 74 . -
-
Grinker, R .. Method for studying and 1 fluencing
cortico- hypothalamic relations . Science , 1938 ,
§1~ 73-74 .
90
11. Guilford, j - . P. Fundamental stati.stics in psycholo~y
and education. ew York: McG aw-Hill, 1950.
12. Guilford, J.P., & Zimmerman, W . S . The Guilford
Zimmerman temperament survey. Beverly Hills,
alif.: Sheridan Supply Co., 1949.
13. Henry, C. E., & Knott , J. R. A note on the relation
ship between "personality" and the alpha rhythm
of the electroencephalogram . J. exp. Psychol.,
1941, _g_§, 362-366.
14. Heppenstall, M., Hill, D., & Slater, E . The EEG i
the pro nosis of war neuroses . Brain, 1945, 68,
17-25.
15. Hoagland, H., Cameron, D .. Rubin, M .A., &
Tegelberg, J. J. Emotion in man as tested by
the delta index of the electroencephalogram :
II: Simul taneo1.1s records from the cortex and
from region near the hypothala us. J. gen.
Psychol., 1938, 1.2_, 247-261. -
16. Jasper, H. H. Electrical signs of cortical activity.
Psychol. Bull., 1937, 34, 411-481.
17. Jasper, H. H., & Shagass, C. Conditioning the occi
pital rhythm in man. l· exp. Psychol., 941 ,
28, 373-388 .
-
18. Kennard, M.A. Inheritance of electroencephalogram
patterns in children with behavior disorders.
Psychosom. Med., 1949, 11, 151-157.
-
19. Kennard, M.A. The EEG in psychological disorders.
y~chosom. Med., 1953, 15, 95-115.
20 . Lazarus, R. s ., Deese, J., Osler, S . F . The ef
fects of psychological stress upon performance.
Pslchol . Bull ., 1952, 49 , 293-317.
21 Lemere, F. The significance of individual differencEB
i the Berger rhythm. Brain, 1936, 59, 366- 375 .
-
22 . Lennox, . G., Gibbs E . L., & Gibbs, F . A. The
inheritance of epilepsy as revealed by the
elec roencephalograph. J. Amer. med. Ass.,
-
1939, 113, 1002-1003.
91
23 . Lindsley, D. B. El ct ical potentials of the brain
in children and adults . I · gen . Psychol . , 1938,
.!.2, 285-306 .
24 . Lindsley, D. B. lectroencephalography . In J . McV .
Hunt (~d . ), Personality and the behavior dis orders . ew York : onald , 1944 . Vol . II ,
pp . 1033-1103 .
25. Li dsley D. B. Emotions and the electroencephalo gram . In M . L. eymert (Ed . ), Feelings a nd Emo tions : The Mooseheart sym~osium. New York :
McGraw-Hill ,1950 . Pp. 23 -246 .
26. May, R. The meaning of anxiety . New York: Ronald,
1950.
27. Murphy, A. T. An e l ectroencephalographic study of
frustration in stutterers . Unpublished doctor's
dissertation, Univer . of Southern alifornia ,
1952 .
28. ubin, s., & Bowman, K. M . Electroencephalic and
personality correlates in peptic ulcer .
Psychosom . Med ., 1942, ±, 309- 318 .
29 . Saul, L. J., Davis, H. , & Davis, P . A. Correlations
between Electroencephalograms and the psychologi cai organizat on of the individual. Trans . A mer .
neurol. Ass., 1937, 63, 167- 169 .
30 . Saul, L. J., Davis, H. , & Davis, P . A. Psychologic
correlations with the electroencephalogram .
Psychosom. Med., 1949, 11, 361-376 .
31 . Shagass, c. Conditioning the human occipital alpha
rhythm to a voluntary stimulus : A quantitative
study. l · exp . Psychol . , 1942, 31, 367- 379 .
32.
Sherman, M . The frustration threshold. Amer . J .
Psychiat . , 1947, 104, 242 - 246 .
Sherman, M . , & Jost , H. Frustration reactio s of
normal and neurotic persons . J . Psychol ., 1942,
lJ., 3-19 .
