University of Southern California Dissertations and Theses

The Visual Fusion Threshold (Vft) Test As A Measure Of Perceptual Efficiency In Kindergarten And First Grade, And As A Possible Predictor Of Later Reading Retardation

(USC Thesis Other)

The Visual Fusion Threshold (Vft) Test As A Measure Of Perceptual Efficiency In Kindergarten And First Grade, And As A Possible Predictor Of Later Reading Retardation

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THE VISUAL FUSION THRESHOLD (VFT) TEST
AS A MEASURE OF PERCEPTUAL EFFICIENCY
IN KINDERGARTEN AND FIRST GRADE,
AND AS A POSSIBLE PREDICTOR
OF LATER READING RETARDATION
by
John Wesley Howe
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
(Educational Psychology)
January 1963
UNIVERSITY O F SO U T H E R N CALIFORNIA
GRADUATE SCHOOL
UNIVERSITY PARK
LOS A NGELES 7. CALIFORNIA
This dissertation, written by
....... JOHN..WESLEY...HOWE.............
under the direction of h.i.s....Dissertation Com
mittee, and approved by all its members, has
been presented to and accepted by the Graduate
School, in partial fulfillment of requirements
for the degree of
D O C T O R O F P H I L O S O P H Y
Dean
Date JAHyARY.1263...............
DISSERTATION COMMITTEE
( l... y —
. Chairman
ACKNOWLEDGEMENTS
To the Board of Trustees of the Rosemead Elementary
School District gratitude is acknowledged for their en
couragement of research and their authorization of this
study within the District.
To Dr. Rodney S. Mahoney, Superintendent, and to
Dr. Arthur A. Attwell, Director of Guidance of the District
warmest thanks are given for their many kindnesses and
repeated aid with schedules, personnel and operations in
the District.
To the teachers, principals and school secretaries
appreciation is expressed for their interest and assistance
in testing, rating, and locating other data on their
children.
Acknowledgement is gratefully made to Dr. Harry W.
Smallenburg and the Division of Research and Guidance,
Los Angeles County Superintendent of Schools Office, for
the use of certain flicker-fusion equipment.
To Mrs. James E. Gibson and Mrs. Thomas E. Butler
thanks are extended for the care and secretarial skill
they devoted to the preparation of this manuscript.
Special thanks are also offered to the augmented
secretarial staff of Sue, Catherine, Janet, and John who
checked, re-checked, and checked again.
ii
TABLE OF CONTENTS
ACKNOWLEDGEMENTS
Page
ii
LIST OF TABLES v
LIST OF FIGURES
Chapter
I. THE PROBLEM 1
Importance of the Problem: Some Effects
and Dynamics
Definitions and Estimated Prevalence Rates
Conflicting Theories and Approaches
A Promising Study Based on the
Visual Fusion Threshold ( ;T T T)
Description and Definition of Visual
Fusion Threshold (VFT)
The Problem Attack of the Present Study
II. REVIEW OF THE LITERATURE.................. 2S
Selective Review of the Literature
on Reading Disability
Review of the Literature on the VFT
Summary of Pertinent Points in the
VFT Literature
III. PROCEDURES................................ 91
Population Selected
Testing Schedule and Numbers Tested
Selection of the Test Battery
The VFT Apparatus and Equipment
Procedures for Testing the VFT
Group-Testing Procedures for the PMA and SPRD
Teachers1 Ratings of Reading Ability
VI. RESULTS.................................. 119
Feasibility of the VFT Test with
Young Children
Validity and Reliability of the
VFT Scores in Young Children
VFT Scores as Predictors of
Other Variables
iii
iv
V. DISCUSSION................................... 143
Administration of the VFT Test
Validity of the VFT Results
VFT Scores as Predictors of
Other Variables
Comparison with the Study by Tait
Long-interval Reliability of the VFT
VI. SUMMARY, CONCLUSIONS AND RECOMMENDATIONS . . . 168
Summary
Conclusions
Recommendations
BIBLIOGRAPHY ....................................... 182
APPENDIX........................................... 189
GLOSSARY........................................... 193
ADDENDUM
198
LIST OF TABLES
Table Pagu
4.1 Means, standard deviations and standard
errors of the VFTs for kindergarten and
first grade classes...................... 124
4.2 Ranges, means, standard deviations and
standard errors of VFT scores on five
ascending (up) and five descending (down)
trials for boys and girls in kindergarten
and first grade.......................... 127
4*3 Ranges and means for the standard devia
tions of the ten VFT trials in each in
dividual case (i.e., the ipsative standard
deviations).............................. 130
4.4 Correlations of selected variables with
VFT scores............................... 133
4. $ — Groups rated by their first grade teachers
as high, medium or low in reading ability:
A Average months of mental a g e ............. 136
B Average scores on the first grade reading
t e s t .................................. 137
C Average scores on the third grade reading
t e s t .................................. 13$
D Average scores on the first grade VFT test 139
v
LIST OF FIGURES
Figure Page
2.1 Schematic illustration of relations
between frequency, intensity and
duration of light pulses and the
fusion threshold ........................ 63
3.1 Stroboscope.......................... 9$
3.2 Viewing-tube........................... 100
3.3 Test-patch assembly ..................... 102
3.4 Stroboscope and viewing tube assembled . . 105
3.5 Test-booth............................. 107
3.6 Detail of test-booth construction .... 109
vi
CHAPTER I
THE PROBLEM
The problem-area investigated in this study is that
of reading retardation or disability in children of appar
ently normal endowment and opportunity to learn.
This problem-area includes zones of interest shared
by many disciplines.
Education seeks to prevent reading retardation
through good initial instruction or to remedy it by spe
cial teaching methods. Psychology seeks to determine the
mental processes or causes involved and to remove, circum
vent or improve them.
Medicine also seeks to furnish some form of preven
tion or correction through certain specialties such as
pediatrics, psychiatry, neurology and endocrinology. The
law has become aware of a relationship between reading
retardation and many cases of truancy, school exclusion
and delinquency.
Political science and public administration are
forced to take account of the problem when they deal with
such matters of general welfare as literacy, employment,
and fitness for civil or military service. Vocational
1
guidance and child guidance by parents or clinics are fre
quently called upon for advice to individual cases.
The general purpose of this study is to provide if
possible some advance in fact or in methodology within
this broad problem-area.
The subsequent sections of this chapter deal with the
following topics; the importance of the problem; defini
tions and estim^t' I prevalence rates; conflicting theories
and approaches* c promising approach based on the visual
fusion threshold; description and definition of the VFT;
and the problem-attack of the present study.
Importance of the Problem:
Some Effects and Dynamics
Reading handicap in a modern or urbanized society
imposes a constant barrier to further achievement or per
sonal development. The almost inevitable results of the
failure and frustration are defense reactions such as
anxiety and hostility. Differing degrees of anxiety and
hostility are manifested in various forms of dissatis
faction, maladjustment, neurosis and delinquency. In all
of these conditions there is regrettable human wastage in
unsatisfying lives or community strife.
The consequences and dynamic effects of reading
retardation are outlined below for each of the parties
concerned: the child; his family; the school; and the
community.
Effects within the child
When a child fails to learn to read in the first
grade of elementary school as expected, emotional pres
sures mount within him. If these pressures are not soon
relieved by success, emotional defense reactions become
necessary.
Almost invariably a child looks forward to learning
to read. This is usually one of his big expectations in
starting to school. He wants to enjoy the stories in
books he has seen. And he wants to be like most of the
children who do learn to read in first grade and are very
proud of it. They could not read at the beginning of the
year either, but soon the retarded reader is surpassed by
them. They can taunt him on the playground, and call him
a dummy because he "can»t even read yet." He tries his
hardest but meets frustration day after day.
Under these conditions anxiety is almost inevitable.
The retarded reader may conclude on his own or from what he
is told that he is either dumb or lazy, or not paying
enough attention. And if he is forced to remain behind
his friends at the end of first or second grade, for the
rest of his school life he is likely to live with the
feeling that he is always "a year behind" and that he has
some explaining to do.
This long-lasting series of events is a public
"humiliation" and a continuous ego-threatening process.
The child's self-concept Is endangered. Outwardly he may
put up the defense that reading is silly, "no fun" or a
waste of time, or that school is a "dumb" place where they
try to make you work, be "good," sit still, pay attention
and so forth. He may over-engage in physical activities
and in fights in order to retaliate or to "prove" his
capabilities in other ways (compensation). Or he may
become fearful, withdrawn, tense, and dejected (reactive
depression). Depending upon whether prolonged fear and
anger are primarily directed inward or outward, the situa
tion is ripe for the development of an "acting-out" psy
chopath (character disorder) or an "acting-in" disturbance
(neurosis or psychosis).
Effects within the family
When his child has continual difficulty with reading,
the parent too is threatened. He realizes education and
advancement are dependent upon this fundamental skill.
Inwardly the parent may wonder about the child's
ability and whether there is some defect in heredity that
may be traced to himself or his family. Or he may question
whether he is handling the child correctly or being too
lenient or too harsh. He may also suspect or accuse his
spouse of these shortcomings. Such processes can place a
great deal of strain on a marriage, and may be the critical
additional stress which causes a marriage failure.
Outwardly the parent may try to "do something" to
help the child with his reading. Depending upon the
parent's own background, the help he tries to give may take
the form of more reading to the child, more drill, more
admonitions, or more punishment.
Another frequent parent reaction is to assume that
the school is at fault. The parent may conclude that
teaching the fundamentals is a lost art and that schools
nowadays should eliminate "frills" and get back to the
three R's or phonics. Friction often develops in confer
ences with the teacher. The parents may then have the
child moved to another teacher, or to another school. Or
they may move the entire family to another district or
another state. If finances will permit, they may place
the child in private school.
Effects within the school
When a school staff considers the problem of its
retarded readers, it may raise a number of questions.
Inwardly, the staff may re-examine its theories, methods
of instruction and the competencies of its members. It
may engage outside authorities to make surveys and recom
mendations. Outwardly it may defend itself by pointing to
its successful graduates, to its test results and its
comparative standing with other schools.
Another defense popular with educators is the theory
that most reading failures are due to "emotional blocks" in
insecure pupils. According to this theory, emotional inse
curities are caused at home and by the parents, not at
school or by the learning failures. The emotional distur
bances at home block the child from learning to read either
by making him too fearful to try, or by imparting some deep
unconscious desire not to read, in order to retaliate
against someone, presumably a parent, and usually the father
or a sibling. Since children are not born with such fears
or hates, it must be the parents who have induced them.
The school thus finds itself in a defensive position against
certain parents, the very ones in fact with whom it needs
the closest collaboration.
The school may also go to quite extensive and earnest
lengths to prevent or correct reading retardation. It may
hire special "remedial** or "developmental** reading teach
ers. It may provide various special reading classes. It
may give numerous group and individual tests to its pupils.
It may schedule frequent lectures and conferences for its
staff. It may study many texts, materials and teaching
methods and try to apply them thoroughly. And it may hold
many parent conferences.
Effects within the community
Within a community a situation can be produced in
which children, parents, and educators may be inwardly
tense about themselves, but outwardly on the defensive
against each other. This discontent and tension may
7
accumulate and under certain conditions erupt into intra-
community warfare. Influential families whose children
have not learned to read as expected, may mobilize this
discontent and other tensions latent in the community and
direct it against the schools. They may attempt to oust
the incumbent board and superintendent, to "clean out
incompetence, eliminate frills, and get back to fundamen
tals.* The friends of the administration may try to hold
the fort against the malcontents. Battles due to mis
understandings of this kind can keep a school district in
foment for years and leave long-lasting scars and tensions
within a community.
This outline of some of the effects and dynamics of
reading retardation serves to suggest the seriousness of
the problem and the scope of its ramifications.
Definitions and Estimated
Prevalence Rate’ s
As is the case with most human variables reading
retardation is manifested to varying degrees. Mild degrees
of the condition are usually considered to include those
whose reading rate is one or two years below their mental
age. More severe degrees are considered to have a wider
discrepancy between mental and reading ages. For the most
severe degrees, terms are employed such as nonreaders,
disabled readers, remedial readers, alexics, and numerous
other designations. No system of designation or nomen
clature appears to exist which is both logically unified
and widely accepted. Perhaps it is for this reason that
the term "retarded reader" is frequently used in the liter
ature and is usually employed to refer to all degrees of
the handicap.
This situation is not ideal. Loosely used nomen
clature obscures the quantitative degrees of retardation
just mentioned. It also obscures important qualitative
distinctions. For example a mildly retarded reader does
read to a certain extent. The extent depends, among other
things, upon his age and the number of years of instruction.
Usually he also writes at approximately the same level as
his reading, or at a slightly more retarded level. On the
other hand, a disabled or severly retarded reader remains
permanently or semi-permanently arrested at the most rudi
mentary or beginning reading level. After several years of
effort, his attempts at "reading" still remain more like
"decoding." Also, he can seldom write or "encode" the
simplest paragraph of connected sentences. This holds true
despite the fact that he may have learned by sheer rote to
"draw" his name, and if he is older, his address, months of
the year, days of the week and a small list of common
words, or uncommon ones with particular meaning for him.
He is not mentally deficient; yet no method of instruction
seems to give him material help, at least not during his
elementary school years.
Smith and Carrigan (1959) directors of the reading
clinic at the University of Michigan, observe a distinction
between the severe and mild degrees of reading retardation.
They define as "retarded readers" all pupils whose reading
age is one or more years below their mental age and who are
unable to learn by the usual methods, but who do profit
from so-called remedial instruction. They define as "dis
abled readers" those who do not appear to learn by any
method, including so-called remedial instruction.
Another drawback of loosely defined terminology is
that it contributes to the difficulty of distinguishing
between mild and severe retardation at the early grade
levels. During first or second grades, all cases of retar
dation may be thought of as mild, since they cannot fall
more than one or two years below expectancy at these grade
levels and ages. Yet with further passage of time some of
the apparently mild cases will fail to make any progress.
Only then will they be recognized as severe or disabled
cases. Nomenclature is not the cause of this problem, of
course. Nevertheless, more accurate definitions and ter
minology can assist in early recognition and early attack
on this problem of differential diagnosis.
For purposes of formal definition, in this study a
retarded reader is defined herein as one who has appar
ently normal health, intelligence and opportunity to learn
and who does not remain arrested at the first or second
10
year level, but whose reading ability is one or more years
below his mental age level (and whose writing ability is
probably equally low or lower). The disabled reader is
defined as one who also has apparently normal health, intel
ligence and opportunity to learn, but who does remain
arrested at about the first or second grade reading level.
Wherever the degree is unspecified, "reading retar
dation" is used as an inclusive term, which covers both
retardation and disability. Whenever the intent is to
indicate only the severest degrees, the term "reading dis
ability* is employed.
No mechanism such as a census exists for establishing
precise numbers or percentages of poor readers. Smith and
Carrigan (1959) estimate that disabled readers constitute
about two per cent of all potential readers. The less
severe group, the retarded readers, they estimate as more
numerous, amounting to perhaps thirteen per cent of all
elementary school children. Many other investigators give
roughly similar estimates. Marion Monroe (1932) estimates
the rate at twelve per cent (perhaps more if "borderline"
cases are included) based on her testing of 101 normal
control-group subjects (page 17).
Rabinovitch et al. (1956) estimate that "at least ten
per cent of children of average intelligence at school in
the United States are reading so inadequately as to impair
their total adjustment" (page 363)- Gillingham and
11
Stillman (I960) claim that “. . . a conservative estimate
of ten per cent of the school population experience suffi
cient difficulty in reading and spelling to be seriously
impeded in their school progress, while an additional five
to ten per cent are on the borderline, falling in skill . ..
far below their ability to comprehend the content" (page
16) .
Conflicting Theories and Approaches
Representative studies of reading disability are re-
veiwed in the next chapter. From these it can be seen that
investigators have adopted a number of conflicting theories
and approaches dealing with the problem of reading dis
ability. For purposes of discussion the approaches may
be classified as emphasizing either an instructional, an
emotional, or a neuro-physiological etiology.
Adherents of the various theories or approaches tend
to formulate their investigations in such a manner as to
support or to confirm one explanation only. The implica
tion usually expressed is that if one approach is strength
ened, the others are thereby in effect weakened. This
logically implies that evidence will ultimately pile up in
favor of one approach, and to the exclusion of the others.
Such an eventuality is of course logically possible.
However, it is also logically possible that all of the
approaches might have some measure of merit. In this
12
eventuality, each approach or etiology could be confirmed
without denial of the others. Several etiologies might
co-exist and reinforce each other.
In regard to the problem of reading disability, the
latter possibility seems quite plausible. It is not at
all difficult to imagine the hypothetical case of a neuro
logical deficit in perception or memory which would make it
difficult for a child to learn to read at a given time by
a given instructional method, and which would therefore
also result in conscious and unconscious emotional problems
because of this failure to learn. Such a case would exem
plify etiology of all three traditional approaches.
The broad etiological approach in the above hypo
thetical case has much to recommend it theoretically. In
a performance complex enough to require the over-learning
of thousands of sound patterns (audio-vocal language sig
nals) which are then encoded into visual symbols (writing)
and finally decoded visually (reading), there are many
"etiologies" which could be involved. It would appear wise
to avoid the disjunctive trap of espousing any one approach,
neurological, emotional, or instructional, to the exclusion
of others.
In the next section an account is given of a recent
study by Tait (1956) which attacks the problem by means
of testing perception with an electronic stroboscopic
apparatus. Application of this tool to the problem of
13
reading retardation is quite new and appears somewhat
promising.
A Promising Study Based on the
Visual Fusion Threshold ivftT
In 1956 J. F. Tait at the University of Southern
California completed a dissertation entitled "Mild neuro
logical impairment as a factor in reading disability: an
experimental study." This study is reviewed in some
detail for several reasons.
First, it is one of the few studies of an experi
mental nature that are aimed directly upon the problem of
retarded reading rather than at reading ability in general.
Second, it adopts a relatively new line of attack on the
problem in that it utilizes the visual fusion threshold
(VFT) test.l Third, the VFT test proved capable of dis
tinguishing a sample of retarded readers from a matched
group of normals. Fourth, the experiment showed that the
VFT test could identify retarded readers at least as early
in their development as the fourth grade.
^VFT is the symbol used throughout this study to des
ignate the subject»s visual fusion threshold* Background
and techniques for the test are given later in this
chapter and in Chapter III. This measure is referred to in
the literature by other terms and abbreviations such as
"critical flicker frequency" (CFF) and "flicker fusion fre
quency" (FFF)• These terms fail to mention that it is a
threshold which is being tested, and that it is the thresh-
old in the visual modality rather than similar thresholds
in auditory, tactile, or kinaesthetic modalities. See the
Glossary in the Appendix for further technical definitions,
principles or laws related to the VFT.
Ik
Tait designed this doctoral study to adduce support
for the neurophysiological approach or explanation of
reading difficulty. He stated the hypothesis as follovs:
• • • neurological malfunction is the primary etiol
ogy for marked reading disability in children of
average intelligence and educational opportunities.
(page k2).
The test battery consisted of two perceptual tests'*', a brief
2
history of illnesses , and most importantly, the VFT.
The subjects selected formed two matched groups, one
composed of retarded readers, the other of normal readers.
Both groups numbered fifty-one cases, and consisted of white
male pupils between the ages eight - thirteen years. All
cases were drawn from the fourth, fifth and sixth grades of
two similar and adjoining school districts in Los Angeles
County. Cases were eliminated from the study if they had
been known to possess physical defects, chronic medical
conditions, or previous symptoms suggestive of neurological
involvement, such as fainting spells, convulsions, or head
injuries. Also eliminated were cases of emotional distur
bances definite enough to have resulted in referral to the
school psychologist or elsewhere for study or treatment.
Two cases of optical deficiency uncorrected by lenses were
^Strauss-Werner Marble Board (Strauss et al. 1955);
Kohs1 Block Design is adapted in the Wechsler Intelligence
Scale for Children (Wechsler 19*+9).
2
A short health history check list filled out by
parents of the children selected.
15
eliminated, one from each group; but cases wearing glasses
to correct simple accommodation difficulties were included.
Subjects were selected by going serially through the
cumulative records of all the eligible cases in both school
districts two times. The first time was to select only the
retarded readers, the experimental group; the second time
was to match each one of them with a control case taken
from the same classroom or grade, and possessing nearly the
same IQ score on a group intelligence test, the California
Test of Mental Maturities, given in the third grade. For
the experimental group the mean, standard deviation and
range of the IQs were respectively 103.8, 8.2k, 9k-120.
For the controls the corresponding figures were 101.9,
6.k0, and 90-117.
Each child in the retarded reading group was at least
1.5 years below his grade level in reading. The range of
retardation was from -2.9 to -1.5 grade levels; mean, -1.96
below expectancy; standard deviation, O.kl. For the
"normal" group the corresponding figures were: -0.5 to
+•2.1+; +-0.27 years above expectancy; and 0.68 grade level.
From the latter figures it can be seen that the experi
mental cases were on the average two grades retarded in
reading, whereas the control cases were on the average
about one quarter grade advanced.
Using VFT apparatus and testing methods generally
similar to that described in Chapter III in the present
16
study^, Tait obtained rather accurate measures of the
groups. The standard deviations were quite small, being
only 1.94 cycles per second (cps) for the retarded readers
and 1.63 cps for the normals. The means for the groups
respectively were 37*9 cps and 41*1 cps, a difference of
3.2 cps. This difference between means was statistically
significant at the .001 level of confidence. Only one
retarded reader scored as high as the average of the con
trols, and no control case scored as low as the average of
the retarded readers. It was possible to choose a cut-off
score (39*5 cps) which correctly identified 71 percent of
the retarded group and mis-identified only 11 percent of
the control group.
This seems to provide clear and convincing evidence
that the VFT can distinguish between most normal readers
and others who appear normal or free of pathology, but who
are about two years retarded in reading in the middle
elementary grades of a representative public school. The
other measures Tait used also differentiated the groups
but not nearly as certainly or speedily.
Although Tait does not make the point, his results
suggest the possible use of the VFT as a screening test
for retarded readers.
iTait used somewhat brighter light though the exact
intensity level is not stated; there were also some dif
ferences in dark adaptation, in the light shielding fea
tures of the apparatus and in the provisions for the
subject's unhampered breathing.
Tait concluded from his results that his hypothesis
regarding a neurological basis for reading retardation was
strengthened. However, he was forced to point out that of
course full confirmation of the hypothesis would require
demonstration of neural pathology by direct examination of
brain tissue. Such brain examinations, however, are not
yet feasible in routine medical practice; and even when
obtainable, neither direct visual inspection nor micro
scopic cellular examinations of the brain can at our
present stage of knowledge be recognized as pathological
unless the lesion or abnormality is quite gross, which is
not often the case with retarded readers of average IQ.
An additional point not stressed by Tait is that the
VFT is also influenced by substances in the blood stream
such as enzymes, hormones, nutrients, toxins, gases, and
drugs. Varying amounts of any or all of these could lower
the VFT, as well as could the structural neurological
damages or faults which Tait more often mentions or implies.
Moreover, Tait does not show that anxiety, which
also lowers VFT, could not have been present in his
retarded readers and could not have accounted for the dif
ferences between his groups. There is no guarantee that
physical and emotional pathology were not co-existent in
his experimental cases.
For these reasons it seems safer and wiser to speak
of the VFT as an indicator of perceptual efficiency, either
high or low, and to leave to other procedures the task of
specifying the many etiological bases for whatever degree
of perceptual efficiency a child shows on the VFT test.
Trying to determine whether psycho-social processes or
neuro-physiological processes are the cause of reading
retardation Is akin to asking whether heredity or envi
ronment determines one's personality. The disjunctive form
of putting the question must be avoided if one is to obtain
a balanced and sensible answer. It is not a question of
whether neuro-physiological or emotional processes are
involved in a retarded reader. Both processes must be
present and interacting within him or he would not be alive
and human.
Description and Definition of
Visual Fusion Threshold (VFT)l
The previous section pointed to a relationship
between reading retardation and low perceptual efficiency
p
as measured by the VFT. If the reader is not fully con
versant with the phenomena of the VFT and what the human
experience is like, he may find the following account help
ful. To introduce and explain the basic concepts of the
VFT, some examples using familiar light sources are given
^For an explanation of the symbols see footnote 1,
page 13.
2If a technical definition is desired at this point
see Glossary in the Appendix.
19
below.
Suppose a person watches an ordinary flashlight go on
and off in the dark as he presses and releases the button
with his finger. He will see a series of light flashes and
dark periods, each separate and distinct, if he presses
and releases the button for exactly the same lengths of
time, the light-dark ratio'*' will be 1:1. If he moves his
finger faster and faster, but still in the same rhythm,
2
there will be more light and dark periods per second , and
each period will be shorter, but the light-dark ratio will
still be 1:1. If he moves his finger as fast as he can, he
will probably still have the impression of seeing separate
’ •flashes" of light, for he is not likely to be able to turn
the light on and off more frequently than five or six times
per second by pressing and releasing the button with his
finger.
But suppose he sets the switch so the light remains
on constantly and he now interrupts the beam by waving his
hand upward and downward rapidly in front of his eyes.
Since he can do this at a faster rate than pressing and
releasing the button, he will now probably experience
"flicker." The light will not appear now as separate
Abbreviated LDR.
An technical terms there will be a higher frequency
(f) of cycles per second (cps).
20
"flashes" which go on and off. Instead it will appear as a
light which is continuous, but not steady; it appears to
alternate or fluctuate in brightness and dullness so rapidly
that it may be said to vibrate, to pulsate, or to "flicker."
One of the most lucid brief descriptions and defini
tions of flicker occurring in modern literature is that
furnished by Peckham and Hart (1958):
Flicker is a purely subjective phenomenon. It
refers to the appearance of an alternating light
and dark visual stimulus at certain specific rates
of alternation. ... As the rate increases, there
will develop a tendency for the discreteness of
darkness and lightness to disappear and the percep
tion will assume a more random character; the light
appears to "flicker", (page W6l).
In the example of the flashlight, the hand cannot be
waved up and down in front of the eyes fast enough to avoid
having some noticeable effect on the light. Hence, some
degree of flicker remains no matter how fast one tries to
wave the hand. But suppose one substitutes for the hand a
disk of cardboard with many sectored openings, and with a
pencil through its center so that the disk can be twirled
or rotated rapidly in front of the eyes.
Now as one looks through the whirling openings, the
interruptions of the light may be so brief that they are
not noticed if the disk rotates rapidly enough. "Fusion"
of the intermittent light sensations has now been achieved
in the observer's nervous system. He now experiences or
perceives the rapid flashes not as a flickering light but
21
as a steady or constant light. Physically it still is an
intermittent stimulus, but the limit or threshold has now
been passed at which the individual’s sensory-perceptual
mechanism can respond to the on-off changes of stimulation.
In a sense, it may be said that the nervous system
cannot "turn on and off" rapidly enough to duplicate or to
register the changes in the light stimuli seen through the
rapidly rotating disk. Hence, there is a loss of accuracy
or fidelity in responding to the actual stimulus conditions.
This loss of sensory-perceptual fidelity or accuracy results
in the individual perceiving the actually separate light
stimuli as though they were "fused" or uninterrupted. The
VFT may thus be said to constitute the individual’s physi
ological limit for discriminating rapid temporal changes
in visual stimulations.
Again in the words of Peckham and Hart (195#) the
fusion experience is described as follows:
As the rate of alternation continues to increase,
the rate is so rapid that the perception becomes
one of "fusion", or smooth coherent lightness.
. . . The transition between "flickering" and
"fusion" is almost [emphasis not in original!
abrupt and it has long been considered to be a
specific and measurable rate . . . (page 461).
Fusion is also known, and perhaps more widely, by the
older term "persistence of vision." This term is often
employed by technicians in other fields 3uch as lighting
and motion pictures. Fluorescent lighting actually turns
on and off (fifty or sixty times per second in the U.S.A.)
22
in accordance with the frequency of the alternating current
supplied. Motion pictures are usually run either at six
teen pictures per second (8mm film) with some observable
flicker, or at twenty-four per second which eliminates the
flicker for most people at the illumination levels permitted
by the film.
With this background perhaps a brief definition of
the term visual fusion threshold may be readily understood.
The VFT for a given individual is that transitional fre
quency or rate of intermittent or cyclically changing stim
ulation below which ho is able to perceive fluctuation
("flicker") in a visual stimulus, and above which he is
unable to do so and hence perceives "fusion,” "steadiness"
or "continuousness" although the stimulus is actually fluc
tuating or intermittent.
This definition is also applicable to sense modal
ities other than vision. When the intermittency is
presented in the auditory modality the perceived fluc
tuating effect (equivalent to flicker) is often referred
to as "flutter" in the sound. Hence the auditory fusion
threshold (AFT) is sometimes called the "flutter-fusion
threshold." Fusion thresholds for tactile and kinaes-
thetlc sensations have also been explored in a few studies,
but no accepted nomenclature had evolved as yet to denote
the fusion threshold in these modalities. Such thresholds
could be represented by the designations TFT and KFT.
23
Nomenclature such as this might be more clear and less cum
bersome than the present alternatives of referring to the
CFF or the FFF for vision, for audition and so forth.
For the purposes of this study, it may be helpful to
mention here that the foregoing descriptions reveal the VFT
as one measure of an individual’s perceptual efficiency.
The more faithfully the individual’s neural mechanism can
respond so that he can perceive changes that are actually
taking place in the environment, the more efficient his
neural perceptual mechanism may be said to be. Hence, the
individual who still is able to perceive flicker at a high
rate of change in the stimulus may be said to possess
greater visual-perceptual efficiency than an individual who
perceives fusion instead of flicker at that same rate, or
at a lower rate of change. Conversely, the individual who®
nervous system reports fusion at a very low rate of inter-
mittency may be said to possess low perceptual efficiency,
at least in regard to this criterion.
It is with young children who possess low perceptual
efficiency that this study is primarily concerned.
The Problem Attack of the Present Study
If it is true that there are several interrelated
causes of reading disability, then in addition to working
with each cause separately, investigation also might well
attempt a more global and empirical attack. Problems in
biological sciences are often solved by a spirally repeated
24
sequence of steps such as the following: l) reformulation
of existing hypotheses or predictions; 2) replications or
modifications of previous measurement techniques; 3) ear
lier identification of cases; and 4) change of contingent
processes or conditions. Continued repetition of this
sequence leads to refinements of each step and ultimately
to control, reduction, or prevention of the problem by
means of changed conditions or treatments.
The orientation for this investigation follows a
similar rationale. It is assumed that the use of the VFT
test in the above study by Tait (1956) represents to some
extent a new and valid way of measuring reading retardation.
Whether the VFT is a valid measure of reading retardation
because of neurological, emotional or even because of
instructional events is not an issue here. It suffices
merely to grant that Tait was correct in showing a corre
lation between reading retardation and a low VFT in pupils
of grades four through six. If this measure can be applied
satisfactorily at still lower grades before reading re
tardation has been manifested, and if the VFT at these
earlier levels still correlates with later retardation,
then greater numbers of cases can be measured, identified,
and predicted in advance.
The dilemma at present is that no one can now predict
which children will later on turn out to be retarded or
disabled readers. Each year thousands of children enter
25
first grade who in time will be identified as retarded
readers, but not until several years later. Not till then
is formal special treatment, either clinical or remedial,
seen as obviously necessary. By then of course it is too
late to prevent the workings of earlier causative processes.
In other words, at present nothing much can be done about
reading retardation until after it has happened. Most
studies, including Tait’s (1956) have been bound by this
limitation and have attempted ex post facto correlation of
the condition rather than prediction of it.
It is true that "reading-readiness” tests are often
administered routinely at kindergarten or first grade level.
In a limited sense this might be considered as testing to
predict reading retardation. However, this is not truly
the rationale underlying such measures. Reading readiness
tests do not purport to identify cases of permanent or
semi-permanent reading retardation or disability. Rather
such tests simply claim to show which children are or are
not "ready" at the moment for reading instruction. The
tacit assumption of the reading readiness test is that any
child not ready this month or this year may be entirely
ready next month or next year depending upon normal matur
ation processes. Lack of "readiness" on these tests is not
looked upon as a pathological condition which is likely to
persist.
What is needed therefore is some kind of index, test,
26
or screening procedure which can be given relatively easily,
quickly, and economically to all kindergarten and first
grade children, and which will predict those who may not
ever be "ready1 * for reading in the usual sense of the term.
For these children intensive study and attempts at "reme
diation" or prevention of difficulty should be started at
once, in this way early efforts might lead in time to
increased understanding of the etiology, and to appreciable
reduction or amelioration of the problem.
The problems or questions to be answered or attacked
inthe present study are as follows:
1. UJan the VFT test be administered successfully to
children prior to or concurrent with the grades
or ages for beginning reading; that is, as early
as kindergarten or first grade levels and ages
5& or 6?
2. If so, can the VFT test be administered at
school?
3» If so, are any adaptations of previous procedures
necessary, desirable or feasible?
4* If VFT testing as early as kindergarten or first
grade is feasible, is it related to reading
retardation at those times, that is to reading
unreadiness or to beginning reading retardation?
5. If early VFT testing is feasible, is it related to
or will it predict later reading retardation?
27
The raising of these questions seems to be a logical
next step in the study of reading retardation.
*
It seems probable from the review of the literature
on the VFT which is presented in the next chapter, that
this test may be regarded as a very sensitive measure of
perceptual efficiency. It is also evident that perceptual
efficiency can be affected either by neurological processes,
by emotional processes, or by educational processes; or by
all of these together. Therefore if any one or all of these
processes are involved in reading retardation, there is a
good chance that the VFT might reveal the presence of the
distressing condition before it is directly recognizable
and measurable by reading tests.
Regardless of etiologies, it was established by Tait
(1956) that low perceptual efficiency (i.e. low VFT) has a
definite relationship to reading retardation after the
latter condition has been discovered. Therefore investi
gation of this relationship should be pressed to see if it
holds true before other recognition of the condition is
possible, that is, while reading retardation is still in
its potential or very early stages.
If one or more of the five questions formulated above
are answered affirmatively by experimentation 3uch new
knowledge might advance our understanding of reading retar
dation, our power to predict it, and thereby prevent or
treat it.
CHAPTER II
REVIEW OF THE LITERATURE
The questions raised for study in this investigation
deal essentially with two main topics. The first of these
is the problem of reading disability, and the second is the
VFT test which appears to offer a promising new line of
attack on the problem.
The literature on both of these topics is presented
in this chapter. While studies on the topic of reading in
general are almost staggering in number, only a small
proportion of them are centered on the problem of severe
reading retardation or disability. And of these, a still
smaller proportion may be viewed as experimental rather
than empirical or speculative in nature. These facts sub
stantially reduce the task of selecting and reviewing a
fairly adequate sample of studies on severely retarded
readers.
On the other hand, the number of studies on the VFT
among retarded readers is extremely small. Yet in order to
understand and apply VFT testing techniques intelligently
in the present study, the investigator found it necessary
to pursue further and further the ramifications of VFT
theory and practice. Thus the task of reviewing the second
2d
29
topic assumed larger proportions than expected.
Selective Keview of the Literature
on Reading Disability
This review deals selectively with reading disability,
not with the extremely broad field of reading in general.
As implied or expressed in earlier paragraphs,
several traditional remedial or operational approaches
toward the problem exist.
Instructional approaches
In these approaches the causes of the disability are
not well stated. It is implied that the child is normal,
and that the fault is primarily poor educational or
instructional practices. The approach most frequently used
by parents seems to be that of trying in some way to change
or improve the child*s instruction. One way is to move the
child to a different classroom, school, or district.
Another approach, used by some parents, is to try to
increase the child’s motivation or effort by various
rewards or punishments.
Educators use several identifiable approaches. One
is to increase or individualize the amount of attention
given to the pupil by the classroom teacher. At early
grade levels or with severe cases, another approach is to
increase the amount of "readiness1 * or pre-reading work,
such as recognizing, naming, drawing, manipulating, or
30
assembling animal forms or geometric designs.
Another educational approach is to include more of
certain kinds of training in the regular classroom, such
as phonics, or the kinaesthetic method, in order to rein
force and extend the usual visual-auditory "look-say"
method. A further adaptation is to shift from study of
wholes, such as whole words, phrases or sentences, to
smaller words or parts such as syllables, prefixes, suf
fixes, and so forth. At higher levels such as the upper
elementary or early secondary grades, special clinics,
classes or teachers are often employed to carry on word-
attack and phonetic drills as intensively and extensively
as possible with all retarded readers identified.
Underlying these remedial approaches is the assump
tion, expressed or implied, that children with reading
disabilities are not very different from other children, in
fact that they are really "normal.” It is frequently said
that "these children just need to have their particular
learning patterns understood and utilized and they will
learn as well as other normal children."
Socio-Psvchological Approaches
If instruction has been of average quality or better,
and if intelligence or medical examinations fail to find
any physical or mental deficiency, the traditional approach
likely to be taken is one of the following.
"Challenge"
It is assumed in this approach that the
child is "not being challenged," or is "just not
interested." "After all," runs this explanation,
"why should he put himself out to learn to read
when there are so many cartoons, comic books,
radio programs, movies, (and in recent years TV
programs)? A person doesn’t really need to read
these days if he doesn’t want to. And why should
this boy want to? Chances are he never sees his
father or mother read a book. So why should he?
He needs to be wakened, stimulated, interested,
challenged in school." In order to "challenge"
the pupil a new teacher is usually sought. This
new teacher then usually applies some of the
instructional approaches already mentioned above.
"Discipline"
"This child needs discipline...a firm hand.
He needs to have the ’board of education1 applied
to the ’seat of learning*...Underneath it all he
is really laughing at you, enjoying all the atten
tion he*s getting from you and everyone else...
He needs to be made to learn...by the ’hickory
stick* like they did in the old days...no monkey
business...no fooling around..."
This is the major traditional approach of
most laymen, and even of some professionals. At
school, according to this approach, discipline is
usually applied by the teacher or principal in
the form of admonitions and restrictions. At
home, it is frequently applied physically by the
father, using hand, belt, or board.
Needless to say, this simple and direct
remedy so widely used by laymen has not solved
nor prevented the problem to any appreciable
extent.
"Emotional Block"
"This child is not a bad child or a dumb
child. He is emotionally disturbed. He needs a
psychiatrist or psychologist. He can learn, but
he is afraid or unwilling to try. As soon as he
sees a book he just gets all emotional. He is
emotionally blocked. If he could get over his
block, he could read."
This has been the major traditional approach
of the past few decades for professional people.
It is still perhaps the most popular of formal
theories. Its roots lie in psychoanalysis,
psychology, and psychiatry and it has grown with
progress in these fields. The underlying
rationale is approximately as follows:
This child is not physically nor mentally
33
defective. Therefore, he is capable of
learning to read. He has had at least
average instruction, and even perhaps some
additional individual attention. Yet he
is not reading. Therefore, something has
blocked his learning. If it is not physi
cal or intellectual, it must be emotional.
Consciously of course the child says he
wants to learn to read, and the parents say
they want this also. Therefore, there must
be some unconscious emotional process that
is blocking this. Now children are not born
with unconscious blocks— it is normal for
a child to want to learn. Therefore, it
must be something the parents have done and
probably are still doing, that created the
child's unconscious block and is still
keeping it alive.
It takes therapy to uncover unconscious
processes. And therapy probably won't help
the child unless the parents change. There
fore, at least one and preferably both
parents should take psychotherapy. Then
play-therapy should also help the child.
However, therapy for the child may not be
the most important step. It may be unnec
essary if the parents change sufficiently;
and it may be useless if they do not.
Therefore, the parents are the ones who
need the therapy most.
Examples in the literature supporting the
basic assumptions in this line of thinking may be
cited briefly and selectively as follows: Freud
(192*0 ascribed hysterical blindness and other
psychogenic visual disturbances to unconscious
emotional factors. Strachey (1930) discussed
unconscious factors in reading. Blanchard (19*+6)
elaborated psychoanalytic concepts of reading
disability. Jarvis (1958) described various clin
ical observations of visual behavior in reading
34
difficulty.
Hecently Walters, Van Loan, and (Jrofts
(1961) conducted an experimental test of several
emotional-block hypotheses. Specifically, this
study tested the hypotheses that retarded readers
would show the following symptoms: (1) greater
than normal hesitation in looking at a sexual
object (undressed male doll) because of uncon
sciously associating "looking" in reading with
"looking" at taboo objects; (2) greater "avoid
ance" in perceptual tasks involving non-sexual
objects also (dressed doll and obscured or incom
plete drawings) because of generalization of the
taboo against looking, and (3) greater hostility
toward the same sex parent, because of "faulty
identification mechanisms," presumably arising
from guilt, fear, or anger engendered by the
parent.
These authors studied a total of fifty-
eight boys (no girls), in grades 3-6, with IQs
from 86-112 and with no obvious physical defects
or behavior problems. The cases were subdivided
into three groups: 20 retarded readers, 20
average readers, and 18 advanced readers.
The retarded readers were found to be
poorer in recognizing obscured or incomplete
35
drawings of neutral objects. They were found to
be slower initially in looking at the nude doll,
but their rates of change or improvement were
the same as the average readers. The retarded
readers chose their fathers less frequently on a
parent preference test. However in projective
stories made up to go with various pictures, the
retarded readers did not show any increase of
hostility toward their fathers, no any increased
"fear of looking" at objects in the pictures.
Walters, Van Loan and Crofts concluded that
some of the findings of the study did support
psychoanalytic theory while others did not. They
themselves finally decided in favor of an alter
native interpretation. They believe that "per
ceptual skills are highly developed among
advanced readers, and poorly developed among
retarded readers" (page 282). They believe that
this difference in perceptual development is not
due to "fear of looking" in the retarded readers.
Instead they attribute the difference in percep
tual development to more practice in the advanced
readers, due to more approval and stimulation by
their parents. In the words of the authors,
"advanced readers have had greater reinforcement
of exploratory behavior by care-taking adults"
(page 282).
These authors do not mention several other
possible explanations, such as the following:
(a) possible higher native perceptual ability in
advanced readers, (b) possible organic inter
ferences with perceptual ability in the retarded
readers, or (c) possible differences in percep
tual maturation rates.
The study just quoted illustrates several
points regarding the emotional block approach.
First, this study is the only one of the
selected related references that is experimental
rather than clinical in nature. Hence it empha
sizes the fact that there have been few experi
mental tests of the theory of unconscious reading
blocks. As the authors of the article express it,
MThe psychoanalytic theory of reading disability
. . . has received relatively little attention
from research psychologists in spite of its pop
ularity among . . . therapists'* (page 277) •
Second, the two tests of perceptual effi
ciency in the study (obscured drawings and incom
plete drawings) did show a difference between the
groups, whereas the projected phantasy material
gave completely negative results. The implica
tion here is that perceptual efficiency is more
37
likely a causative factor of reading handicap
than are unconscious emotional processes.
Third, there is no doubt that much can be
said for the traditional emotional block explana
tion as a humane, thoughtful, and sophisticated
approach. However, in the last analysis it pre
dicates that the source of the handicap is really
not in the child, but in the parents. Therefore,
it tries to interest the parents in treatment or
to focus treatment on them rather than on the
child.
Even the brief list of selected studies
mentioned above makes it evident that the emotional
block approach has existed for about forty years
and has been quite widely tried clinically.
Unfortunately it is also quite evident that this
approach has not materially reduced the problem,
nor does it seem likely that it soon will do so.
Neurophvslologlcal approaches
The possibility of a neurological basis for reading
difficulty was envisioned very early.
Hinshelwood (1917 clearly described cases of dis
abled readers. To denote this condition he used the
colorful and expressive term "congenital word-blindness.1 1
He rather shrewdly suspected that it involved the region of
supra-marginal and angular gyri of the parietal lobe, a
38
region of the brain that is quite suspect today.
Henry Head (1926) used the term "reading aphasia,1 *
ascribed it to localized brain lesions, and described adult
patients with brain injuries who could not read words but
could still recognize isolated letters.
The neurologist, Samuel T. Orton (192S) employed the
term "strephosymbolia" to denote interference or reversal
of images or symbols by competing activity from the non
dominant hemisphere of the brain. The question of the
effects of "mixed" cerebral dominance has been debated many
times since then, and is still an unresolved issue.
Dearborn (1931) also subscribed to the influence of
cerebral dominance. He reported that mixed-dominance and
left-handedness were strongly associated with reading dif
ficulty. However, he did not believe the interference
effects of "mirror images" were as important in reading
disability as were the effects of opposed laterality or
confused directional tendencies. A good deal of controversy-
developed around this issue. It was shown that the obverse
of the proposition did not hold true; not all ambidextrous
and left-handed pupils manifest difficulty in reading. This
and other neuro-medical approaches then became unpopular,
as compared to the very popular "emotional block" or instruc
tional approaches.
Thurstone (1944) investigated various psychological
processes by means of factor analytic methods. In one of
39
his studies on perception he used two groups of college
freshmen, one composed of slow readers and the other of
fast readers. He found statistically significant differ
ences between them and he concluded that reading is pri
marily a perceptual function in which the task of the
reader is to form associations quickly with "rapidly
changing visual stimuli." He also expressed the opinion
that in regard to reading retardation too much attention
had been centered on problems of motivation and personality
adjustment. He recommended more investigation of visual
perceptual functions.
Strauss and Lehtinen (1947) and other collaborators
conducted long-term studies of mentally deficient pupils
of both the hereditary (endogenous) and the brain-injured
(exogenous) type. They considered that visual perceptual
defects were so common among the injured type that the
diagnosis of brain injury could be partly determined from
this symptom.