34 .
35.
36 .
37 .
38 .
39 .
40.
41 .
42.
92
Travis, L. E . Brain potentials and the temporal
course of consciousn ss. ~- exp . Pslchol ., 1937,
21, 302-309 .
-
Travis, L. E., & Egan, J .P. Increase in frequency
of alpha rhythm by verbal stimulation. J. exp.
Psychol., 1938, §, 384-393.
Travis, L. E., and Gottleber, A. How consistent are
an individual's brain potentials from ay to
day? Science, 1937, ~, 233-234 .
Ulett, G. A., & Gleser, G. Effect of experimental
stress upon the photically activated EEG .
Science, 1952, 115, 678-682 .
Ulett, G. A., Gleser, G., W inokur, G., & Lowler, A.
The EEG and reaction to photic stimulation as an
index to anxiety- proneness. EEG clin. neuro
physiol., 1953, 2, 23- 32.
alter, . 3. Critical review : The technique and
application of electroencephalography. J. Neural .
Psychiat., 1938, l, 359- 389 . -
alter, W. G., & !alter, V. J. The electrical
activity of the brain . Annu. Rev . Physiol .,
1949, _g, 199-230.
W illiams, A. C. Facilitation of the alpha rhythm of
the electroencephalogram. J. exp . Psychol .,
1940, 26, 413-422.
-
W illiams, D. The significance of an abnormal electro encephalogram. J . Neural . Psychiat ., 1941, i,
257-268 .
Zander, A. F . A study of experimental frustration .
Ps~chol. Monogr., 1944, 56, No . 3 ( hole No .
25 ).
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
The relationship of the pretreatment strength of social responses of schizophrenic patients to improvement with electric shock treatment
PDF
Sociological aspects of training programs in selected industrial organizations in the Los Angeles metropolitan area
PDF
The isolation of personality traits in the domain of military leadership
PDF
A factor-analytic study of the Minnesota multiphasic personality inventory using a transpose matrix (Q-technique)
PDF
Social functions of the churches in a changing community with special reference to social processes
PDF
Studies on cactus alkaloids and related synthesic [i.e. Synthetic] compounds
PDF
The effects of partial reinforcement on a whole human response
PDF
An analysis of the research literature used by American writers in the field of speech
PDF
Claus Spreckels of California
PDF
Clothing as a factor in the social status rating of men
PDF
An experimental study in fluctuations among successive play performances
PDF
The effectiveness of criminal justice sanction strategies in the deterrence of drug offenders
PDF
Women and the city, 1870-1920
PDF
The selection of textbooks and other instructional materials in the high schools of California
PDF
Social stratification as reflected in selected novels of Sinclair Lewis
PDF
W-B data as indices of aptitude for law and engineering a contribution to vocational counselling
PDF
The literary reputation of F. Scott Fitzgerald
PDF
An investigation of the chronic physiological changes due to relaxation training and practice
PDF
Action of DDT on two clones of a marine green alga Pyramimonas
PDF
Phineas Banning with special reference to the development of transportation in Southern California
Asset Metadata
Creator
Dunbar, Waldo
(author)
Core Title
The effects of anxiety on the alpha rhythm of the electroencephalogram
School
College of Letters, Arts and Sciences
Degree
Doctor of Philosophy
Degree Program
Psychology
Degree Conferral Date
1956-05
Publication Date
05/01/1956
Defense Date
01/01/1956
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
anxiety,electroencephalography,OAI-PMH Harvest
Format
theses
(aat)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-oUC112724843
Unique identifier
UC112724843
Identifier
A No.16 (Film) (call number),Ph.D. Ps '56 D899 (call number),etd-DunbarWaldo-1956.pdf (filename)
Legacy Identifier
etd-DunbarWaldo-1956
Document Type
Dissertation
Format
theses (aat)
Rights
Dunbar, Waldo
Internet Media Type
application/pdf
Type
texts
Source
20230207-usctheses-microfilm-box7
(batch),
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Repository Email
cisadmin@lib.usc.edu
Tags
anxiety
electroencephalography