Lillian Wagenheim (1949) reported a statistically
significant relation between later reading disability and
measles contracted at age 2 and also at age 5» She con
siders that these two ages might be unusually important in
that one coincides with the development of speech and the
other with entrance into school. Encephalitic sequelae are
known to be frequently associated with at least one form of
measles (Rubella).
40
Richardson (1950) in England completed a factor ana
lytic study of reading ability which agreed with Thurstonefs
interpretations in some respects and differed in others.
Instead of two groups of college students, Richardson used
ten-year old children with a normal range of reading ability
and gave them a variety of intellectual, auditory, visual
and personality tests. Three major fcCtors were found to
account for most of the statistical variance. These fac
tors were designated as intelligence, visual-auditory
mechanics and emotional stability.
Also very interestingly, a combined factor, made up
of both the visual-auditory mechanics and the emotional
stability factors, accounted for over 50 percent of the
variance. This tends to agree with Thurstonefs interpre
tation, in that the visual-auditory mechanics factor is
probably the equivalent of his perceptual factor. But it
also differs in that considerable weight is given to an
emotional or personality factor for which Thurstone recom
mended de-emphasis.
In what has been termed one of the most exhaustive
and detailed studies of disabled readers, Helen Robinson
(1946) reported the findings from detailed examinations of
disabled readers by an inter-disciplinary team of thirteen
specialists from various educational, psychological,
optical, and medical subfields. Represented among these
were such specialists as remedial teachers, speech
41
pathologists, oculists, audiologists, neurologists, psy
chologists, psychiatrists and endocrinologists. For the 22
cases examined by all specialists, no one cause could be
cited for reading disability. But it was noted that as the
number of abnormal physical findings increased, so did the
degree or severity of reading retardation. Visual anom
alies of some sort were present in 73 percent of the cases
and were judged as contributing to the reading disability
in 50 percent.
Eames (1953) examined the blood analyses of thirty
reading failures. In results similar to Hobinson*s he
found no one abnormality related to all cases, but the
general category of "abnormal blood cell forms" was present
to a statistically significant degree.
Statten (1953), working in a children*s outpatient
psychiatric clinic in Montreal, noted that visuo-motor
disturbances, reading disability, and school failure were
very common complaints. He found that the Goodenough
(1926) Draw-A-Man score of his patients was on the average
markedly below their mean on the Wechsler (1949) Intel
ligence Scale. Moreover, on their electro-encephalograms
(EEGs) most of his cases showed diffuse abnormalities,
especially featuring slow waves of 2-4 cycles per second
in the occipital regions of the brain. Statten suspected
that all these symptoms may have a neurophysiological
basis.
42
Burks (1955) sought to develop a rating scale which
elementary school teachers could use to screen classes for
cases of neurologically impaired children. He used two
groups of subjects, one consisting of 94 l *unselectedH
children from a parochial school and 137 children from
public school who had been referred to the school psychol
ogist on account of faulty behavior. Both groups were
given EEGs. The incidence of abnormal tracings was about
9 percent in the parochial group, and over 60 percent in
the referred group. Heading difficulty was also one of the
single most frequent symptoms of the referred group. The
inference here is that reading retardation is associated
with abnormal brain function.
Many other studies exist, especially in medical pub
lications, which are not primarily concerned with reading
problems, but which nevertheless give evidence supporting
the neurophysiological approach. Examples of reading dis
abilities (alexias or dyslexias) are numerous in medical
reports of war casualities, automobile or industrial acci
dents, tumors, strokes, and other conditions. Such condi
tions have repeatedly been demonstrated to produce an entire
spectrum of language disorders or aphasias. Aphasia is
broadly defined by Barry (1955) as applying to disturbance
of any or all linguistic functions including reading,
writing, thinking, and even the control of emotions or
impulses.
43
The evidence for a neurophysiological cause of reading
retardation is clear when the alexia occurs in adult cases
who previously possessed normal language functions which
were lost after the onset of a known illness or injury.
The evidence for a neurophysiological cause is often less
clear of course in childhood cases because they have not
first possessed fully and then lost the reading or other
language function. Nevertheless in numerous instances the
parallelism between cases of adults and children is com
plete enough to give strong support to the neurophysiol
ogical explanation of reading retardation.
In summarizing this section it may be pointed out
that there are many theories of reading retardation. They
may be broadly divided into three classifications: instruc
tional, socio-psychological and neurophysiological.
The single most widely held theory, at least among
professional personnel, appears to belong to the socio-
psychological classification. It rests on the hypothesis
that an unconscious emotional block exists in the child
due to some fault in parent behavior. This theory has
received little experimental study and has not made mate
rial inroads on the problem in the past quarter century.
Neurophysiological theories are also quite old. They
are quite widely accepted when there is evidence of ante
cedent reading and subsequent disease or injury resulting
in loss (dyslexia). Acceptance of the theory is not wide
44
when antecedent disease or injury is posited to explain
subsequent non-acquisition (alexia).
It would seem that much experimental study, espe
cially predictive study, is still required for material
reduction of the problem.
Review of the Literature on the VFT^
The history of studies of visual fusion phenomena
reaches back into antiquity. Landis (19510 considers that
it began with observations on the "persistence1 * of vision
and that in an informal or pre-scientific sense the history
extends as far back as Plato. In the formal sense of truly
experimental study, he considers it to have begun on
June 19, 1834, with the presentation of a paper presented
to the Royal Society of London by Henry Fox Talbot, and
with the contemporaneous work of J. Plateau in France.
Talbot formulated the relationship between rapidly
alternating light and dark periods as they combine to pro
duce the perception that the intermittent light is dimmer
than the unobstructed light. Plateau (1835) confirmed and
refined this relationship, which is referred to in the lit-
2
erature as Talbot*s Law or the Talbot-Plateau Law .
•^VFT is the designation used in the present study to
refer to the visual fusion threshold in preference to such
designations as GFF and FFF. See footnote 1, page 13 and
glossary.
2
See glossary at end of this chapter far fhller definition.
In view of this long history, it is somewhat sur
prising that as recently as the year 1953 no investigator
had previously attempted or completed a survey of VFT lit
erature and published the results in a comprehensive bib
liography. This fact became evident in 194^ in connection
with the Columbia University-Greystone Brain Research Pro
ject, in which Landis (1949) and others (Young 1949, et al.)
found that the VFT was lowered in patients after surgical
removal of the frontal lobes. In order to explain these
surgical results Landis searched further and further
through the literature. He could find no general prin
ciples or theory explaining the results nor relating the
many separate studies that had been done on the VFT.
With aid from several sources'1 ', Landis continued and
expanded his search for several years. At length he suc
ceeded in compiling and annotating a bibliography containing
more than 2,000 related studies (Landis 1953). Over half
of these bore directly upon VFT phenomena. The studies
were drawn from more than twenty disciplines including
physics, chemistry, mathematics, physiology, neurology,
opthalmology, pathology, internal medicine, endocrinology,
pharmacology, biochemistry, psychology, and electrical
engineering. All major abstract journals and indices were
searched regardless of language, and qualified translators
■'"Carnegie Corporation; Rockefeller Foundation; Armed
Forces National Research Council.
46
were used for publications in Russian, Czech, Polish, Nor
wegian, Japanese, and Latin. Unfortunately however, this
valuable bibliography is still not available as a commercial
publication^, nor is it widely known or quoted in subse
quent studies.
While this literature search was still in progress,
Landis (1951) wrote a short but fairly comprehensive sum
mary. He published it, understandably, in a multidisci
plinary journal (Scientific Monthly), under the modest
title: “Something About Flicker-Fusion.” This article
constitutes a good orientation to the subject. It listed
the major factors or conditions then known to affect an
individuals VFT. The list was substantially as follows:
1. The intensity of the light.
2. The color of the light.
3. The ratio of light to dark periods.
4» The size, shape, and distance of the light patch
from the eye.
5* The color and intensity of the field surrounding
the light patch.
6. The condition of the eye with respect to dark-
adaptation.
7* Monocular versus binocular viewing.
S. Age of the subject.
9« A host of biological and psychological conditions
such as: body temperature, vitamin deficiencies,
central nervous system injuries, and many chem
icals or drugs.
In surveying this list one sees that the factors
It is available on request from the Armed Forces
National Research Council Vision Committee Secretariat,
3433 Mason Hall, University of Michigan, Ann Arbor, Mich
igan.
47
might be subsumed under two categories^-: The first five
factors2 seem to fall into the category of variations in
the external stimuli. The remainder fall into a second cat
egory of variations within or between subjects.
ttegarding the second category, variables within the
organism, Landis offered the following conclusion as a gen
eral principle or theory to account for all the facts in the
field:
Any condition or agent which acts to decrease the
available blood sugar and/or oxygen available to the
retina or to the brain, decreases CFF, whereas any
condition increasing the efficiency of the vascular
supply increases UFF. In other words, CFF seems to
be a measure of functional efficiency of the retina
and/or the visual cortex." (Landis, 1951, page 314*
Empnasis nrt in the original.)
Among those substances or conditions known to decrease
the threshold, Landis listed the following: sedatives,
vitamin deficiencies (A and u), excess carbon dioxide, cer
tain opthalmological conditions, acute physical exhaustion
and military "combat exhaustion," anemia, brain tumors and
brain injuries, certain neuroses and psychoses, acute
depression, and exogenous mental retardation. In contrast,
he listed the following as those conditions which increase
the threshold: strychnine, thyroxin, benzedrine, adrenalin
small doses of glucose, dinitrophenol, pervitin, and testos
terone.
•^These categories were not explicitly utilized by
Landis.
^Numbered 1, 2, 3, 4, and 6 by Landis.
k8
In what might be termed a third category in this
article, Landis was forced to list the following factors or
effects as still unknown in their influences on the VFT:
1. The effect of learning.
2. The effect of genetic factors.
3. The effect of color blindness.
h . The relative influences of retinal tissue versus
brain tissue.
One might suppose that as a result of the many
studies of VFT, a legacy of well-established norms and
expectancies would be available, which would hold true
under usual conditions or at least under certain standard
testing conditions. Unfortunately, this desirable state
definitely has not been achieved. The lack is clearly
evidenced in the studies reviewed below.
Presumably this lack of normative data is due at
least in part to the following factors: the confusing
multiplicity of conditions affecting the VFT; the diffi
culty of specifying at a given moment the exact internal
conditions of the organism; the delicacy, precision and
cost of adequate VFT instrumentation; the lack of standard
ized testing procedures; and the experimenter's dependence
upon the testee for accurately reporting his subjective
experience of flicker or fusion, particularly at the thresh
old moment when his perception shifts from one effect to
the other, i.e. from flicker to fusion or vice versa.
Shortly after Landis* (1951) summary, a very exten
sive and useful review of the background and various
49
applications of VFT testing was published by Simonson and
Brozek (1953)* At the time of its appearance, and still
today, it is one of the best compilations of the known
facts relating to methods, variables and the effects of
various physiological and clinical conditions.
A year following completion of his annotated bibli
ography, Landis (1954) published in the journal Physiol
ogical Review the most complete attempt up to that time to
list and to briefly assess all the known and even some of
the "unknown" factors affecting the VFT. The list of
"determinants" mentioned in this treatment includes fre
quency, luminance, visual angle, retinal area and location,
dark adaptation, exposure time per trial, color, wave form,
eye shape (of species) and pupil size, monocular and binoc
ular stimulation, retinal and cortical factors, body temp
erature, diurnal variation, age, physique; concurring
tones, odors and tastes; fatigue, practice-effects,
instructions, attitude, and interaction effects, in these
areas he finds considerable agreement between investi
gators, but also much disagreement and confusion because of
incompletely controlled and widely differing conditions.
One topic in particular, the light-dark ratio (LDR),
Landis (1954) considers as thoroughly unsettled, since
various investigators have reported diametrically oppo
site results from variations of the ratio:
. . . the confusion is evident and other than
50
attributing the difficulties to the interaction of
variables and the lack of independence of c.p.s. and
LDR no clarification can be offered. . . Perhaps LDR
is no more than an arithmetical device which never
should have been applied tc such measures or otherwise
stated, perhaps LDR is a pseudo-problem. If it is a
real problem, it is certain no one has offered the
beginning of a solution, (page 274)*
Surveying the field of VFT investigation as a whole
and considering the much that is known and the greater
amount that remains to be known, Landis (1954) concluded:
This review of the determinants of CFF has not
provided an adequate basis of explanation of the
flicker-fusion phenomenon. At present no one of the
hypotheses advanced by various investigators will
cover more than a fraction of the experimentally
determined observations which have been made and
recorded, (page 2$4)•
Pertinent studies or parts of them are reviewed
below and are presented under the following major topics.
The order of studies under each topic is chronological
except in a few instances.
1. Methodology of VFT testing.
2. Age and sex differences in VFT.
3. VFT and intelligence.
4. VFT and physical pathology.
5. VFT and anxiety.
Methodology of VFT testing
In the period from Talbot and Plateau up to decades
as recent as 1930-1940, experimenters chiefly were con
fined in their instrumentation to some form of sector-
disk (see glossary) which was placed in the path of a
beam of light and rotated, in one arrangement the sectors
of the disk were partly solid and partly open, and the
51
light was transmitted intermittently through the open
sectors of the spinning disk. In another arrangement the
sectors were left solid, but were painted different colors,
(e.g., half the disk was painted white and the other half
black); the light was then reflected for an instant from
each colored sector as it whirled past a small opening in
a screen. Early investigators used muscle power and
gears to rotate the disks, and vibrating strings of cer
tain lengths viewed through slots in the disks to deter
mine the speed of rotation. Later investigators used
electric motors to drive the disks and tachometers to
register the speeds. Extraneous sounds and vibrations
were the natural accompaniments of these methods. Not
until the latest few decades were electronic instruments
available in sufficient supply to gradually supplant the
older methods and to eliminate the possibly complicating
auditory and vibratory cues. Most of the studies reviewed
here are "modern1 * in this sense and may be presumed to
have used electronic instruments unless sector-disks are
specifically mentioned.
Knox (1945) investigated the influence of the
subject’s mental "set** toward the test, i.e., whether the
subject was set to observe flicker, or to observe fusion.
This is of methodological importance because when one is
set to observe fusion, the stimulus must be started at a
52
low rate (for flicker) and must then proceed to progres
sively higher rates until fusion is reported; this is
often called the “ascending order** of stimulation. Con
versely, when one is set to observe flicker, the stimula
tion must start at a fast rate (for fusion) and “descend**
to lower and lower frequencies until the subject reports
flicker. Theoretically, the threshold should be reported
at the same frequency, but seldom do subjects respond
with such accuracy. Knox concluded that the threshold
is reported as slightly higher when the subject is set
for flicker (descending trials) and lower when he is set
for fusion (ascending trials). He also concluded that
this was due not entirely to “set” but also to practice
effect. He used only six subjects, but gave 20 trials
per day for 10 days.
Landis (1951) listed the following as recommenda
tions for achieving adequate testing procedure:
1) Subject should be comfortable, and there should
be a minimum of distractions or noise which
complicate results.
2) An equal number of flicker-to-fusion (“ascend
ing") trials and fusion-to-flicker (“descending”)
trials should be given and the results should
be averaged.
3) if an individual*s own (ipsative) measures vary
by more than 4 or 5 cycles per second (c.p.s.)
they should be suspected because the typical
standard deviation of a normal subject’s scores
is less than 2 c.p.s.
53
4) The subject should not be required to view the
flickering light for more than 3 or 4 seconds
at a time, since it is an unpleasant sensation
to most people; the more intense the light, the
more true this is likely to be.
5) The size and shape of the test patch of light,
and its distance from the eye, should be con
trolled and reported.!
6) The visual angle should be between 1.0 degrees
and 1.3 degrees, since complications by other
factors are likely to arise if smaller or larger
angles are used.
7) The light-dark ratio must be specified, since
the threshold changes with this ratio.2 (The
less time the light is on during each cycle,
the longer and hence more noticeable is the
dark period, and hence the higher the frequency
at which the subject can still perceive flicker.)
(Compare with the later comment by Landis (1954)
regarding LDR quoted above.)
^The size and the distance of the patch from the eye
will of course determine the size of the visual angle or
are of the light rays as they enter the pupil and nence
the size or total area of the retina upon which they will
then fall. The area of clearest visbn, the fovea (also
called the central or photoptic or cone-cell area, as
distinguished from the rest of the retina which is called
the peripheral or scotopic or rod-cell area) is only a
few millimeters wide in man. This is the area used for
most conscious, voluntary visual work. In order to test
just this visual area without possible complicating
effects from the peripheral areas of the retina, one must
use either a small test patch (small visual angle), or a
mask with pinholes serving as “artificial pupils1 * which
will achieve the same effect.
2
The apparatus used in the present study and in many
others utilizes a Strobotac (see details in Chapter III)
which gives a changing LDR since the flash or light period
of the cycle is always the same in length, while the dark
period is shortened or lengthened as the frequency is
raised or lowered.
54
3) Attention must be paid to the brightness and
hue of the * immediate surround* of the test
patch for this too can influence the results.
9) One must be sure to instruct the subject and to
demonstrate to him what a 'flickering* test
patch looks like as compared with a ’steady*
(fused) one.
Ricciuti and Misiak (1954) offer serious criticisms
of the widely used variable stimuli or '*just noticeable
difference'* (j.n.d.) method. The j.n.d. method is the one
discussed thus far, in which the stimulus is varied by
slight and continuous gradations, it is the method
utilized by all other investigators reported in this study
unless expressly mentioned otherwise. According to
Ricciuti and Misiak the traditional j.n.d. method suffers
from the following four short-comings:
1) it requires relatively long exposures of
approximately £ - 15 seconds for the experimen
ter and the subject to go gradually from
flickering frequencies to a fusion frequency or
vice versa at a pace at which the experimenter
can be relatively sure of the frequency to
which he had adjusted the machine at the
instant the subject indicated his threshold.
These relatively long exposures, which must be
repeated, require relatively sustained and
well-mobilized attention. The experimenter
cannot always count on this when testing
children, the mentally ill, the aged, or the
mentally deficient.
2) It encourages the subject, in going from flicker
to fusion and vice versa, to make errors of
anticipation. This was referred to by Knox
(1945) above as mental set, and is the chief
reason why Landis advocated averaging an equal
number ascending and descending trials.
55
3) It yields relatively wide variability in the
individual’s report of his thresholds.
4) It does not permit the experimenter to be sure
that he was always changing the stimulus at
the same rate from flicker to fusion or vice
versa. Thus he might make the transition too
slowly for the subject on one occasion and too
fast on another.
In order to overcome these objections, these
authors utilized the other equally classic and traditional
method of judging increments or differences between
stimulus gradations, namely, wthe method of constant
stimuli.u They used seven different frequencies,
arranged in random orders, and presented each frequency
six times for a total of 42 exposures to each subject.
The seven frequencies were graduated in steps of .£ of a
cycle per second, and hence covered a range of 4»3 c.p.s.
(This approximates the ipsative standard deviation men
tioned by Landis, 1951.) The exposures were only seven
seconds long, thus presumably reducing intra-trial
visual changes. The rest intervals consisted of 30
seconds of dark between each exposure, except after the
twelfth and twenty-seventh exposures when the dark rest
period was lengthened to one minute. Testing was pre
ceded by a five-minute dark adaptation period. Total
testing time per subject was reported as approximately
15 minutes.-*- On each trial the subjects merely had to
Vrhis is puzzling, since the stated exposures and
rest intervals amount to 15&4 seconds or 26.4 minutes.
56
give a simple yes or no response as to whether the light
was flickering.
The authors concluded from a comparison of the two
methods on 12 cases that the method of constant stimuli is
superior to the j.n.d. method for the following reasons.
1) Intra-individual variability is lower. 2) Inter-
individual variability is higher. 3) Test-retest relia
bility is considerably higher. Therefore they consider
the method of constant stimuli preferable unless one is
dealing with well-trained and motivated subjects capable
of maintaining sustained attention and making difficult
judgments. They admit however that this method is some
what more complicated than the j.n.d. method both in
administration and in the computation of results.
Axel Mahneke (1956), a Scandinavian investigator,
agrees in substance with Ricciuti and Misiak above, at
least in regard to some of the drawbacks of the j.n.d.
method. He points out that in averaging ascending and
descending trials, two assumptions are implied: 1) that
accelerating (ascending) or decelerating (descending)
change in stimulation is uniform throughout each and every
trial; and 2) that acceleration has the same effect upon
the threshold in an ascending trial as deceleration does
in a descending trial. These assumptions are technically
very difficult to test, and apparently there is as yet
no study which has undertaken or succeeded at this task.
57
Brown (1956) used a line several inches long formed
by the sweep of an electron beam on a cathode ray oscillo
scope to demonstrate that foveal and peripheral thresholds
are different. He instructed the subject to look fixedly
at one end of the line (foveal vision) and to adjust the
frequency for himself until the flicker vanished at that
end. At the same instant the subject could report that
a portion of the line toward the other end, seen
indirectly (peripheral vision), was still flickering
though of course the oscilloscope was presenting the
entire line at exactly the same frequency. This is a sim
ple and clear demonstration that for intermittent stimuli,
peripheral vision is more acuto than foveal vision, it is
also an illustration of why a small visual angle should be
used in order to avoid complicating effects, as recommend
ed above by Landis (1951)*
Foley (1956) conducted a somewhat similar demonstra
tion showing the effects of peripheral vision and of
backgrounds of various sizes and brightnesses. He used a
test patch that was square in shape and only 0.5 of a
degree in size of visual angle when the subject was
seated at a distance of one meter. While the test patch
was held constant in size, the luminous background was
progressively increased in size from 1° to 16° of visual
angle. The VFT increased linearly with the size of the
background. The effect was greatest when brightness of the
background equalled the brightness of the test patch. This
study illustrates the sensitivity of the threshold to the
total quantity of light striking the retina, and hence
the necessity for controlling the size and intensity not
only of the test patch but also of the surrounding sur
faces or sources of light.
Doehring, Ward, and Hixson (1956), in a study under
the auspices of the U. S. Naval School of Aviation,
developed and standardized a method for group testing the
VFT in adults. Their apparatus was housed in an air-
conditioned, sound-proofed trailer. It consisted of 10
separate booths, each containing a neon lamp (Drake type
105 "Post-lite"), all lamps being operated by the same
current from a control room compartment. The current was
a series of impulses ("square1 1 wave) formed by an audio
oscillator (Hewlett-Packard Model 200AB) and yielding a
1:1 LDR at frequencies governed by turning the dial of the
oscillator. The j.n.d. method was used, the subjects
receiving "descending" trials only. They signalled their
perception of the change from fusion to flicker by
releasing (not pressing) a telegraph key (pressing pro
duced an audible click). The responses of all ten tele
graph keys were automatically recorded on a graphic record.
They found they could test 10 naval servicemen at a time
59
in a period of £ minutes.^
This study represents an improvement on the only
other group test study previously known which was conduct
ed by Mucher and Wendt (1951) on the effects of fatigue and
caffeine. In that study the subjects sat in a circle
around a single light-source and indicated their response
by pressing a telegraph key. It is noteworthy that no
group-testing methods for VFT were apparently ever re
ported prior to 1951, and no group VFT tests have ever
been used with school children so far as this reviewer can
determine.
Ronchi and Bittini (1957), Italian investigators,
explored the possible effects of light flashes of two
different wave-patterns of intensity. In the first wave-
pattern, the flash of light reached its maximum intensity
almost immediately and stayed at that intensity until
cut-off abruptly. When represented graphically, this
wave-pattern is called the "square or rectangular wave."2
The other form of light flash increased gradually to its
peak of intensity and then also decreased gradually. This
^In comparing the results on 100 naval cadets and
174 enlisted men, they report that the results yielded by
this method were significantly different for the two
groups and that intra-individual variability was minimal.
For further discussion of the implication of these find
ings, see the section on the VFT and intelligence, page
75.
2.
Light period
60
is the so-called "saw-tooth wave" pattern.^ They found
that for foveal vision the two kinds of waves produced
the same VFT. However, in peripheral vision, more sen
sitivity vas obtained using the saw-tooth wave. These
similarities and distinctions held true for white, red,
and green light (but not for light of wave lengths below
U90 millimicrons).
Bujas (1957) a Jugoslavian worker, investigated the
effects of stimulation both by intermittent light and by
intermittent electrical stimulation of the eye. He found
that previous stimulation of the eye by intermittent
current has no effect on the VFT for light. However,
previous stimulation by intermittent light at sub-fusion
frequencies did have the effect of slightly lowering the
VFT. The effect was maximal after an exposure of 20
seconds. This is of some importance methodologically. It
raises the question of whether or not some degree of
adaptation is bound to take place in the subject during
successive trials when testing by the j.n.d. method.
This seems highly probable, since the j.n.d. method cus
tomarily requires several alternate ascending and
descending trials. Before and during each ascending
trial the subject customarily looks for several seconds
at sub-fusional frequencies, and hence his VFT should be
iLight period A_A_ dark period.
61
lowered on ascending trials. Moreover, if the investigator
happens to pause for some reason (for example, to record
the score from the preceding descending trial) and allows
the subject to look for a number of seconds at a sub-
fusional frequency, the subject*s VFT on the forthcoming
ascending trial might thus be lowered. This point is not
fully or frequently discussed in the literature.
Also bearing on the question of visual adaptation is
the work of Dondero, Hofstaetter and 0*Connor (1958).
They tested 78 male college students under both light-
adapted and dark-adapted conditions of the eye. Three
ascending and three descending trials were given under
each condition. They found that under light-adapted
conditions the threshold was higher, and variability
between subjects was greater.1
Bartley (1958) investigated the effects on the VFT
of using differing intensities of light and also differ
ing light-dark ratios (LDRs). He found that at various
levels of light intensity more than one LDR could be used
to produce fusion at a given frequency. Figure 2.1
(modified from the original) presents schematic curves to
illustrate the preceding statement. This study is another
^However, this effect was not universally true; it
appeared to be evidenced most among those who tested high
on the Taylor Manifest Anxiety Scale, and absent in those
who scored low. The effects on VFT of anxiety and other
conditions are dealt with in subsequent sections of this
chapter.
Frequencies i n cps
62
Intensities of light
LDR 1:9
LDR 1:1
LDR 9:1
20
10
Initial
Intensity Multiples (logarithmic)
of initial intensity
FIGURE 2.1
SCHEMATIC ILLUSTRATION OF RELATIONS
BETWEEN FREQUENCY, INTENSITY AND DURATION
OF LIGHT PULSES AND THE FUSION THRESHOLD
Adapted from Bartley, H. S. (195#)
Exposing the eye to a cycle with a short
light period and a long dark period, for
example a LDR of 1:9, will give a fusion
threshold at higher frequencies than cycles
with shorter and less noticeable dark
periods, such as LDRs of 1:1 or 9:1. And,
two different LDRs may yield a VFT at the
same frequency if the intensity of the
light is adjusted to a certain requisite
level, such as the level at point X in the
schematic diagram. This illustrates the
interdependence of variables affecting the
threshold.
63
illustration of the fact that the perceptual threshold of
flicker or fusion is determined by a complex inter
relationship between the amount of light energy received
by the eye (luminance, the visual angle, etc.) and the
time or rate (frequency, LDR, etc.) of receiving the
amount.
Peckham and Hart (1953) projected light flashes
which measured 14 inches in diameter on a screen 10 feet
away from the subject. Under these conditions they
recommend the following methodological specifications:
1) The light-dark ratio should be 1:1.
2) The surrounding field must closely approach the
same intensity of luminance (Talbot level) as
the flickering area.
3) Either the LDR or the cps may be varied.
4) The stimuli must be presented in a standard
fashion and the results must be treated
statistically.
These authors pdnt out the following drawbacks
shared by many studies:
. . . unless some precaution is devised (such as
artificial pinhole pupils) a small undetected change
in fixation during the testing will result in error
of the estimate of critical flicker frequency. Those
tests which are made in darkness or in dim rooms with
bright and relatively small centrally fixated stimulus
areas could well permit some rod function in the peri-
macular field of vision. This could cause a serious
error and could explain some of the disputation
between investigators. . . . (Also, they point out,
current investigators are using stroboscopic lamps in
which the light-dark ratio necessarily varies with the
frequency, since the duration of each light flash is
the same— a few millionths of a second.) .... Unless
reports specifically state both the background
64
conditions and the light-dark ratio, no absolute
critical flicker-frequency values are meaningful out
side of their specific context. Nevertheless, when
flicker fusion frequency is measured on a number of
persons within the same set of stimulus conditions,
responses can be treated as group results, with great
reliability. . . . (page 470).
in closing and summarizing this sub-section of the
literature, the following points may be considered to have
been established regarding VFT methodology:
1) No standardized method of measuring the VFT is
available or in general use at present either
for group or individual testing. Hence, no norms
have ever been developed and published, and there
is no direct comparability of scores from one
investigation to another.
2) The method of ♦constant stimuli1 is probably
more accurate than the method of * just noticeable
differences*, but it appears to be more compli
cated and time-consuming to administer, and to
score.
3) The VFT is higher when the eye is light-adapted
than when it is dark-adapted.
4) The VFT is higher when brighter intensities of
light are used, and lower when weaker intensities
are used. This is the Talbot-Plateau Law.
Intensity is much more important than color of
light in determining the VFT.
5) Higher intensities are unpleasant to look at
for more than a few seconds. (Therefore, lower
intensities were used in the present study.)
6) The VFT is higher when the length of the dark-
period of the cycle is greater, and lower when
the dark period of the cycle is shorter. An
LDR of 1:1 has usually been recommended for this
reason.
7) The VFT is higher when peripheral rather than
central (foveal) portions of the retina are
tested. Therefore, if foveal vision (the usual
and most serviceable vision for accurate and
sustained visual work) is to be tested, a small
65
test patch should be used which subtends a
visual angle between 1 and 1.8° and foveal
fixation should be encouraged by artificial
pupils or some other method.
8) Backgrounds or borders surrounding the test
patch have varying effects on the VFT and hence
should either match the test patch in intensity
or be eliminated altogether.
9) Fairly convenient methods of measuring the VFT
consist of holding constant the intensity of
the light and the LDR while varying the frequency
of the cycle; or holding frequency and intensity
constant and varying the LDR; or by holding
frequency and LDR constant and varying the inten
sity of the light. The first of these three
alternatives is the conventional method and the
one which has received most experimental atten
tion.
10) In determining the VFT, there is an interdepend
ency or interrelationship between at least the
following nine variables: 1) the intensity of
the light (luminance); 2) the duration of each
flash, and each dark period (LDR); 3) the
frequency (F) of each flash, and each dark
period (CPS); and *+) the amount or size of the
retinal area stimulated (size of visual angle);
5) the color of the light; 6) the location of
the retinal area stimulated (foveal versus
peripheral); 7) the adaptive state of the retina
(light or dark adaptation); 8) the duration of
the testing trials; 9) the nature of the im
mediately preceding stimulation (fatigue, effect
of sub-fusional frequencies, practice effect,
set to respond). Any of these variables may be
manipulated in such a way as to counter-act or
facilitate the effects of others.
Age and Sex Differences in VFT
Comparatively few studies have been conducted on the
effects of age and sex differences upon the visual fusion
threshold. Conflicting results have been encountered.
The age range investigated has usually included adolescents,
adults and aged persons. Seldom have studies included
66
children under age eight. And as regards infants, no
studies at all were located in the literature.
The dearth of investigations on infants and young
children may partially explain the conflicting results
reported on the effects of "age” on the VFT. In most of
the studies on this topic, the term "age” is construed as
meaning "old age." Since advanced age usually brings a
lowering of organic efficiencies, it is not surprising
that negative correlations are frequently reported between
age and VFT. In the mid-ranges of adulthood, the corre
lation isiBtally nil or negligibly low. Thus far only two
studies, those by Hartmann (1934) and by Miller (1942)
reviewed below, seem to have investigated the possibility
that age may be positively correlated with VFT throughout
early childhood where organic efficiencies are usually
increasing. Very likely the reason for this is the rela
tive difficulty in securing satisfactory cooperation, con
centration and verbal report from very young subjects.
Hartmann (1934) studied a total group of 31 young
school children of ages 6 to 11, and a group of 32 young
adults of ages IS to 25. These totals were made up of
three series of cases tested under slightly differing con
ditions. Hartmann utilized a rotating cardboard sector
disc half white and half black, or half yellow and half
blue seen by reflected light "coming from a 100 watt Mazda
blue Tdaylight* lamp suspended from the ceiling." (page 22).
67
Tachometer readings of the electric motor speeds were taken
for each subject to measure the frequency at which he per
ceived the transition from flicker to fusion and vice versa.
Three (later five) trials were given in each direction
across the threshold. Deviant results were replaced by
another trial.
In his first series of cases, numbering five girls
(ages 7, 8, 9, 10, 11), six boys (ages 6, 7, 8, 8, 8, 9),
and twelve older males (ages 18-25), Hartmann found higher
thresholds for the children compared with the adults, and
higher thresholds for the boys compared with the girls. In
his next series, he increased the number of both ascending
and descending trials from three to five and used 10 boys
(ages 6-10) and 10 men (ages 19-25). Here the differences
in favor of the youngsters practically disappeared and was
no longer statistically significant. In his third series,
conditions were the same as the second except that blue and
yellow sectors replaced the black and white, and 10 girls
(ages 6, 7, 8) and 10 women (ages 19-25) were used. The
girls were from the second grade and were "of more than
average brightness." (page 125). This series again yielded
a slight but not statistically significant difference in
favor of the children. Hartmann concluded therefore that
"the basic visual mechanisms involved mature relatively
early." (page 126). Regarding sex differences, he con
cluded that the threshold was slightly higher in males in
68
both the age groups studied.
In all, Hartman used only six children as young as
age six. He was interested in testing Koffka's Gestaltist
hypothesis that children's thresholds should be lower than
adults' due to children's greater susceptibility to fatigue.
He therefore explored the feasibility of testing still
younger children. His is the only account of such an
attempt unearthed in this search of the literature. Since
it bears directly upon a basic question for this study it
is quoted below in fulls
It may be argued that these children, whose average
age was about 8 years, are actually adults from the
standpoint of sensory functioning, and that a fair
test of Koffka's hypothesis (lower VFT in children
due to fatigue) would require the use of younger
subjects. Unfortunately, this proved to be an
impractical suggestion, for an attempt to use a few
bright nursery school children collapsed entirely
because of their high distractibility and the exces
sive irregularity of their "Judgments." A child
must evidently be of school age before he can be of
service in this type of experimentation (page 125,
emphasis not in original).
V. L. Miller (19^2) investigated a group of HU boys,
ages 7-11 to 18-7, and 3*+ girls, ages 6-0 to 18-11. These
numbers permitted only 2 or 3 subjects at each year of age.
The subjects viewed a light beam interrupted by a rotating
sector-disk, and signalled the experimenter by pressing a
telegraph key when they perceived flicker change to fusion
(ascending trials only). Trials were always started with
speeds far below the fusion threshold, and five trials per
day were given for three days. The range obtained was from
69
31.6 to 73.0 cps. There was some evidence for an increase
of the threshold with age, though the increase was not
clearly consistent and reliable. The bovs showed higher
thresholds than the girls in each age group. (This corrob
orated the tendencies noted by Hartmann (193*+)» especially
in his first series of cases.) Practice effects were also
interpreted as being present and visible in the slight
decrease of variability in later trials, coupled with a
slight increase in the threshold frequencies.
Simonson, Enzer and Blankstein (19*+1) studied a
group of ky subjects ranging from 10 - 80 years of age,
including equal numbers of males and females. No measures
of central tendency or variability of age are given. Sub
jects were grouped into 10 year intervals up to age *+0.
There were only U cases in the 10-20 interval; 18 cases in
the 20-30; 10 in 30-^0; and 15 cases between *+0 and 80.
The apparatus consisted of a rotating sector disk which
interrupted a beam of light from an electric bulb. The
ranges and averages for each group were determined and it
was found that as age increased, the average and maximum
threshold frequencies decreased.
Brozek and Keys (19*+5), interested in gerontolog
ical and fatigue measures, studied a group of 56 working
women of ages 18-60. They employed a rotating sector disk
interrupting a light beam and took readings during the
first and last hour of the working day. Two ascending and
two descending trials were given at each sitting and the
subject’s score was taken as the average of these four
trials. The age-intervals were 18-25, 26-35, 36-45, and
46-60 years, respectively containing 19, 17, 6, and 12
cases. The respective means (and standard deviations) for
these age-intervals showed a tendency to decrease as age
increased: 46.70 (+4.16), 45.74 (±3*59), 45.39 (.£2.89) and
40.92 (£3*27). The lower score of the 46-60 year group
is significantly different from the other age groups at
the .05 level of confidence.
Father Henryk Misiak (1947) of Fordham University
investigated age and sex differences among 100 subjects,
half of whom were between ages 19-30 and half between 63-87,
with equal numbers of females in each group. The apparatus
consisted of a 3 watt neon glow lamp viewed through a 5 mm
circular window at a distance of 12.8 inches. This gave a
test patch subtending a visual angle of 46 minutes, and a
brightness of 6 foot-candles. Flicker was produced and
controlled by variable resistors, calibrated and monitored
by a small cathode ray oscilloscope. Before testing, the
subjects were adapted to the light intensity of a 3 watt
glow lamp. Misiak concluded that there was a significant
decrease with age between the two groups. (The difference
was demonstrable when tested in either the dominant eye,
non-dominant eye, or both eyes together.) He also con
cluded that there were no sex differences in either his
71
younger or older group. This conclusion is contrary to
that of Miller (1942) above, but the age range in one
investigation was from 19 years downward, and in the other
from 19 years upward.
The above investigation was extended in a second
study by Misiak (1951).^ Using essentially the same
apparatus he tested 182 males and 137 females ranging in
age from 7 to 91. They were grouped into seventeen 5-
year intervals, with about 35 cases each in the first four
intervals, and half that number in the remainder (except
the two final intervals which contained only 6 and 2). No
children younger than 7 were tested. And only a few cases
(approximately seven) were tested at each year level above
this age. The threshold readings covered the fairly large
range of 37.35 cps (from 18,29 up to 55.64), and the
standard deviation for the total group was 4.22 cps. The
mean for the males was 41.18 cps; for the females 41.08
cps. According to Misiak the findings revealed no sex
differences, and a decrease in threshold with age.
. • . The relationship between age and cff is linear
and negative, the correlation coefficient, based on
all individual cases, being significant but small,
-.52. The differences between groups are statis
tically insignificant except between the groups
below 30 and above 55 years of age. . . . The inter
individual variability tends to increase with age,
■^This second study was reported earlier (1948) in
abbreviated form, American Psychologist 3:246.
72
a tendency generally characteristic of many mental
and physical capacities. The intra-individual
variability, on the contrary, shows a tendency to
decrease..." (page 552).
These conclusions differ from the earlier work of
Hartmann (1934) and of Miller (1942) reviewed above.
Misiak*s study contained a greater total number of
subjects but no more at each year of age in the 7 to 11
year interval than did Hartmann*s. Coarseness of grouping
may affect both studies as a factor in their results.
Both studies used a five-year interval over the approxi
mate ages 7 - 11* Hartmann omitted all other intervals
except ages lS-25. Though Misiak does not mention it, a
trend in the direction of increasing scores during youth
is discernible in his data despite the coarse five-year
intervals: The average score for the 7-11 year group
was 42.91 cps. For the 12 - 16 year-olds it rose to
44.&5 cps, which was the highest average for any of his
five-year intervals. For the 17 - 21 year group it fell
again to 43-63 cps. It continued to fall slightly with
each successive five-year interval (except for minor
fluctuations in intervals above 50 years of age and having
fewer than 15 cases). Thus despite his generalization to
the contrary, Misiak*s results may have reflected an
increase of VFT with age up to some point in the second
decade of life. The first decade of life in his study,
as in other studies, is not well sampled since so few
73
children under age 8 are included.
Coppinger (1955) tested 120 men ranging in age from
20 to 79 years. The apparatus afforded a light-dark ratio
of .5, a test-patch subtending a visual angle of 44 min
utes, and three different levels of intensity: .09, .54,
and 1.74 millilamberts. As did Misiak and others who test
ed chiefly mature subjects, Coppinger concluded that the
threshold decreases significantly with age, and that the
decrease in the elderly was most pronounced in the trials
using the highest brightness level.
Summarizing the literature on the effects of age and
sex on the VFT, it is thus apparent that there is no full
and thorough agreement. Sex is variously reported as
showing slightly higher thresholds in males, or as showing
no differences. Age is usually reported as showing a
negative correlation, i.e., a decrease of VFT as age
increases, especially a3 senescence is approached.
Landis and Hamwi (1956) summarize the effects of
age on VFT as follows:
"There are more than a dozen published investigations
of the associations between CFF and age, but there
is little uniformity in results reported, because of
the diversity of methods used, number of persons
tested, and the range of age examined....Various
investigators have previously published correlations
between CFF and age ranging from 0.00 to -0.74, the
values seemingly depending on the age-range and age-
composition of the group studied.••.About all that
can be concluded from the studies is that there is
a tendency for CFF to be negatively correlated with
chronological age." (page 4o0).
In the really early stages of the maturation process,
that is, in young age as contrasted with old age, changes
in individual VFTs remain virtually unexplored. There
appear to be no systematic studies whatever of VFT in
infancy, that is in ages 0 to 2. The same conditions
hold true also for the important years of early childhood,
covering approximately ages 2 to 4- This is a noteworthy
lack of data in view of the fact that children of these
nursery school or pre-school ages are fairly readily
accessible for testing and research. The lack is probably
explainable by the relatively high level of cooperation
and comprehension required of the subjects, and by
extraneous and distracting features of older apparatus
such as Hartmann’s (1934) and Miller’s (1942). Less
explainable is the lack of systematic coverage of the
kindergarten and early elementary school ages with modern
equipment. Few studies have been done at these ages and
even these have used only 4 or 5 cases at each year of age.
These studies have been variously interpreted as showing
an increase in threshold with age, a decrease with age,
or no relationship to age. There is similar uncertainty
with respect to possible sex differences.
Moreover, all of the studies reported thus far appear
to have been done with a cross-sectional approach, rather
than a longitudinal approach to changes with age. Landis
(1954) stated: "There is no systematic study of the
change in CFF in the same observers over a period of one
75
or more years" (page 282). It is quite possible that
maturational changes in the individual’s VFT are masked
when readings are measured and averaged for groups of
people at various ages instead of being taken on the same
individuals repeatedly over a period of years. It appears
that much more work is needed with the VFT in the early
years of human development in order to answer basic
questions of age and sex differences in the maturation of
this important visual function.
VFT Studies and Intelligence
Considering the number of studies of intelligence, and
the interest in that topic, it is remarkable that there
are so few studies relating it to the VFT. In the summary
by Landis and Hamwi (1956) quoted earlier, the following
statement is made with respect to the factor of
intelligence.
"Not enough relevant and comparable information has
been included in these publications to ascertain
whether the CFF is or is not affected by intelligence
....Only Tanner and Colgan have reported significant
correlations between CFF and indices of intelligence^
no confirmation of their results has been found.
There is no obvious reason why intelligence should
act as a determinant of CFF. There are enough
established determinants of CFF to suggest that the
correlations reported by Tanner and Colgan are
examples of random fluctuation." (page *+6l)
In view of this statement by Landis and Hamwi, a
review of the studies by Tanner (1950) and by Colgan (195*0
seems warranted.
Both phases of the study by Tanner (1950) were under
taken and reported first as an unpublished master's thesis
at the University of Florida. The apparatus and method
ology employed seem basically sound, though somewhat
different from most previous studies. Essentially Tanner
tested the VFT by holding the duration of the light flash
constant and varying both the frequency and the LDR by
lengthening or shortening the dark periods. He connected
two single-cycle electric multi-vibrators so that each
quickly interrupted the current from the other. The
length of the light period, which remained constant, was
controlled by the current from one vibrator, and the
length of the dark period, which was varied, was controlled
by the current from the other vibrator. The light flashes
were produced by fluorescence from a gas discharge tube,
and their constancy was monitored by means of a photo
electric cell which picked them up and transformed them
into an image on an oscilloscope, thus permitting visual
correction and adjustment.
In the first phase of the study four durations of
light period were used: 8, 16, 38, and 8U milliseconds.
In the second phase these were supplemented by two longer
durations of 135 and 250 milliseconds. With these
arrangements Tanner measured the shortest noticeable dark
period which his subjects could perceive as flicker. At
each light duration tested (8, 16, 38 milliseconds, etc.)
the light-portion of the period remained the same, but the
77
dark-period grew shorter and shorter until the testee
could see no flicker. The frequency varied also of
course, since the smaller the duration of the dark-period,
the higher became the frequency. In Tanner's procedure,
therefore, lower scores (shorter dark periods) and
negative correlations were comparable to higher scores
(higher frequencies) and positive correlations in other
investigations, since Tanner measured the decreasing
dark periods instead of the Increasing frequencies.
For two groups of male college students, numbering
25 in the first series (light periods of 8, 16, 3U,
and 8U milliseconds) and 21 in the second series (135
and 250 milliseconds) Tanner found the dark period
scores correlated negatively with the raw scores on a
group intelligence test. The intelligence test used
was the American Council on Education (ACE) Psyshological
Examination, College Edition. The correlations were
lowest for the 8, and 135 millisecond light-periods,
highest for the 8^ millisecond light-period. For the
second series of students, who were tested only in the
morning one hour after breakfast, and at a lower light
intensity, the correlations were -.1 +78 with the Quanti
tative section of the test, -.532 with the Language
section, and -.*+85 with the total raw score. These
correlations were significant at the .05 level of
confidence. Tanner reached the following conclusions:
7$
"The correlation between the visual threshold
measures for the 34 millisecond light flash
and the ACE scores are strikingly high in view
of the homogeneity of the groups studied..•
From the results of this experiment, it may be
concluded that the shortest noticeable dark
period for a light flash of some critical length
promises to be a significant physiological
correlate of intelligence.* * (page 203).
Colgan (1954), using the Wechsler Bellevue and the
VFT, tested 40 elderly men in a home for the aged. Of
these men 13 were in the 65 - SO year range and 22 in
the extremely elderly 31 - 95 year range. As might be
expected from the discussion of old age in the previous
section, the correlation between VFT and age in these
elderly men was -.32. However, there was also a still
higher negative correlation of -.54 between age and total
Wechsler Bellevue IQ; and when this latter correlation
was partialled out, the correlation between VFT and
age fell to -.10. When age was partialled out, the
correlation between VFT and IQ was +.36. Significances
of all the correlations (except that of -.10) were at
or above the .05 level of confidence. Colgan*s con
clusions from this were as follows:
“There is a significant relationship between
CFF and intelligence which is independent of age;
and previous studies relating CFF to age must be
reexamined to determine whether or not intelligence
was a controlled variable— -if it was not, the
conclusions of those studies must be qualified
accordingly. It may be that the relation of age
to intelligence, and to CFF when intelligence is not
a controlled factor, stems from the fact that after
the peak of development is reached, the passage of
time provides opportunity for the accumulation of
physiological accidents and for events that contribute
79
bit by bit to the impairment of neural functions upon
which both intelligence and CFF depend.*1 (page 713)*
Halstead (1947) appears to be the only other con
tributor to the topic of VFT and intelligence. However,
his conception of the term "intelligence** differs a good
deal from the usual one, which seems to him to be too
dependent upon educationally and culturally determined
processes. Halstead prefers to speak of biological
intelligence. For him this term denotes a broad range of
human responses and capacities determined or naturally
defined by the neurological structures and functions of
the brain. These he subsumes under four factors which
he believes to be the fundamental factors of biological
intelligence. They were derived by factor-analysis from
a total battery of 27 tests most of which he classified as
neuro-psychological.
Halstead regards the VFT as a direct and sensitive
measure of the P or **power-control** factor^. He conceives
of this factor as neural processes responsible for control
of impulses and emotional tensions, and for the focusing
of attention and concentration. Thus, the function of the
"power-control** factor is to permit or regulate the
iThe other three factors are: G, a central-
integrative function of the brain; A, an abstract-
categorization ability necessary for logical and associa-
tional operations; and D, a directional (left-right,
etc.) factor.
80
engagement of the three other factors or processes.
As measured by the VFT in cycles per second, Halstead
found various amounts of P in the following groups of
subjects, arranged in decreasing order: Military controls
24*6, normal controls, 23*2; nonfrontal lobectomies 21.7,
emotionally disturbed under psychiatric care 18.6, frontal
lobectomies 17*2. The difference between the means for
the controls and the nonfrontal lobectomies is significant
at the .05 level; between the controls and the frontals,
the difference in means is significant at the .001 level.
Halstead certainly sees the VFT as correlated with
intelligence though his use of the latter term is not
typical.
In the studyl by Doehring, Ward and Hixon (1956)
significant differences in VFTs between two groups of
naval servicemen were reported. For 100 naval aviation
cadets, the methods and apparatus used yielded VFTs
centering around 40-43 cps with a mean of 41.51? whereas
for 174 enlisted men, similar (with exceptions) to the
cadets in age (17-25) but mainly engaged in aircraft
maintenance work, the comparable figures were 38-41 cps
with a mean of 39*79* The difference between the two
groups of men was significant at greater than the .01
■^Cited earlier as an example of group testing,
page 56.
61
level of confidence. Though this study does not explicitly
discuss the point, presumably the cadets were the some
what more intelligent group. Hence this study probably
furnishes additional evidence to indicate that intelli
gence and VFT may be correlated.
In attempting to summarize the relationship between
intelligence and VFT, it would seem that there is much
room for further investigation. At least three investi
gators have directly reported some type or degree of
correlation between these measures. One of these in
vestigators examined young adult males with a group
intelligence test; another utilized elderly males and
an individual intelligence test; a third made use of
factor-analyzed intelligence constructs. A fourth study
implicitly indicated an association between IQ and VFT
by showing higher threshold scores in a presumably more
intelligent military group.
Therefore it appears that Landis and Hamwi (1956)
may have taken a somewhat extreme position in holding that
there is no reason to suppose a relationship between VFT
and intelligence. By Landis* own earlier statement (1951)
that VFT is lower in cases of brain injury, some degree
of relationship might be expected between low VFT and
exogenous mental retardation. Indeed this has been
reported, as is seen in the next two sections covering
anxiety and organic pathology.
B2
VFT and Anxiety or Emotional Conditions
With respect to anxiety, a number of studies agree
that this factor or process is related to lower VFT.
Krugman (1947) compared two groups of Air Force
combat returnees, one composed of normals, and one composed
of those with anxiety reactions known as "operational
fatigue". Although there was considerable overlap in
the VFT scores, those for the anxiety cases were on the
whole significantly lower. Moreover this appeared to be
at least partially reversible; for as the anxiety
symptoms abated, the lowered VFT scores tended to rise
toward presumed former levels.
Friedl (1954) found a correlation of -*34, signi
ficant at the .05 level of confidence, between VFT and
scores on the Taylor Scale of Manifest Anxiety.
Goldstone (1955) in a resume of his doctoral
dissertation, reported VFT differences significant at the
.005 level between one group of 35 patients who rated low
on an unpublished anxiety-tension rating scale developed
by M. Lorr and a similar group of 39 patients who rated
high on the scale. The cases came from the outpatient
psychiatric clinic at Duke University.
One systematic limitation or possible source of
error with which these studies of actual patients have
had to contend is the fact that VFT scores were not
available prior to the "onset" of anxiety. Thus it could
33
not be directly proved that anxiety had lowered the VFT.
A study by Kushner (1955) sought to circumvent
this difficulty by inducing stress experimentally in
normal subjects and comparing their pre- and post-stress
VFTs.
To induce stress, Kushner utilized the following
methods or procedures: 1) delayed speech feed-back test
(reading a fairly difficult selection with the sound of
one’s own voice being fed back .3 seconds later by ear
phones and a modified tape recorder); 2) mirror drawing
task (tracing a difficult pattern by viewing it in a
mirror, being interrupted and informed from false norms
that performance was inferior); 3) electric shock of 3
milliamperes, unexpected.
Four groups of college students were utilized, one
for each of the three stress tests and one control group.
Each group numbered 15 males and 15 females and received
three VFT trials. In the experimental groups the last
VFT trial followed the stress.
In comparison with the control group, all three
stress groups showed decreases in VFT after stress. The
decreases were statistically significant at the .01 level
of confidence. This experimental evidence thus appears
to support the clinical evidence that anxiety, or at
least some of its effects, can be detected by changes in
the VFT.
8k
Vier (1956) has also demonstrated experimentally
that the VFT is lowered by strenuous mental tasks or
by situations which appear to threaten the subject's
social prestige.
Summarizing this section of the literature, there
seems to be agreement at present among a small number of
later studies that anxiety is associated with lower VFT
scores. Clinical studies have had to contend with the
fact that for their subjects VFT readings were not
available prior to the onset of the anxiety which
supposedly lowered them. Much longitudinal case study
remains to be done in order to clarify such processes or
conditions as: 1) reduction of VFT prior to or
simultaneously with anxiety, 2) physical exertion,
fatigue, exhaustion, 3) illness or injury, and/or k)
purely psychological processes other than anxiety. The
clinical studies have received some support from a small
number of recent experimental studies. The latter have
shown that the VFT can be reduced temporarily by ex
perimentally induced stresses of a psychological or emo
tional nature.
VFT and Organic Pathology
It is well established that organic pathology and
other variations in physical condition can lower the VFT,
and that drugs or certain metabolic substances can raise
it. Landis (1951) stated that the VFT is influenced by a
great number of biological and psychological conditions
such as body temperature, vitamin deficiencies, central
nervous system injuries, and many chemicals or drugs.
This statement is supported by studies too numerous to
cite in detail. Among them are studies by Enzer and
Simonson (1940), Werner and Thuma (1942), Simonson, Fox,
and Enzer (1943), Halstead (1947), Teuber and Bender
(194$), Bender and Teuber (1949), Battersby (1949),
Schaefer and Carey (1954), Reuning (1955), Stauffacher,
Marks, and Ax (1955), Halpern (1957), and Mark and
Pasamanick (195$). The only point of disagreement that
appears to exist at present is the failure of a study by
Keller (195$) to confirm the earlier finding of Werner
and Thuma (1942) that VFT differentiates between two
sub-types of mental retardation. The point in question
is not whether VFT is low in retardation; this seems
agreed. The question is merely whether the VFT is
lower in the traumatically produced type (exogenous) than
in the genetically produced type (endogenous). Part of
the difficulty in answering this question may lie in the
difficulty of obtaining two completely »pureT criterion
groups to test.
Summary of Pertinent Points
in the ^FT Literature
To summarize this review of the VFT literature with
S6
respect to those aspects most pertinent to the problem
dealt with in the present study, the following points
may be listed:
1) There is no standard apparatus or equipment
available for use by various investigators.
Each investigator must construct his own
apparatus and report its stimulus features
in detail (for example, intensity of light,
minutes of visual angle, LDR, and so forth).
Many current investigators are using strob
oscopes as a basic part of their apparatus
even though these do not give a constant LDR.
Various aspects of the LDR, both theoretical
and practical, are still quite confused and
uncertain.
2) No standardized testing procedure has been
evolved or adopted and given widescale use.
Only one study was reported using the "method
of constant stimuli." This is probably a better
technique than the "method of varying stimuli
or just noticeable difference (j.n.d.)", which
imposes a harder and more complex perceptual
task on the subjects and yields more variable
results. Despite this, most present investiga
tors use the j.n.d. method because it is
simpler to instrument, to administer and to
87
score* Standard numbers of trials, standard
conditions of dark-adapted vision, standard
control of luminance, visual angle, diurnal
variations, temperature, ventilation, sound
background, and so forth have not been agreed
upon and observed.
Only two reports of group-testing methods
have been discovered and both of these were
utilized with adults only.
3) As a result of the foregoing points, there
is no possibility for direct comparison of
results from one study with results from
another. Each testee’s score is meaningful
only within the distribution of other scores,
all of which are by the same investigator
with the same apparatus and under the same
test conditions. Hence no norms of any kind
have ever been established for VFT scores.
4) Regarding sex-differences, there is confusion
and uncertainty as to whether male thresholds
tend to be slightly higher than those for
females. The differences, if any, are probably
small, and there is a tendency in adult studies
to deny them.
5) Regarding age-differences, there is fair
agreement that VFT tends to decrease with old age.
But there is marked confusion and disagreement
about whether the VFT increases during early
maturation. No studies whatever seem to have
been done with infants. The same is true for
children of nursery school ages. Only three
studies have included children of early
elementary school ages, and most of these
were children eight years of age or older.
6) None of the age studies followed up the same
cases with testing after an interval of one
or more years. Test-retest reliability of
the VFT over a lengthy interval appears
unstudied.
7) There is marked conflict as to whether the
VFT is or is not associated with differences
in intelligence. Several investigators have
conducted studies which implied or directly
claimed a correlation between these variables.
Others have claimed that there is no relation
ship, and no rationale for any.
£) It is well established that the VFT is a very
sensitive measure of visual perceptual
efficiency. There is wide agreement that any
condition which interferes with the efficiency
of brain or retinal tissue is likely to result
in a lowered VFT. The obverse proposition is
also widely accepted: that a lowered VFT is
likely to indicate some unhealthy or inter
fering tissue condition. It is not expected
or claimed however, that a low VFT by itself
can determine or diagnose which one of many
possible interfering conditions is present in
a given case. The evidence that the inter
fering conditions can consist of many physio
logical states is extensive. There is also
a fair amount of recent evidence that psycho
logical states such as conflicting stimuli,
stress or anxiety, can also reduce the VFT
at least temporarily.
9) With respect to difficulties in written
language facility (reading and writing) there
are few or no studies on the VFT except those
of Tait (1956) and Smith and Carrigan (1959).
However, these two studies are quite recent
and bear directly upon the basic concerns of
this report. The first of these is reviewed
in the preceding chapter (page 13). The
second was published after the inception of
the present report. Both of these studies
present strong evidence that the VFT is a
good indication of reading disability in
subjects in middle or later childhood, or in
early adolescence, that Is, from ages about
3 to 14, corresponding to the intermediate
and upper grades in elementary school. But
these studies did not deal with the questions
of how early the VFT might be used; or whether
it might indicate reading difficulty at earlier
ages; or preferably, whether it might predict
reading difficulty that is still potential rather
than actual.
CHAPTER III
PROCEDURES
This chapter describes the population selected, the
schedule and numbers of children tested, the measures and
apparatus making up the test battery, and the procedures
followed in administering the battery.
Population Selected
Three factors were chiefly responsible for the
selection of the particular school district for this
study.First, the district is of relatively small
size, which meant that an attempt could be made to test
its entire population of kindergarten and first grade
children. Second, the population appeared relatively
free from extreme characteristics of a socioeconomic
nature. Third, owing to prior working relationships
with the district's administrative and teaching per
sonnel and to their interest in research, conditions
were favorable for securing official approval.
The school district is located approximately
*The Rosemead Elementary School District. Thanks
are extended to the administrative and teaching personnel
of this district for their very kind and continuous assist
ance, and especially to Dr. Rodney S. Mahoney, Superintend
ent and to Dr. Arthur A. Attwell, Director of Guidance.
91
92
twelve miles due east of metropolitan Los Angeles,
California. It lies between some very expensive and
restricted residential areas to the north and west, and
some mixed residential, industrial, and agricultural
areas to the south and east. It is bounded on the south
by a busy freeway, and is divided roughly into four
quadrants by two main thoroughfares which cross near its
center.
In 1959 the area became an incorporated city with
a population of approximately 16,000. Its size is
2,355 square miles. It contains approximately 6100
dwellings, of which about 5500 are of the single family
type, 300 are duplexes, and 300 are of the ’multiple*
or apartment type. Along its thoroughfares are roughly
225 commercial establishments; at least three quarters
of these are classifiable as business or light industry;
the balance is mainly semi-heavy manufacturing or
fabricating plants.
The school district boundaries are practically
coterminous with those of the city. The school system
is composed of four elementary schools including levels
from kindergarten through grade six, and one intermediate
school for grades seven and eight. On standardized
group tests of intelligence the average score for the
district over a period of several years has been
93
approximately 108, with a similar standing on school
achievement tests. Children of Mexican-American parentage
compose about ten per cent of the school population.
There are one or two dozen Negro families and a similar
number of oriental families.
In attempting to arrive at an overall evaluation or
classification of the district population, perhaps it
is best characterized as a mid-type suburban community
with a few features which are slightly above average.
Testing Schedule and Numbers Tested
Testing was carried out in two periods occupying
the last two weeks of school in two successive years.
These two testing periods are hereafter referred to
respectively as Year I and Year II. In the Year I
test-period (June 195&) the subjects were finishing
the last two weeks of kindergarten. One year later
the pupils tested in the Year II test-period (June 1959)
were of course finishing the last two weeks of first
grade.
In both testing periods, an attempt was made to
test all children in the grade and to proceed at about
the rate of one full class per day. This general plan
was accomplished, but with appreciable exceptions. The
exceptions were due to the following factors: 1) ill
nesses and absences of the children or other key per
sonnel! 2) interruption by school or PTA assemblies, or
94
mechanical or electrical failures; 3) 1 spoilage* of some
test results (one entire first grade class was discarded
because of lack of "cubicles" for group-testing, as
described below in the section on group-testing pro
cedures); 4) "missed cases" due to failure of the admini
strator to test all children on the VFT apparatus before
dismissal time for the class; 5) occasional interference
by other commitments of the investigator.
In Year I, ISO children were tested, constituting
about 75 per cent of the kindergarten population. In
Year II, 205 children constituting slightly more than
SO per cent of the first grade population were tested.
Ninety-two (92) of the cases tested in Year I were
also re-tested in Year II, i.e., were included in both
the kindergarten and the first grade testing.
Selection of the Test Battery
The questions raised for answer in this study
dictated the choice of tests selected for the battery.
The VFT test of course had a major place in the
battery, since the primary interest of the investiga
tion was to see whether it could be applied at early
age levels and, if so, whether it could then predict
future reading disability.
A test of later reading ability was the next most
important requirement. This need was met at the third
grade level by the California Achievement Test. This
95
test was given routinely by the district every two years
commencing at this level.
In order to check on the pupils’ reading ability at
the same time as the VFT test was given, the Harsh-
Soeberg Survey of Primary heading Disability was chosen
for use at the first grade level. This test was selected
because it contains enough easy reading or pre-reading
tasks to compare with a reading readiness test (kinder
garten or early first grade level) and also enough
harder reading tasks to challenge pupils of good ability
in second or even third grade.
In order to investigate whether the VFT test is
related to the mental ability of pupils of these ages
an intelligence test was also included in the battery.
Because of its sampling of identifiable factor-analyzed
abilities, the choice fell on the Primary Mental
Abilities Test for Age3 5 - 7 (195A-)• In addition to
a possible relationship with "total scores” of intelli
gence, it was thought that the VFT test might be parti
cularly correlated with certain PMA subtests such as
those of perceptual speed (factor P) and of visual
manipulation (factor S).
As a check on the possibility that teachers1
opinions of reading ability might prove to be related
to the VFT test, teachers* ratings of each first grade
child were obtained for the battery.
96
Although not a part of the battery administered by
the investigator, another important measure was automati
cally provided by the district*s regular testing program.
A reading readiness test, the Lee-Glark, was routinely
given to all pupils at the beginning of first grade. This
test was administered by the first grade teachers. It
followed the kindergarten VFT by about four months and pre
ceded the first grade VFT by about seven months. The
California Achievement Test, previously mentioned, was also
given routinely to all third grade pupils by their teachers
at the end of October. It therefore followed the kinder
garten VFT test by two years and four months and the first
grade VFT test by one year and four months.
The VFT Apparatus and Equipment
The VFT apparatus employed is similar to that
utilized by Tait (1956) in his work with fourth, fifth,
and sixth grade pupils. The two chief components of the
apparatus consist of a stroboscopel commercially obtained,
and a viewing tube constructed by the investigator.
^Type number 631-BL (Serial number 25,194), brand
name “Strobotac,* manufactured by General Radio Co., 275
Massachusetts Avenue, Cambridge 39, Massachusetts, U.S.A.
The type of neon bulo used was type 631-PI, brand name
"Strobotron,* manufactured by Sylvania Electric Products,
Inc., Woburn, Massachusetts.
97
The stroboscope (Figure 3.1) presented light flashes
of a nearly constant and extremely short duration. From
correspondence with the manufacturer, and from the
company's technical manuals, it was learned that the
duration of each flash is only 10 - 40 microseconds,
i.e., 10 - 40 millionths of a second. With this small
variation (30 microseconds) and with the threshold
frequencies ranging as they did from a low of 6 cps
to a high of 30 cps, the LDRs (light-dark ratios)
could not have varied more than from a low of 0.6 :
10,000 to a high of 12 : 10,000 cps.
Historically a constant LDR of 1 : 1 has been
regarded as preferable; however, this was probably a
matter of convention or convenience associated with
the solid-epen proportions of sector-disks, or with
the psychological and mathematical advantage of using
such a simple ratio. Moreover, great doubt has recently
been cast on this preference, or on the necessity for
any particular LDR, as long as it does not change
markedly or in a haphazard manner within or between
trials (Landis 1954, page 274)* Thus, good quality
commercial stroboscope units such as the one used in
the present study are being used by many other current
investigators. Commercial stroboscopes are precision
instruments which ordinarily do not vary from one another,
nor from their own previous performance, by more than one
LUMINOUS DIAL
INOICATING- FLASH RATE
(
3
LINE DIRECT
M l O H N . xLOW
LC W v
<^p
OFF ^ ^ N REED
8"
^OWER INPUT JACK ^-SELECTOR SWITCH
. H m ---------»
SIDE VIEW
ADJUSTING KNOB POR
CHANGING- FLASH RA TE MANUALLY
LENS (5“ DIAMETER)
PROHT VIEW
FIGURE 3.1 STROBOSCOPE
99
or two per cent. Hence, the conditions created by these
instruments remain practically constant and comparable
for all subjects in an experiment.
The light impulses emitted by the stroboscope used
in this study consist of intermittent orange-colored
flashes of fluorescence from the neon gas in the light
bulb. The fluorescent flashes of course are caused
by intermittent bursts of electrons passing through the
gas from the cathode of the light bulb to the anode.
As far as human perception is concerned, the
fluorescence of the gas is virtually an instantaneous
and an all-or-none process. Nevertheless some mensurable
interval is required for the build up and also for the
fade-out of the light flash. Therefore, if graphed the
light impulses resemble a pulse spike followed by a long
low plateau and a final slow drop, rather than a
"rectangular” wave pattern.1
The viewing tube employed (Figure 3.2) was con
structed from a 2& inch section of heavy cardboard tubing,
the inside diameter of which was 5 inches and the out
side diameter 5& inches.^ The inside surface was care-
3-For brief discussion of wave patterns, see
Chapter II.
^Such tubing is often procurable from rug factories
or furniture dealers where it is used as a core upon
which to roll up lengths of carpet.
a
CAAAY 1 N Q
H A N P L E
_Letf.5 ASSEMBLY
(see etc't'o.i I
« > » Fifri-3)
W o o p e ti a d a p t e r - H o l d e k E ye-P iece: r u b b e r
SWIM MASK, LENS
Re M o v e D
Wo o d e n a d a p t e r - h o l d t r } f o r
F I T T I N G T U B E t o S T R O B O iC O P E
0
T U B E
F ROMT V I E W
A d a p t e r - holder ,
FRONT view
FIGURE 3.2 V IE W IN G TUBE ASSEMBLY
100
101
fully lines with a layer of black construction paper
having a soft dull surface rather than a glossy finish.
This provided a light absorbent non-reflective surface
for the inside of the tube.
Mounted at one end of the tube were parts from a
childfs black rubber under-water diving mask from
which the lens had been removed. This created a
comfortable and snug-fitting eyepiece which conformed
to the subjects forehead, temples, and bridge of the
nose. It served to position the subject quickly and
accurately and to shield his eyes from any source of
light other than the test-patch at the opposite end of
the tube. Since the subject’s nose and mouth projected
below the mask, there was no interference with breathing.
Where the edges of the mask made contact with the skin,
sanitation was maintained by occasional cleansing with
alcohol or a mild commercial antiseptic.
The test-patch assembly (Figure 3*3) at the opposite
end of the viewing tube was constructed in a manner
similar to Tait’s (1956). It consisted of a diamond
shaped aperture measuring 1 inch in height by 7/& inches
in width cut into the center of a 5 inch circle of black
construction paper. The visual angle subtended on the
retina by this test-patch was approximately 2 degrees
and 3 minutes (twice the tangent of 0. 5* ,/2fS**).
MILK GLASS" L E N5E5
5"
BLUE FIXATION
D OT
VJRATT E H N E U T R A L
D E N S I T Y F I L T E R S
c
FIGURE 3.3 T E 5 T - P A T C H ASSEMBLY
Over the aperture, to reduce the intensity of the
light from the stroboscope and to neutralize its orange
color, was placed a gray photographic gelatin filter
(neutral density Wratten Filter #61) similar to the one
used by Tait. The filter and aperture-paper were then
sandwiched between two flat circular glass plates or
lenses to protect and hold them rigidly. To match the
inside diameter of the viewing tube and to fit snugly
within it, the lenses were also cut to a 5 inch diameter.
They were cut from a pane of commercial white-frosted
"milk glass** and thus aided the filter in diffusing and
controlling the intensity and color of the light. On the
inner surface of one lens, and in the center of the
aperture area, was pasted a small "fixation** dot of
black paper. Like the aperture, it too was diamond
shaped, but it was only 1/4 inch tall and 3/16 inch wide.
Its purpose, as with Tait»s apparatus, was to provide a
specific point upon which the subjects could fixate
while looking at the test patch. This was believed to be
conducive to a steady gaze and maximal use of the foveal
area of the retina.
Preliminary trials with this test patch assembly
produced a light which, in the opinion of the investiga
tor and other adults, was not restful or comfortable to
look at for more than a few seconds. Small children
104
tested In the investigator's neighborhood made remarks
about it "hurting" their eyes. Also at certain fre
quencies there were suggestions of color changes at the
edge of the test patch. Therefore, it was decided to
reduce the intensity of the light still further and to
eliminate if possible all vestiges of discomfort or
suggestions of color effects at the border of the test
patch.
After trials with various modifications, a satis
factory result was obtained, in the subjective opinion
of the experimenter and of several adults and children,
when a second gray filter (Wratten #61) and a sheet
of light blue sixteen-pound mimeograph paper were added.
The illumination level, as seen by the testee at the eye
piece of the viewing tube, barely registered 1 or 2 fc
on a light-meter.1 At the source of the light,
measured at the lens of the stroboscope and before
passing through the test patch and the viewing tube,
the luminance registered about 6 fc at low frequencies
(6 cps) and about 32 fc at fusional frequencies (20 cps
and above).
Figure 3*4 shows a diagram of the fully assembled
■^Universal Exposure Meter Model 745 (Serial
#12557, brand name Weston Master IV) manufactured by
Daystrom, Inc., Weston Instrument Division, Newark,
New Jersey, U.S.A.
FIGURE 3 . 4 VFT APPARATU5 A5SEriBLEP
So
106
stimulus apparatus. The test-patch assembly was snugly
fitted into the viewing tube and held there securely
by black tape. The viewing tube was positioned and
supported on the table by two wooden adapter-holders
(#1 and #2 in Figure 3.2). Adapter-holder #1 merely
supported the viewing tube near the eye-piece. Adapter-
holder #2 held the other end of the viewing tube and was
joined to the stroboscope by means of heavy elastic
bands. Light leaks were prevented at this union by
lining the inside edge of adapter-holder #2 with a sponge
rubber grommet and black tape.
Transportation of the apparatus was facilitated by
simply removing the two heavy rubber bands from adapter-
holder #2. The two main components, stroboscope and
viewing tube, could then be carried separately and
easily, one in each hand.
In Year II additional apparatus was constructed in
order to control dark-adaptation of the subjects. The
apparatus consisted of a dismountable testing booth.
The booth was cubic in shape and measured 6£ feet
in each dimension. It was constructed fairly simply
from panels of heavy black plastic stretched over a
framework of metal bars. The construction is shown in
the diagram, Figure 3*5- The bars were 6i foot lengths
of light-weight steel tubing of the type ordinarily
" rC o R N E R l e f t LOOSE FOR D o o r
VeNTlUTWGr SLITS IK IN K E R CEIUN&
i r "
3 " LIGHTING A M D VENTILATION SAP
SlbE VIEW FRONT VIEW
FIGURE 3.5
DIAGRAM O F TESTING- BOOTH FOR M EASUREM ENT OF V F T
103
used for electrical conduit. They were coupled with
elbows and with compression fittings which held securely
when tightened by hand or by very light pressure from
a wrench.1 The black plastic sheeting was approximately
l/64 inch thick and completely opaque to even the strong
est light.2 As shown in Figure 3*6 the plastic was cut
into two rolls or panels, one slightly longer than the
other and each panel was stretched over the bars of the
booth to form two of its sides and one of its two (inner
and outer) ceilings. The shorter panel was used for the
left and right sides and the inner ceiling. The longer
panel was used for the front and back sides and the
outer ceiling.
Lighting and ventilation were provided by the
following means: a) by leaving a three-inch gap between
the floor and the bottom of the rear wall; b) by slitting
elliptical holes in the inner ceiling; and c) by sloping
the inner ceiling upwards towards the front, and sloping
the outer ceiling upwards towards the rear. By con-
■^The conduit tubing and the compression fittings
were purchased at a local hardware store.
2The black plastic was purchased by the yard from
a roll 54 inches wide from a local millinery store.
Seams (a 54 inch and a 27 inch width were joined) were
made very simply and satisfactorily by joining the edges
with masking tape. Sleeves through which the bars could
be slipped were also seamed in this way. Cost of the
booth was approximately forty dollars for materials.
PANELS W E R E O R IE N T E D A S SHOW N, AND ASSEMBLED
AS IN FIGURE 3 .5
\ ^ &
FIGURE 3 .6
DETAIL. OF P L A S T IC P A N E L S FOR T E S T B O O T H
109
110
vection cool air entered low through the gap at the rear,
and as it grew warmer it moved upward and forward, then
through the elliptical slits of the inner ceiling, and
finally out of the booth at the top rear. At the corners
of the booth, the edges of the plastic panel were secured,
and released, or re-secured to the metal bars in light
tight fashion by use of masking tape.l
The time required to erect the booth was approxi
mately fifteen minutes, and a little less to dismantle
it. After familiarization, one person could accomplish
the task, but it was facilitated if another person
assisted. The booth was moved into each first grade
room on the afternoon preceding testing in that room.
It was carried from school to school quite easily in a
station wagon or passenger car. The booth was custom
arily set up in the rear of the classroom and did not
occupy enough space to interfere with regular classroom
activities. The door flap was turned towards the rear
wall so that children did not attempt to peek in or out.
Under operating conditions, two to five children
at a time occupied the booth for their turn to be
tested. They sat with their backs toward the rear
this purpose sin „ - - .
the bar or the plastic without appreciable loss of
adhesiveness.
1
It proved possible to use the tape re
Ill
wall so that the small amount of light from the three
inch lighting and ventilating gap at the bottom of that
wall was not directly visible to them. Nor was the light
from the luminous dial of the stroboscope visible to
them directly. The diffused and indirect light from these
two sources was the only light in the booth when the
door-flap was closed. By measurement with a light-meter,^"
the illumination read approximately zero fc when the
meter was directed at the walls or ceiling; approximately
0.2 fc when directed at the floor; and approximately 0.3
to 0.8 fc when directed at the light and ventilation gap.
It varied only momentarily when the door-flap was pushed
aside as a child entered or left the booth. These con
ditions were such as to allow just sufficient light for
orientation within the booth, and sufficient darkness
to produce at least partial dark-adaptation in the
subject during the five minutes or more while he was
waiting for his turn to be tested.
In Year I, only an approximation of the test booth
conditions was possible, since the booth had not then
been designed or constructed. Dimly lit hallways, cloak
rooms, or storage rooms were used as substitutes. Need-
"Hjniversal Exposure Meter Model 7*+5 (Serial
#Y2557* brand name Weston Master IV) manufactured by
Daystrom, Inc., Weston Instrument Division, Newark,
New Jersey, U.S.A.
112
less to say, they were less satisfactory and more
variable. The illumination level in them varied from
about 6.5 to 8.5 foot-candles. With respect to dark-
adaptation in the subjects prior to testing therefore,
the environmental conditions for Year I were roughly
similar to Tait's and the subjects could probably not
be considered as dark adapted. Kindergarten classes
1 and 2 constitute exceptions, for they were tested
in a storeroom which had black shielding over its
windows and was therefore quite dark, measuring only
about 0.9 fc on the light meter.
Procedures for Testing the VFT
Preliminary trials with young children revealed
the necessity of keeping verbal instructions at minimal
length and language level. Following are the standard
procedures used and the standard directions read (or
later given from memory) by the investigator.
"This is the machine that makes the tricky
little light for the eye-test game you were told
about. Look right in there (eye-piece of viewing
tube) and you will see the little light going on
and off. (Stroboscope was set at approximately
6 flashes per second). Can you see it now?
[Ttnswer: Yes^] Now I can make it go faster...
and faster. (Examiner gradually increases
frequency.) Is it going faster now? [Answers
YesT] That's right. I can make it go still
faster...and faster — until pretty.soon it
stops moving and just stands still:1
^After preliminary experimentation, the words
113
(Stroboscope set at speeds in excess of 50
flashes per second.) What is it doing now?
Is it moving or standing still? (Answer:
Standing stiliTT That’s right.
"Now keep watching and tell me when it
starts to move again. When it moves just a
little teeny bit, you say ’Now.* (Examiner
starts ’descending* trial, reducing flashes
per second rapidly until fusional frequencies
are approached and then continuing the re
duction at a slower rate. When the subject
says ’Now,’ or ’It’s moving a little bit.’
examiner notes the dial-reading and says)
All right. Now I’m going to make it move
more and more...(Examiner decreases flash
rate further) until it moves a lot. Now
it’s moving a lot, isn’t it? (Flashes
are now back at 6 per second and subject
answers ’Yes.’)
"Now we’re going to do that five more
times. You keep watching and tell me when
it stops moving. Wait ’til it stops moving
and stands still and then say ’Now.’
(Examiner slowly increases flash rate ’til
subject says ’Now’ or ’It’s stopped.*
Examiner notes the flash rate at that
instant, and writes it down a moment later
after first quickly raising the flash rate to
50 or 60 cps; this ensures that the subject
is not left to struggle with possible per
ceptual fluctuations due to stimuli of
near-threshold value.)
"All right. This time say ’Now’ when
it starts to move again, even a little
teeny bit. Remember it’s a tricky light.
Don’t let me catch you. Watch it carefully
and say ’Now’ when you are sure it starts
to move a teeny bit." (Examiner gradually
decreases flash rate until subject says
’Now,’ notes the reading, then rapidly
’moving* and ’still* were chosen because they were under
stood better by the kindergarten and first grade children
than alternatives such as blinking, flickering, wiggling;
or steady, standing, staying on, etc.
114
decreases rate again to 6 flashes per second,
and keeps it there while recording the
descending threshold value.)
Five ascending and five descending readings were
taken in this manner for all the first grade classes.
The reverse order (descending-ascending) was used
for the kindergarten classes.
Children were taken from the class in small groups
of three or four at a time. Five small chairs were
placed aide by side in the testing booth extending in a
line toward the VFT apparatus. The fifth chair was
directly in front of the eye-piece. Each child
advanced one chair at a time, at length reaching the
fifth chair and receiving his turn at the test.
Immediately after his test he was dismissed and re
joined the class. When two or three chairs were vacant,
the next group was allowed to come and take seats.
This procedure exposed each child (except the first
one for the session) to one or more repetitions of the
instructions, as he watched one or more children go
through the test. This appeared to have an expediting
and reassuring effect on the children.
In Tear II, when the testing booth was used, the
interval of waiting one’s turn also provided time for
dark-adaptation for all subjects (except the first one
who was given a special five-minute delay).
An administrative step which proved helpful was
115
to utilize a roster previously prepared by the teacher.
Thus even if shyness or poor speech were present, the
children*s names were quickly recognized, were already
written and spelled correctly, and required only an
instant for the examiner to verify with each child.
The roster also contained space after each name for
entering the child*s ten VFT scores.
Additional preparation and reassurance was given
to the class as a whole by a preliminary statement from
the teacher and from the investigator, whom the teacher
introduced. The statement mentioned that today was a
special day in which everyone would receive a turn
**to play a special eye-test-game with a tricky little
light that can blink on and off very slowly, or faster
and faster and faster until finally it goes so fast
you can’t see it move!** This orientation proved highly
effective in motivating the young children. It usually
brought expressions of glee from some of the class.
Group-Testing Procedures for the FMA and SPRD
In Year II, the VFT was preceded by group tests of
intelligence and achievement. All tests were given on the
same day, starting at the beginning of class in the morn
ing and extending through the full school day for the
first grade.^ The order and approximate time schedule
^In two instances parts of consecutive days were used.
116
was as follows:
1) PMA (Primary Mental Abilities)
6:30 AM to 10:15 or 10:30 AM with one or two
recesses
2) SPRD (Survey of Primary Reading Development)
10:15 or 10:30 AM to 11:30 or 11:45 AM with
one or two recesses
3) Lunch
12:00 M to 1:00 PM
4) VFT testing in small groups, the remainder
of the class carrying on normal routines.
1:00 PM to 2:30 PM
The class was prepared for the group testing by
the teacher, who introduced the investigator to the
class. The teacher then served as proctor while the
investigator served as administrator for the tests.
Thus, the group tests, as well as all VFT tests, had
the same administrator. This served to control and
minimize factors of variation which otherwise might
have arisen from differences between multiple
administrators.
In first grade, getting help from one’s neighbor
or giving it to him is very natural and spontaneous.
Therefore, in the group testing, in addition to fre
quent verbal directions and reminders for each child to
do the test-games just by himself, it was found that
physical barriers were not only helpful but necessary.
The need for these barriers was not confirmed until
the first of the classes had been tested. Thus, the
117
results for this class were invalid. They were discarded
and the VFT test which ordinarily followed the group
test in Year II was not given to this class.
For all subsequent group testing physical barriers
were provided in the form of temporary cubicles. The
cubicle walls were composed of sections of cardboard
cut from large packing cartons discarded by a market.
For each table, which regularly seated four children,
two cardboard sections were cut, slotted, fitted together
at right angles and placed on the table so as to form
four individual cubicles, 18 inches high.^- The cubicles
prevented any sustained or extensive copying. There
were still occasional instances in which a child
attempted to stand up and look over, or lean back and
look around the wall. But this could be immediately
noted by the teacher or the investigator and corrected.
To facilitate keeping the right place on the page,
place-markers (cards approximately the size of a post
card) were used by the pupils. Red crayons were used
^■The tables at which the children worked were 4
feet long by 2 feet wide. One cardboard section was 5
feet long, the other 3 feet long, and both were l£ feet
tall. They were slotted thus I n T I 0 l and the slots
were fitted together at right angles forming a walled
•cross* which sat upon and was taped to the table top
and divided it into four *cubicles.f Since the walls
exceeded the length and width of the tables by one-
half foot at each end and projected between the children,
it was difficult for one to lean back and look around
the wall into his neighbor*s cubicle.
118
to circle choices from multiple choice items. Black
crayons were used on two subtests which called for
additional black lines to complete a drawing partially
printed in the test booklet. Most children were able
to place their own first names on their test booklets,
followed by their last names or initials. Before or
during collection of the test booklets the teacher
checked the names and completed them when necessary.
Teachers1 Ratings of Reading Ability
In Year II, at the conclusion of the day's testing,
the teacher was asked by the investigator to rate each
child as to whether he fell into the top, middle, or
lowest part of the class with respect to reading ability.
General academic ability or intelligence was not
initially mentioned in this connection by the investiga
tor. About half of the teachers asked whether reading
ability should be rated separately from intelligence.
They were informed this was not necessary. In practice
it was found that the teachers gave the ratings quite
readily. Most of them utilized high, middle, and low
reading groupings for instructional purposes.
CHAPTER IV
RESULTS
The results sought in this investigation are des
igned to answer one or more of the five questions formula
ted in the first chapter.
These questions relate to the feasibility of admini-
*3tering the VFT test to children as young as five and one-
half or six years; the feasibility of carrying on such
testing at school; the adaptations of equipment and pro
cedures that may be required for the attempt; the extent
to which such early VFT scores, if obtainable, correlate
with reading success or failure at these early ages;
and finally the extent to which such scores might predict
reading retardation in later grades.
From the standpoint of obtaining data and econom
ically reporting the results, these five questions may be
reduced to three main issues. They are as follows.
First, is it physically possible to administer the
VFT test to such young children under any conditions at
all? The questions of whether it can be done at school,
or whether modified procedures and equipment are necessary,
are important but subsidiary aspects of this issue.
Second, if it is physically possible to test the
120
children, will the scores obtained be meaningful? That
is, how valid and reliable will the VFT scores turn out
to be?
Third, if such VFT testing proves both possible and
meaningful, will the scores correlate with success or
failure on reading tasks in kindergarten or first grade,
or with later ("predicted") reading retardation?
Evidence on these three issues is presented in this
chapter. The subdivisions follow the logical order of the
issues. The results of course bear on all five questions
raised for the study.
Additional results which developed in the course
of the study are also included where appropriate or
convenient.
The results are presented with a minimum of comment.
Fuller discussion of the findings is reserved for the
following chapter.
The raw data were obtained by the procedures des
cribed in the preceding chapter.
Median VFT scores for all individuals were calculated
by hand. These and all raw data were entered into
punched cards and most of the calculations were then per
formed by a high speed computer.1
^An IBM 7090 Computer housed at the Western Data
Processing Center, School of Business Administration,
University of California of Los Angeles, Los Angeles,
California•
121
Feasibility of the VFT
Teat with Toung Children
Using the equipment and the procedures described
in the preceding chapter, the VFT test was administered
to one-hundred and eighty-three (183) kindergarten
children in June 1958. Half (92) of these children
had not yet reached their sixth birthday; that is, one-
half the group fell into the interval five and one-half
to six years of age, and the other half into the interval
six to six and one-half years of age. No case had to be
discontinued by the administrator on account of inability
to control the subject or to secure and sustain sufficient
orientation or cooperation from the subject.
This result is in opposition to the findings of
Hartmann (1934)* Discussion of possible reasons for
this discrepancy is presented in the next chapter.
The result just stated does not mean that for every
child each one of his ten responses was perfectly clear
and credible. In a small number of cases, amounting to
approximately three per cent, the child gave apparently
valid and credible responses for the first few trials
and then hesitated, gave an extremely high or low response,
or even failed to give any response at all on the next
trial. This usually was not accompanied by removal of
his head from the eye-piece or by any other observable
122
extraneous activity. When briefly questioned in these
instances, the child seldom gave an adequate verbal
explanation, saying only nI don't know* or "I forget"
or merely remaining silent. In most of these cases
when the administrator said "Let's try it again" and
repeated the trial, the child responded adequately
on the second chance.
In a few of the cases to whom a second chance
was given, improvement was not secured. In these rare
cases the administrator discontinued the testing, re
assured the child that he had taken his turn well,
and released him. There were three cases of this kind
in the kindergarten testing, one of which was later
diagnosed as a mentally retarded case. In the first
grade there were also three such cases but only one
of these was discontinued by the administrator.
The percentage of cases which for any reason
were non-scorable amounted to less than two per cent
in kindergarten and less than one per cent in first
grade.
There were two cases in which trials were in
advertently omitted or "spoiled" by the experimenter.
These cases were not counted as "unscorable." The
missing scores were filled in by proration from the re
maining eight or nine valid trials for the individual.
123
Validity and Reliability of the
VfrT Scores In Young Children
How accurate and reliable are the VFT scores at
the ages and under the conditions reported? Central
tendencies, variabilities, and test-retest reliability
of the scores are presented In the following subsections.
Comparison of class groups
Table 4.1 presents the means, standard deviations
and standard errors for each of the class groups tested.
These measures are based upon a single score represent
ing each individual's VFT. This single score for the
individual was the median of his ten trials. The median
was used for this purpose because it is theoretically
a more accurate index of the individual's threshold than is
the mean. The mean is more subject to distortion by one
or more extreme or erroneous reponses either by the
subject or by the administrator. In many individual
cases the difference between the mean and the median
was two or three cps, a difference theoretically worth
taking into account. For groups of pupils, the in
stances in which the median was higher than the mean
were offset by those in which it was lower. Therefore,
the means for class groups were virtually the same
whether calculated from individual means or medians.
Table 4.1
Means, Standard Deviations and Standard Errors
of the VFTs for Kindergarten and First Grade Classes
Grade Class
Number Class
Number Tested Mean
First 1 20 18.88
a
2 15
19.46
a
3
20 19.00
t t
4 24 18.65
a
5*
20 22.05
a
6
20
19.93
a
7
18 19.46
a
8 20 19.46
a
9
18
(175)
13.45
K’gtn. 1 17
20.08
a
2 25
19.31
a
3 Sec. A 17
17.36
a
Sec. B 11 19.92
a
4* Sec. A 11 22.08
a
Sec. B 9 22.33
a
5*
22 22.42
a
6*
23 22.77
a
7*
19
22.70
a
8* 24 22.90
m
Standard
Deviation
1.76
1.07
1.63
1.4$
2.08
3.11
1.91
1.29
0.96
2.74
2.67
3.43
3.79
2.16
2.29
2.63
1.85
2.24
1.61
Standard
Error
.396
.276
.363
.301
.463
.694
.450
.288
.224
.399
.319
.345
1.142
.651
.763
.559
.204
.512
.328
Test Conditions
Booth
M
«
a
n
n
a
a
tt
Dark store room 9:30 - 12:00
Same; 1:30 - 3:00
“Bright* supply room; 6/10
Same; 6/ll
Same; 6/10 12:30 - 2:00
Same; 6/ll 12:00 - 1:00
Alcove in hallway
Hall between K*gtn. rooms
Same
Same
♦Conditions possibly affecting these classes are discussed in the following
chapter.
125
The order of the classes in Table 4.1 is the same
as the testing order. Notes on the physical surround
ings and the lighting conditions are also given for the
kindergarten classes, since for these groups the test
ing booth was not employed to control such conditions.
Inspection of Table 4.1 shows that eight of the nine
first grade classes and the first two kindergarten classes
have means that are highly consistent and standard
deviations and standard errors that are relatively small.
First grade class number 5 and kindergarten class number 3
have means that are numerically higher or lower than most
of the others. These differences are probably attributable
to an artifact caused by confusion of several similar
switch settings on the stroboscope, as discussed in the
following chapter along with other possible causes.
Despite these slight differences the general consistency
of the means is clearly visible.
In order to adjust for the differences in means
and standard deviations just mentioned, the VFTs for
each individual were converted into T-scale scores
(mean of 50 cps and standard deviation of 10 cps).
Each class distribution was used as the basis for
converting the scores for the Individuals within that
class. Thus the differences in numerical values in
Table 4.1 did not influence subsequent calculations,
since the T-scores were used thereafter instead of the
126
original raw scores.
The test-retest reliability of the VFT was calculated
by correlating the two sets of scores for 92 children who
took the test at the end of kindergarten and again at the
end of first grade. The correlation turned out to be quite
low. The coefficient was .174* and reached only the .09
level of confidence.
Comparison of age (grade) differences
From Table 4*1 it can also be seen that as a whole
the kindergarten scores are somewhat higher than those
in first grade. The difference, however, is possibly
attributable not only to age differences, but to differ
ences of dark adaptation or to some artifact affecting
certain classes as discussed in the next chapter. There
fore, the effect of age or grade differences in these
results is not clear.
Inter-trial differences
In order to examine the variability reported in
the literature due to crossing the threshold with an
ascending series (increasing cps) versus a descending
series (decreasing cps), the means and variabilities
were calculated separately for the five "up" trials
and five “down* trials. Lest the possible small differ
ences be concealed by possible small sex differences (and
vice versa) a further separation was made for boys and
girls. The results are presented in Table 4»2.
Table 4.2
Ranges, Means, Standard Deviations and Standard Errors
of VFT Scores on Five Ascending (Up) and Five
Descending (Down) Trials, for Boys and Girls
in First Grade and Kindergarten
(Figures in parentheses are based on individual
medians for 10 trials instead of individual means)
Grade No. Means of Trials Standard
Deviations
Standard
Error
Range
First
Grade
Boys
35
5 up
19.30
5 down
19.90
all 10
19.60
(19.55)
10 trials
4.36
(4.12)
10 trials
.513
(.268)
7.5 - 30.0
Girls 90
19.31 19.75 19.53
(19.42)
4.90
(4.12)
.512
(.432)
3.5 - 31.3
Totals 175
19.30 19.31 19.56
(19.30)
4.95
(4.10)
.332
(.310)
7.5 - 31.3
Kinder
garten
Boys
97 13.74 19.57
19.16
(19.13)
5.29
(4•34)
.533
(.492)
7.0 - 33-5
Girls ai 19.70 19.39
19.30
(19.34)
5.28
(5.15)
.537
(.572)
6.1 - 30.0
Totals 173
19.17
19.72
19.45
(19.43)
5.29
(4.93)
.407
(.373)
6.1 - 33-5
i —1
»o
128
From Table 4*2 it can be seen that there are slight
but consistent inter-trial differences, the means being
higher for the descending trials. The magnitude of this
difference was about .5 cps. This inter-trial difference
is not dependent upon sex differences for it is found
in the two subgroups for boys as well as in the two
subgroups for girls. Nor is it due to the order of the
trials, since in kindergarten the pupils began with a
descending trial while in first grade they began with
an ascending trial.
The ascending-descending differences are not large
enough to satisfy statistical tests of significance.
The largest such difference in the table, that between
the five ’ •up" and five "down1 * trials for kindergarten
boys, has a t-ratio of less than 1.04 which is far below
the usual criteria for statistical significance.
That the differences are not due to chance fluctua
tion, however, is evidenced by the fact, not shown in
Table 4*2 but evident in the raw data, that the average
for each one of the "up1 * trials differed from that of
its paired "down" trial by about the same amount, and
by the fact that there were no pairs of trials which
failed to show these differences. A sign test of course
would show this patterned difference to be statistically
significant since there were no reversals of sign.
129
Inspection of the raw data also revealed a
possible tendency for maximal scores to occur in the
middle or at the end of the 10 trial series, but this
tendency was indeed very slight and certainly does not
approach statistical significance in this study.
Sex differences
Sex differences are not evident or pronounced in
the figures of Table 4*2. There is little difference
between the means and standard deviations of the boys
and girls. In kindergarten the highest single score
belonged to a boy and the lowest to a girl, but the
opposite occurrence took place in the first grade.
Intra-individual (ipsative) variability
In order to determine the amount of variability
each individual showed on the scores from his ten trials,
the standard deviation and standard error of these scores
was computed for each case. Table 4*3 shows the range,
mean and standard deviation of these ipsative scores.
For the first grade group, Table 4*3 shows that the
individual standard deviations range from a low of .40 cps
(case #58) to a high of 4«71 cps (case #112), with a mean
of 1.293 cps. For classes 4 and 5, in which as discussed
in the next chapter, the switches on the stroboscope may
have been set in the wrong "low speed" combination, the
mean of the standard deviations was 2.125 cps. For all
Table 4*3
Ranges and Means for the Standard Deviations of
the Ten VFT Trials in Each Individual Case
(i.e. the Ipsative Standard Deviations)
Classes Number
of Cases
Standard Deviations on Ten Trials
First Grade
Range
(in cps)
Mean
(in cps)
All Classes 175
.40 to 4*71 1.293
Classes 4 and 5 40 .62 to 4.71 2.125
All Classes Less
4 and 5 135
.40 to 2.4# 1.046
Kindergarten
All Classes 173 .50 to S.12 1.353
Classes 1 and 2 42 .50 to
2.39 1.273
Class 4 20
.95
to 3.12 3.222
131
the remaining first grade classes except 4 and 5, the
mean was I.O46 cps. (The range for the individual
standard errors was from 0.15(3 cps to 1.86 cps.)
For the kindergarten group, the individual standard
deviations range from a low of .50 cps (case #K 165) to
a high of 8.12 cps (case #K 72), with a mean of 1.853 cps.
(The standard errors ranged from .196 cps to 3«2l6 cps.)
The chief reason for the difference in magnitude between
most first grade and most kindergarten classes is
attributed to different switch settings on the strobo
scope, as discussed in the next chapter.
VFT Scores as Predictors
of Other Variables
Correlations
In order to determine which other test variables
the VFT might predict and the accuracy of such pre
dictions, the inter-correlations were first calculated
for most of the variables contained in the battery.
Twenty-eight such variables were included in these com
putations. They consisted of the VFT scores taken in
kindergarten and first grade; the tests and subtests
employed in the first grade battery; and all third grade
reading test scores which could be located for these same
children.
132
Of course many of the third grade reading scores
were lacking because of students moving away, being
absent, missing the testing for various reasons, and so
forth. This attrition greatly reduced the number of those
for whom scores were available on all twenty-eight
variables. To partially offset these reduced numbers,
the data were submitted to the computer in three separate
runs, each of which covered many but not all of the
variables. The data from these runs are presented in
the three matrices given in the Appendix.
Inspection of the matrices reveals that most of the
correlations involving the VFT are negligibly low.
As a check on the accuracy of the results, and as a
means of making maximal use of the applicable data in a
number of partially complete cases, additional correla
tions were calculated, taking the VFT and only one other
variable at a time. Only those variables which seemed
substantially related to the VFT or which seemed to need
further checking were studied in this manner.
These results as well as those from the first three
runs with the larger number of variables are summarized in
Table 4.4. For each correlation coefficient the number of
cases (N) is shown on which it was based. The larger sam
ples of course include most of the cases in the smaller
samples.
133
Table
Correlations of Selected1 Variables with VFT Scores
Variable Number VFT VFT
Grade 1 Kgtn.
Calif. Achvt. Test
10/60 (early Grade 3)
Total Reading Score 103 .026
it it tt
k 8
.022
n it it
55 .085
Reading Vocab. Score
3*+
.058
.225
Lee-Clark Reading Readiness
10/£8 (early Grade 1)
Total Reading Readiness
95 .033
ti ti it
55 .020
it ii it
b2
.119 .090
ii ii it
3b
.155 .131
Vocab. -&-Foll • -Direc. b2
.069 .25*+
1 1 ti »
3b .166
, 1 +79* *
1 1 it 1 1
5b .318*
Cross-Out
9^ .258*
it 1 1
5b .221
1 1 ti
b2
.339* .3^+2*
• 1 ti
bO .306#
Primary Mental Abilities
6/59 (end Grade 1)
Mental Age in Months
17*+ .032
1 1 ti 1 1
b2 .0*+7 . 02*4
Survey of Primary Redg. Dev'mt
6/59 (end Grade 1)
Total Raw Score 17b
.059
1 1 it 1 1
b2
.091 .1 b7
Sentence Comprehension 17b ,6 b b
it 1 1
b2 .212 -.268
it 1 1
92 .d b b -.196#
Word Form Comparison b2 .020
.295#
1 1 it 1 1
17b .120
Form Comparison 17b
.1^1#
Kindergarten VFT
6/58 (end of Kgtn.) 92 • 1 7 b /
b2 .260
3b .210
**p
*p
#p
/p
approximately .01
approximately .05
approxi ma tely .06
approximately .09
134
Most of the VFT relationships shown in Table 4»4
are negligibly low. However, one of them is statis
tically significant at the .01 level of confidence.
Another is significant at about the .05 level. Three
more approach significance at about the .06 level.
The highest indicated relationship, an r of .479,
is that between the kindergarten VFT scores and the
reading readiness pictorial subtest entitled "Vocabulary
and Following Directions" (Lee-Clark). The next highest
relationship, an r of about .34, is between another
reading readiness subtest, "Gross-Out" (Lee-Clark) and
both sets of VFT scores. And finally, at about the .06
level, three possible relationships are indicated between
VFTs and certain subtests of the Harsh-Soeberg Survey of
Primary Reading Development: a) Form Comparison and
first grade VFT (r * .141); b) Word Form Comparison and
kindergarten VFT (r = .295); and c) Sentence Comprehension
and kindergarten VFT, a negative relationship (r = -.196).
Differences between means
a. Third grade reading scores
An effort was made to see whether the means for
certain portions of the distributions of the first grade
VFT scores are related to the third grade reading scores.
A scatter plot prepared for this purpose gave the appear
ance of a rather weak and somewhat curvilinear relation
ship. The extent of the relationship did not appear
great enough to warrant calculation of eta,
b. Teachers1 ratings
As mentioned in the procedures, first grade teachers
were asked to rank the students in their classes as
falling into high, medium or low groups of the class
on the basis of reading ability (sometimes including
general "mental ability" also). Comparisons of these
three groups with respect to their reading scores on
standardized tests, their mental ages, and their VFT
scores are given below in Tables 4»5 A, B, C and D.
I
Table 4.5 - A
Groups Rated by their First Grade Teachers as High, Medium
or Low in Reading Ability: Average Months of Mental Age
on the Intelligence Test
(Primary Mental Abilities)
Group
Rating
N Mean
Months
of M.A.
Standard
Deviation
Standard
Error
Difference
of Means
Standard
Error,
Difference
t-ratio
High 79 90.41
Medium 4$ $5*46
6.SO
6.42
.76 5
.927 4.95
(High-Med)
1.20 4.12
Low
47 $1.19
6.16
.$99 4*47
(Med-Low)
1.35 3.31
P
.001
.001
VjJ
O'
Table 4.5 - B
Group
Rating
High
Medium
Low
Groups Rated by Their First Grade Teachers as High,
Medium, or Low in Reading Ability: Average Scores
(in Grade Levels) on the First Grade Reading Test
(Survey of Primary Reading Development, Harsh-Soeberg)
I Mean
Grade
Score
Standard
Deviation
Standard
Error
Difference
of Means
Standard
Error
Difference
t-ratio
2.2d
>7 1.92
.311
.311
.035
,046 0.36
(High-Med)
.059
6.66 .001
k7 1.57 .477
,070
0.35
(Med-Low)
,065 4.12 .001
-4
Table 4.5 - C
Groups Rated by Their First Grade Teachers as
High, Medium, or Low in Reading Ability:
Average Scores on the Third Grade
Reading Test (California Achievement Test)
Group
Rating
N Percentage Mean
Reading
Grade
Standard
Deviation
Standard
Error
Difference
of
Means
Standard
Error
Difference
t-
ratio
High 56 71$* 4.54 .645 .113
Medium 26 56$*
3-14 .547 .103 1.40
(High-Med)
.153 9.1
Low
19 39$* 2.54
.603 .163
.60
(Med-Low) .210 2.66
.001
♦Percentage of pupils receiving this rating (high, medium, or low) at the end of
first grade who were still present and took the test in beginning third grade.
H
C O -
Table 4.5 - D
Groups Rated by Their First Grade Teachers as
High, Medium or Low in Reading Ability:
Average Scores on the First Grade VFT
Test (in Standard Scores of Cycles per Second)
Group
Rating
N Mean
VFT
Score
Standard
Deviation
Standard
Error
Difference
of Means
Standard t
Error
Difference
High 79 50.37 10.27
1.16
Medium
4 3 49.35 9.43
1.36
Low
4^ 43.94
10. IS
1.47 1.43 1.S7
(High-Low)
.76
140
Tables 4*5 A, B, and G show that high, medium and low
groups of pupils, as based on teachers ratings of reading
(and sometimes general) ability, yielded respective
averages that were very significantly different from each
other on all three standardized group tests (Primary
Mental Abilities, Survey of Primary Reading Development,
and the California Achievement Test).
Those whom the teachers rated high in reading ability
(including general ability also if the teacher raised the
question) averaged 2.23 grades of achievement on the
Survey of Primary Reading Development (SPRD) at the end
of first grade. At the beginning of third grade, they
averaged 4*54 grades of reading achievement on the
California Achievement Test (CAT).
Those whom the teachers rated medium were almost
exactly at the expected level.
Those whom the teachers rated low averaged about one-
half year below expectancy.
Significant differences in mental age scores on the
Primary Mental Abilities Test coincided with the teachers
ratings also. Those whom the teachers rated high in
reading or general ability earned mental ages that averag
ed about five months higher than for those rated medium.
And those rated medium exceeded by almost the same amount
those rated low.
141
The VFT scores also showed a discernible trend
coinciding with the teachers1 ratings, but the magni
tudes of the differences were far below the usual
criteria for statistical significance. Those rated high
had VFT scores which averaged about 1 cps above those
rated medium, and about 1.5 cps above those rated low.
Another result, reflected in Table 4*5C, is the
fact that those rated low in reading or general ability
tend to leave the district or become unavailable for
further testing much more frequently than do those
rated high. Of the high group, 71 per cent remained and
were retested in third grade according to the regular
schedule. But only 39 per cent of the low group re
mained for such testing. The middle group again was
represented by an intermediate figure of 5# per cent,
c. Variability ratings
An attempt was made to see whether variability of
an individuals ten VFT responses is correlated with the
other variables measured. The standard deviation was
computed for each individual*s ten VFT scores (ipsative
standard deviations). Each kindergarten or first grade
class was then divided into four groups. The fifty per
cent (approximate) showing the smallest deviations were
labelled the high-consistency half of the grade; and the
twenty-five per cent (approximate) with the smallest
deviations were labelled the high-consistency quarter.
142
Similarly the 50 per cent showing the largest deviations
were labelled the low-consistency half; and the 25 per
cent with the largest deviations were labelled the low-
consistency quarter. The mean performances of these
groups on other variables was then inspected to see
whether related differences were apparent.
In a number of the variables differences between the
means were noted that favored the high-consistency half
or quarter. Most of the differences between means, how
ever, were slight. Only one of them approached a satis
factory level of statistical significance. This was
the difference with respect to total scores on the
reading readiness test (Lee-Clark). The first grade
high-consistency half (N“45) had a mean total readiness
raw score of 54»44 as compared with 51»53 for the low-
consistency half (N“49)« The standard errors of these
means were 1.05 and 1.20 respectively. This yields a
t-ratio of 1.33 which approaches significance at the
.07 level of confidence.
This concludes the presentation of the observational
and statistical results of the study. Some of the results
are discussed in the following chapter.
CHAPTER V
DISCUSSION
In the preceding chapter evidence is presented on
the three key questions of the study, namely the feasi
bility of administering the VFT test to young children,
the possible validity of such VFT scores, and finally
their possible usefulness in predicting reading dis
ability. The present chapter discusses certain aspects
of this evidence, clarifies some points, and raises
some others for further question.
Administration of the VFT Test
From the ages and numbers of children tested in
this study at the end of their kindergarten year, it
seems clear that the VFT test can be administered succes-
fully to children as young as the year age level.
This was the youngest age tested in this investigation.
It may not be the youngest age testable under these or
other conditions.
Using the test-booth described, the VFT test can be
given in the child's own classroom at school, with com
paratively little effort in test administration or
advance preparation of the children. Most children were
1^3
144
willing and many were even eager to take the VFT test.
Not a single child refused to be tested. Only one
required brief coaxing.
Contrary to expectation, maintenance of order did not
constitute a major problem. In fact it did not even call
for any special effort. The children for the most part
sat and listened or watched or whispered quietly while
waiting their turns. No child was uncontrolled in the
VFT testing situation.
This is at odds with Hartmann’s (1934) earlier
report. Perhaps this difference in findings is due to
several factors such as the following.
First, it is certain that the apparatus used in the
present study provided less extraneous and distracting
stimulation than did Hartmann’s, since he used a sector-
disk which undoubtedly presented sounds and other vibra
tions as it rotated at various speeds.
Second, it may be that the viewing tube and eye-piece
used in the present apparatus served to orient the
children more fully and to focus and maintain their
attention upon the desired stimulus. As long as the
child’s head was in the correct position at the eye
piece, there was literally nothing else that he could
easily look at except the test-patch. And if for any
reason he did happen to remove his head from the viewing
tube for a moment it was almost instantly detected and
145
corrected by the administrator.
Third, Hartmann conducted his study in his labora
tory. The children were probably brought to these strange
surroundings singly or in very small groups. Hartmann's
(1934) entire account of this is interesting but sketchy:
"The confidence of the children was gained by
a timely distribution of candy and by encouraging
them to inspect some of the curios of a psycho
logical laboratory. The nature of the task was
carefully and patiently explained to the young
sters ." (page 123).
In contrast, the present study took place in the chil
dren's own schools and in the company of their regular
classmates. Thus, the situation was much more familiar
to them and therefore probably less threatening, dis
tracting, or over-stimulating. When the test booth was
used, the testing was carried on quietly and successfully
in the rear of the child's own regular classroom.
Hartmann does not give the size of his test patch.
His account sounds as though the entire sector-disc was
visible and the children were faced with the difficult
task of disregarding part of it.
"The subjects were asked to watch the flicker
at the edge of the disc and to look directly at that
area. Although few persons normally take advantage
of the higher sensitivity of the peripheral retina,
precautions were taken against this possibility by
adjusting the observers chair so that his eyes were
in a straight line with the mixer disc ...w
(page 123).
The present apparatus might also be an improvement
over that used by Tait (1956). Tait used a head-cloth
146
instead of the eye-piece and test-booth used in the
present study. Use of a head-cloth is open to the follow
ing objections: 1) variations in lighting due to shifts,
spaces or variations in the drape of the cloth; 2) dis
traction caused by the strangeness of being under the
cloth; by the rubbing of the cloth on the forehead or
cheek; and also by the subjectTs responsibility for
keeping it in place; 3) muffling of the subjects hearing
and speaking; 4) creation of a visual barrier so that
the administrator cannot detect when the subject is or
is not looking into the viewing tube; 5) and perhaps
most importantly, the possible accumulation of carbon
dioxide in the exhaled air underneath the head-cloth.
Change of oxygen or carbon dioxide content of
inspired air can very definitely affect the VFT.
Several additional points may be of importance
in the administration of the VFT test.
The first point is related to the main purposes
or tasks of the administrator. The main purposes of
the present VFT testing were not those of providing
a maximally accurate reading for each individual
case, such as might be desired in a clinical study or.
in an established visual-testing program. On the
contrary, the testing done in the present study possessed
only the quality of a rapid survey or preliminary
screening. In each of the year-end testing periods the
administrator faced the task of testing approximately
two-hundred children in a relatively short time. There
fore, the pupils had to be run through the procedure in
rapid assembly-line fashion. Even then, as previously
mentioned, three or four children from a classroom fre
quently had to be omitted because of lack of time.
Therefore, no pains were taken to detain a particular
child, to repeat all questionable trials, to refine or
to vary the testing approach by manipulating the frequency
control carefully up and down at about the threshold
level, or to make other finer adaptations. Such refine
ments would be very much in order in a real-life clinical
situation, or in further research with young children.
For the present study, however, the investigator tried
consciously to set and to maintain throughout the
experiment rather minimal standards and a routine,
mechanistic attitude of "let the chips fall where they
may," so that positive results, if any, would not depend
upon unusual administrative care and skill.
The second point is related to variations in the
administrator due to lack of extensive VFT experience
before the experiment, and to practice effects during
the experiment, such as greater facility in manipulating
the frequency control knob, reading off the figures
from the scales, writing them down, keeping an eye on the
148
child, and so forth. As the study progressed these and
other repetitive activities became more automatic.
Subjectively the investigator had "more chance to think"
about the testee, to anticipate his next response, to
move the control knob more carefully and perhaps more
slowly through the range where the response was
"expected," and to read the figures a little more
accurately at the moment when the subject said "now."
Nevertheless, a conscious effort was made to maintain
essentially the same rapid and "mechanistic" screening
approach from first to last.
The investigator had no prior experience with the
VFT except that which resulted from testing several
selected groups of older children (N = approximately 45)
in another school district, plus some young elementary
children (N = 6) in his neighborhood. Hence, changes in
administrative performance were bound to occur during the
course of several hundred repetitions of the test
procedure.
Validity of the VFT Results
Comparison of class groups
For an untrained administrator, working with very
young children, and attempting rapid screening rather
than more precise measurements, the agreement of most
149
of the class means seems quite close. The standard
deviations for the classes also compare well in size (2
or 3 cps) with those reported in the literature.
As shown in Table 4«1, the last six kindergarten
classes and first grade class number 5 have means of
about 22 cps. The remaining kindergarten and first
grade classes have means of about 19 cps. This differ
ence is probably attributable to operative error
involved in the settings of the switches on the
stroboscope.1
Another possible explanation of the difference
in means for the class groups is that they were due to
sampling fluctuations. Theoretically such sampling
fluctuations could have produced the differences
just mentioned; however, this does not seem likely,
considering the fairly small standard errors of the
class means.
■^The instrument used has four controls which present
many possible combinations, some of which overlap in the
frequency-ranges they produce. There is a SLOW and a LOW
setting, two HIGH settings, and two settings for incoming
power frequencies, LINE and DIRECT. Also on the luminous
dial, there are two scales, one of which must be selected
and read, depending on the settings of the two switches.
The scales differ by a factor of 4, and must be multiplied
by either 10 or 100, depending upon the switch setting, in
order to get the -correct readings in terms of rpm. These
then must be divided by 60 to convert them to cps. One of
the switches is a toggle switch which can be accidentally
changed to its opposite position by some inadvertent cortact.
No change of switch setting was knowingly permitted while
the testing of any given class was underway. However,
150
Still another possible explanation of differences
in class means involves differences in dark-adaptation.
Such differences were more likely to affect the kinder
garten classes, since they lacked use of the testing
booth. For example, the difference between sections A
and B of kindergarten class number 3 might theoretically
be due to differences of morning versus afternoon sun
light illuminating the supply room where the testing
was done.
The dark-adaptation explanation is weakened by the
fact that only two of the first three kindergarten
classes used a "dark" supply room, and the third class
used a "bright1 * supply room, yet all three classes have
means of roughly the same magnitude. The adaptation
explanation is further weakened by the fact that dark-
adaptation was controlled by the testing booth for all
first grade classes including number 5; hence a differ
ence in dark-adaptation could not have made this mean
more like the bulk of the kindergarten means.
It is observable that the most deviant first grade
mean (class number 5) resembles the majority of the
kindergarten means. Conversely, the deviant means of
there is no guarantee against occurrence of such an
artifact either by accident or by confusion. The possi
bility of such error could be avoided in the future
by taping the switches In place.
151
the first three kindergarten classes resemble most of
the first grade means. This strengthens the case for
assuming a repeatable and reversible artifact.
The upshot of these considerations seems to impli
cate a switching error as the probable source of any
appreciable differences apparent in the class means.
Comparison of age (grade) differences
Theoretically, of course, the difference between
most of the kindergarten and most of the first grade
means in Table H.l could also be attributed to age.
However, as just explained it is probably due to an
artifact relating to switch settings. If any differ
ences due to age exist between kindergarten and first
grade, they are small and are obscured in this study
by other factors.
From the standpoint of advancing our knowledge
of the possible changes in the VFT during childhood
it is unfortunate, and a limitation of this study, that
all factors except age were not kept constant from Year I
to Year II. This disadvantage seemed outweighed by
changing in Year II to the use of the test booth. The
advantage of the booth lay in the better control of dark-
adaptation, and in testing a methodological advance
whereby VFT can be measured in the regular classroom.
Inter-trial differences
As shown in Table W.2 the differences observed in
favor of descending versus ascending trials are slight.
They amount on the average to about .5 cps. Differences
of this magnitude are not large enough to be convincing
on t-ratio tests of statistical significance. However,
they are probably true rather than chance differences
as judged from the previously mentioned fact that they
are consistent throughout all five of the paired trials
in the study.
The fact that the difference in each pair of trials
is in favor of the descending or flicker trial is of
some interest because of the confused state of the
literature of this topic. There is agreement among all
authorities that descending and ascending thresholds are
usually different. However, there is no unanimity re
garding this difference or the reasons for it. Indeed
it is not yet established whether the descending or the
ascending threshold is "normally" higher. Simonson and
Brozek (1952)* and Landis (195*02 state that the
^...'’Uie fusion (ascending) threshold is higher.
The difference represents largely a disguised reaction
time which would depend to a large extent on the speed of
acceleration or deceleration and the initial rate of
light flashes." (Simonson & Brozek, 1952, page 2.)
p
..."Under most circumstances the frequency in cps
is higher for the ascending threshold than for the des
cending threshold...The fact that the...thresholds are
not the same is a phenomenon of interest which has never,
to my knowledge, been systematically studied. The differ
ence between the two thresholds acts as if an established
153
ascending (fusion) threshold is the higher. But Knox
(1945) concluded that the descending threshold is re
ported as slightly higher when the subject is ••set* 1 to
observe flicker. And Bujas (1957) concluded that ex
posure to a few seconds of subfusional frequencies, which
is necessary in ascending trials, has the effect of lower
ing the threshold.
The first two of the above articles agree that the
ascending threshold usually tests higher, but there is
doubt as to why this is so. The second two articles
give experimental findings tending to explain why the
opposite should be the case.
Perhaps the ascending trials were lower in the pre
sent study because the investigator always turned the
frequency rate to its lowest setting (about 6 cps) while
recording the reading from the preceding descending
trial. Thus before each ascending trial there was
routinely an appreciable exposure of from 3 to 5 or more
seconds to subfusional frequencies, a factor which
Bujas (1957) claims has its maximal effect within 20
seconds.
steady state of adaptation or an established inter
mittent state tends to persist, or as if an established
appearance has a certain inertia which resists change."
(Landis, 1954, page 268.)
154
Sex differences
As shown in Table 4*2 there is very little differ
ence between the means for boys and for girls on the VFT
in this study. The largest such difference of .96 cps
between the means of kindergarten boys and girls on the
five descending trials yields a t-ratio of less than
1.20 which is clearly not significant. It is possible,
of course, that small differences do exist but were
obscured by imprecision of the procedures used in this
study.
Intra-individual variability
The trends of differences between high-consistency
and low-consistency VFT groups with respect to their
means on other measures, including the reading readi
ness total test raw score in which P = .07, are interest
ing and suggest that variability of the VFT may be worthy
of further investigation. Small but important differ
ences in variability may have been concealed in this
study because of using a mechanistic rapid survey
approach which may have made too many cases look more
variable than they truly were.
VFT Scores as Predictors
of Other Variables
It is interesting to speculate about the possible
meanings involved in those few variables which were
155
found significantly correlated with the VFT scores in
kindergarten or first grade. The two highest correla
tions were between the VFT scores at the end of kinder
garten and two subtests of the Lee-Clark reading readi
ness test taken at the beginning of first grade. These
two subte3ts are designated "Vocabulary-and-Following
Directions," and "Cross-Out."
Vocabulary-and-Following Directions consists of
drawings of animals, people, and familiar objects such
as houses, autos, and boats. The drawings are grouped
into twenty numbered sections or "boxes" and each box
constitutes one subtest item. The earlier boxes contain
only two drawings, the later ones three or four. For
each box the administrator pronounces the name of one
object or indicates one activity out of several repre
sented in the drawings. The child is supposed to put a
mark on whichever drawing shows the object or activity
mentioned. No reading is involved. The obvious ability
requirements are visual perception, auditory perception,
comprehension, and the motor ability to mark with a
pencil.
Cross-Out consists of only 12 lines of print. Each
line contains only 4 capital letters. Three of them are
the same but the fourth is a different letter which must
be "crossed out." In the first three lines there is no
difference in the size of the capital letters. Eight of
156
the lines use two sizes of type, and the one remaining
line uses three type sizes. The style or typeface is
the same in spite of the size changes. Given enough
auditory perception, comprehension and motion to get
started on the test, the only other obvious ability
required is enough visual perception to perceive same
ness or differences of letter-forms which sometimes
differ in size. The letters do not have to be named or
sounded, but merely visually perceived as same or differ
ent in form regardless of size.
Time or speed is probably not an important factor
in either of the two above subtests. On the Vocabulary
subtest the test-adrainistrator controls the pace by
naming the objects or activities one at a time, and
pausing after each while the marks are made. On the
Cross-Out subtest, the children proceed at their own
rate, but no time limit is set, and most administrators
allow time for all to finish.^
If the question is posed as to what these two
highest correlations with VFT (respectively .479 and
•342) have in common, it seems self-evident that
^■It bears repeating here that the reading readiness
test in first grade and the reading achievement test in
thifd grade were routinely given by the district, hence
these two tests were not administered and scored by the
investigator, but by the teachers.
157
aside from the audition, comprehension, and motor abili
ties involved in taking tests and marking answers, the
key requirement in both of these subtests is visual per
ceptual efficiency.
Evidently most of the pupils found both of these
tasks quite easy. On Gross-Out the maximum score was 12,
the mean was 11.43, and the standard deviation was 1.76.
This is obviously a skewed distribution which indicates
that only a few atypical pupils had any difficulty on
this subtest. It seems quite safe to regard such pupils
as cases of low perceptual efficiency. And it is also
obvious that those same few cases tended to have low VFT
scores, thus producing the correlation.
A skewed distribution also was yielded by the
Vocabulary-Following Directions subtest, where the
possible score was 20, the mean was IS.4, and the standard
deviation was 1.16. Obviously this too was an easy task
for all except a few. Comprehension may be involved more
in this subtest than in the foregoing one. However, the
items were so familiar to most pupils that perhaps it was
perception of them in the drawings that posed some diffi
culty for those with low VFTs.
Looking next at the relationships which fall just
short of the customary five per cent criterion level, the
VFT appears to be correlated with Form Comparison (.141),
15S
with Word-Form Comparison (.295), and (negatively with
Sentence Comprehension (-.196).
Form Comparison consists of twelve pairs of drawings
of irregular shapes that represent no familiar objects and
hence are "meaningless" for most people. In each pair
the pupils must mark the shapes if they are alike, not
mark them if different. The task seems to be essentially
the same as in Cross-Out, i.e., simple visual perceptual
discrimination of shapes. Again most of the pupils found
the task quite easy. The possible score was 12, the
mean was 9*90 (2 points are deducted for each error),
and the standard deviation was 2.33. Few pupils made more
then one error out of the twelve items. And these few
also tended to have low VFT scores.
Word-Form Comparison is the same essentially as
Form Comparison except that the "shapes" to be marked
if they are the same are "wordsi* The pupils do not
necessarily have to recognize either the sounds of the
words or their meanings. The words must merely be dis
criminated as identical or non-identical visual gestalts.
The possible score was 14, the mean was 13*57, and the
standard deviation was 1.29. Again this shows that re
latively few children encountered any difficulty with
the task. And again these few tended also to have low VFT
scores and thus produce the correlation.
159
Sentence Comprehension, the subtest with a negative
correlation, consists of nine short sentences. Each must
be logically completed by marking one of three possible
words offered. Examples: Susan is a (house, tree, girl).
I like to (jump, red, home). We can (green, funny, run).
The three alternative words are printed several spaces
away from the rest of the sentence. Only one of the
three alternative words is on the same line as the rest
of the sentence; another is on the line above; the re
maining one is on the line below. Thus the correct word
to complete the sentence is frequently out of position
for normal reading habits. Each sentence is numbered and
separated from the other sentences by a heavy black line
running across the width of the page.
Contrary to the preceding subtests, this one appears
to require much more than easy visual perception and dis
crimination. The words must be recognized (read), the
meanings must be comprehended, and logical reasoning must
be used to reject false choices. Moreover some visual
perceptual habits of reading must even be opposed when
the correct completion word is not on the same line as
the rest of the sentence but is on the “wrong" line.
The highest possible score on this subtest is 9,
one point for each correct choice. The obtained mean was
4.19 and the standard deviation was 2.13. Contrary to
160
the foregoing instances, the average performance does not
approach the highest possible score. Evidently this task
or set of tasks definitely challenged the pupils.
Probably the "intellectual" nature of this subtest,
which requires and rewards "high-level" functions that go
beyond and sometimes even oppose simpler perceptual habits
in reading, explains its negative correlation with the
VFT. The greater the role of factors other than simple
perceptual efficiency, the greater is the chance to com
pensate for defects in perceptual efficiency. And when
these other factors even oppose routine perceptual habits,
the probability of a negative correlation is increased.
It is tempting to hypothesize that if other factors are
equal, those beginning readers who have set up the most
efficient perceptual habits are also the ones who would
encounter the greatest difficulty in breaking those
habits in order to select a "better" word on a "wrong"
line.
The possible value of the VFT in predicting future
cases of reading retardation would therefore seem to lie
in its capacity to identify those whose visual perceptual
efficiency is so low, and remains so low, that they have
continuing difficulty on "first level" functions, such as
recognizing shapes or seeing the differences between
shapes, letters, or words.
161
As to how long in advance the VFT can predict re
tarded (perceptually inefficient) readers, the evidence
from this study indicates only modest usefulness on a
relatively short-term basis. The longest interval studied
and showing some promise is the year between the kinder
garten and the first grade testing. The kindergarten VFT
yielded a correlation of .295 with Word Form Comparison
and a negative correlation of -.196 with Sentence Com
prehension. Both correlations approached significance
at about the .06 level (Table 4*4).
Correlations for the longer period between kinder
garten or first grade VFTs and the third grade reading
tests fell to magnitudes too small to appear of con
sequence.
Perhaps the explanation for these facts is two-fold.
First, the nature of reading obviously changes from
"readiness1 * in kindergarten or first grade to “compre
hension1 * in the third grade. Perceptual efficiency plays
a crucial role in the beginning stages, and a continuing
but decreasingly important role ever thereafter. Second,
the organism is changing quite rapidly at these ages and
perceptual efficiencies may change a great deal due to
maturation, as well as to illnesses and accidents. The
low correlation (.174) between the two VFT tests a year
apart may be due primarily to such changes.
Reading is a visual language skill found only in
humans and only after considerable auditory language skill
has been acquired slowly and overlearned. Of the total
activities which are needed to achieve normal reading,
the visual perceptual functions tapped by the VFT test
appear quite simple and mundane, yet quite necessary.
At least a critical amount of the visual perceptual func
tion would appear to be necessary as a "first level”
function before the higher level reading functions such
as comprehension and reasoning can be called into play.
Perhaps this explains why many retarded readers are
of average or better intelligence, comprehension or
reasoning, and can demonstrate these functions in auditory
spoken language, but not in visual language. It may be
that retarded readers do not possess sufficient "low-level
visual perceptual efficiency to master the rudimentary but
necessary "first-level” function of accurate form-
perception. Or if they possess the requisite perception,
they may not possess the requisite visual memories from
their perceptions to go on to "second-level” functions
such as form-recognition, form recall, form-decoding, and
form-encoding, all of which are necessary for reading,
spelling and writing. In any one modality, the first
and second level functions may have to become overlearned,
automatic, and unconscious before a pupil can carry on
163
"third-level" functions such as comprehension, reasoning,
judging, changing habits and so forth in terms of that
modality.
Summarizing this discussion of the VFT as a possible
predictor of reading retardation, perhaps the following
statement reflects the present situation. The only ex
perimental evidence thus far comes from the present study,
and it indicates that the VFT might predict reading re
tardation that is due primarily to inefficiency on the
relatively simple visual perceptual functions which form
an important part of the beginning stages of reading.
To the extent that the individual can compensate for his
visual perceptual inefficiency by memory, comprehension, or
reasoning the prediction from the VFT seems less likely to
be of aid.
Comparison with the Study by Tait
The present study did not parallel the methodology
of the study by Tait (1956). Tait did not work with an
entire population of a grade, including both sexes and
all levels of reading ability. Nor did he attempt to
correlate the VFT with reading ability. He selected from
three grades in two school systems enough disability cases
of one sex only (boys) to form an experimental group,
which he then compared to a matched group of normal cases.
164
This method could not be followed in the present study
because too few reading disability cases could be identi
fied. At least three reasons are readily apparent for
this.
First, the total number of pupils (about 200) was a
far smaller pool than that from which Tait drew his 51
disability cases. Second, as mentioned in connection with
Table 4*5, the poorer readers moved away at a much faster
rate than the able readers. This left very few retarded
readers in the third grade who had been in the district
since first grade and on whom a VFT score was therefore
available. Third, disability cases could not be identi
fied from the larger available number with scores on the
first grade reading test (Harsh-Soeberg SPED) because
none of them could as yet be classified as reading dis
abled according to Tait*s criterion of being two years
below reading expectancy.
It would seem desirable in the future to apply Tait»s
procedures to children as early as the end of second grade.
The second grade is the earliest possible one in which the
criterion of two years below reading expectancy could be
satisfied.
Long-interval Reliability of the VFT
Finally a comment seems called for regarding the
165
correlation between the two sets of VFT scores taken a
year apart. Judging from the literature survey and
from the statement by Landis (1954, page 282),^ this
may be one of the first reports of such re-testing. The
coefficient obtained, approximately .17, is quite low and
only approaches significance at the .09 level.
This immediately raises a number of questions. Is
this low figure truly representative of changes in child
ren^ VFT scores at these young ages? Is it also repre
sentative of changes in adults* scores taken a year apart?
If children and adults differ in this respect, at what age
is the difference greatest? Or is the coefficient spuri
ously low in this study because each set of VFT measure
ments was not taken repeatedly over a forty-eight or
seventy-two hour period, nor with clinical accuracy, nor
with the method of constant stimuli, nor with other
similar refinements?
Presumably the low re-test correlation must mean
either that the testing was not sufficiently precise for
long-term prediction; or that many organisms of these ages
manifest considerable change in the VFT over a one year
period; or that both of these explanations are true. But
since there are higher and more statistically significant
correlations between the VFT in kindergarten (June) and
^Quoted on page 74, Chapter II.
166
certain reading readiness subtests approximately four months
later (October), perhaps it can be expected that the
correlation of the VFT with itself, had it been done at
the same interval, would have been at least as high as
those correlations if not higher. Therefore, it might be
inferred that the lower correlation of the VFT with itself
at the longer interval is not entirely due to imprecise
testing. Thus, it is implied that some of the low reli
ability of the VFT at the one year interval is probably
due to changes in the organisms.
If the long-term reliability of the VFT is actually
no greater than .17, the prospect for long-term pre
dictions seems poor. It would mean that an individual’s
VFT score could not be considered a highly stable
characteristic of him. However, the literature on adults,
based on several sittings per day for a week or two,
suggests that it is stable and grows more so; that is, as
age increases, intra-individual differences decrease
(Misiak, 1951, page 552). And it also holds that an
individual’s pattern of diurnal variation in the VFT is
quite characteristic of him (Landis, 1954, page 572).
These facts have contributed to the widely held impression
that an individual’s VFT is a relatively stable personal
(and generic) characteristic, similar perhaps to body
temperature•
Perhaps a series of daily measurements taken over a
one or two week period in one year and repeated in the
same way in subsequent years might reveal a greater reli
ability. Such a procedure would not be difficult and
could minimize the effects of temporary fluctuations or
inaccuracies. It would be desirable in future studies
to ascertain when, in the human life span, the greatest
stability of the VFT is achieved, and how stable or
characteristic it is at best, barring illness or accident.
Discovery of the amount of VFT "drift" if any which is
normal or tolerable for certain ages or periods of time
would be valuable both for theoretical purposes and for
the early detection of pathological processes.
CHAPTER VI
SUMMARY, CONCLUSIONS AND RECOMMENDATIONS
Summary
The problem for attack by this study is that of
beginning to predict cases of reading retardation before
they have fully developed rather than afterward.
A survey of the literature on reading retardation
shows that several theories of causation exist. They
may be classified as primarily instructional, emotional,
or neuro-physiological. Many professional articles have
been written in the past half-century in support of
variations or combinations of one or more of these
theories. Most of the studies are discursive rather
than experimental. Even the experimental studies deal
almost exclusively with ex post facto cases. Few studies
attempt to predict future reading retardates in advance.
A previous study by Tait (1956) with selected and
matched groups demonstrated that a significant rela
tionship exists between reading retardation and low
sensitivity of the visual fusion threshold (VFT) in
children of fourth, fifth, and sixth grade ages.
16S
169
The present study explores the possible use of the
VFT as a measure -which might be used for a fresh attack
upon the problem at early ages. The study considers
such questions as whether the VFT can be applied to
children of pre-reading or beginning reading ages, and
if so, under what conditions, and with what results.
A survey of the VFT literature shows that no
standardized apparatus, procedure or norms yet exist
for measuring this threshold. The VFT is complexly
affected by many variables. The list includes fre
quency, duration, and intensity of the light flashes;
dark adaptation of the eye, amount of light admitted,
and its location (central or peripheral) on the retina;
body conditions of temperature, fatigue, health, toxi-
medication, carbon dioxide concentration, and
brain lesions; attention, attitude, emotional state,
and practice effect.
The effect of intelligence on VFT is undertermined,
some investigators claiming a relationship, others
denying it.
The effect of old age is to lower the sensitivity
of the VFT but the effect of young age is as yet unknown.
The ages of increasing or peak sensitivity are unknown.
Previous studies include very few children under age
eight, none under age six. An attempt by Hartmann (193*0
to test "a few bright nursery school children collapsed
170
entirely because of their high distractibility and the
excessive irregularity of their *judgments*1 *... (page 125) •
At the end of one school year (June 195&) an attempt
was made to administer the VFT test to the entire kinder
garten population of approximately two hundred children
in a suburban district of four elementary schools in
Los Angeles County. One year later an attempt was made
to re-administer the test to the entire first grade
population. Also included in the second testing were
three additional measures: a group test of intellectual
abilities (SRA Primary Mental Abilities), a group test of
reading ability (Harsh-Soeberg Survey of Primary Reading
Development), and teachers* ratings of reading ability
including general ability. From routine district testing,
results were also secured on a reading readiness test
(Lee-Clark) given at the beginning of first grade
(October 195#), and on a reading achievement test
(California Achievement Test) taken two years later at
the beginning of third grade (October I960). The two
latter measures were given and scored by the district
teachers; all others were given and scored by the
investigator.
The following modifications of VFT apparatus and
testing procedures were undertaken: 1) Construction and
use of an eye-piece and viewing tube assembly which
171
which quickly positions the child's head, limits his
vision and attention, yet does not interfere with his
breathing. 2) Construction and use of a low intensity
filter (two foot-candles) which eliminated irritating
effects of the intermittent light. 3) Construction and
use of a portable dark room or test-booth to control dark-
adaptation of the eye, and to permit testing to be carried
on inside the child's own classroom without disturbing
the class. 4) Use of a 'mechanistic' rapid-screening
approach with all pupils rather than a careful clinical
approach involving more time with each testee.
Of the statistical results obtained, the following
are the more important: 1) The means for classes ranged
from 17.86 to 22.90 cps, and the standard deviations from
O.96 to 3.79 cps. 2) The means for individuals, based on
ten trials given to each case, ranged from 6.1 to 33.5 cps,
and the standard deviations from 0.4o to 8.12 cps. These
means and standard deviations compare favorably with those
generally reported in studies with older subjects. Hie
class means of the individual standard deviations ranged
from 1.04 to 3.22 cps.
A small systematic difference of about 0.5 cps was
found in favor of trials descending rather than ascending
across the threshold.
Small differences due to sex, age, or intelligence
172
were not visible* They may have been obscured by tech
nical conditions related to dark adaptation (test booth),
variable switch settings, the rapid-screening (non-
clinical) approach, or the use of the method of variable
stimuli (j.n.d.) instead of the method of constant
stimuli•
The VFT did appear to be correlated with several
reading readiness subtests at or above the .05 level of
confidence. These subtests appeared to involve visual
perceptual efficiency in identifying drawings of familiar
objects (pictorial vocabulary), and in discriminating
identity or difference in pairs of ‘ •meaningless** forms,
letter forms, and (at about the .06 level) word forms
(not meanings)* In one subtest emphasizing meaning (com
prehension) rather than visual perceptual efficiency, a
correlation was obtained which was negative and approached
significance at about the .06 level.
In spite of the above correlations with kindergarten
or first grade reading readiness tasks, the VFT correla
tions with third grade tests of reading vocabulary or
comprehension appeared negligibly low in this study.
The reasons for this are not clear. One possible factor
is the low reliability of the VFT. Another is the change
in the nature of the reading processes from first to
third grade where perceptual processes are still necessary
but far less important than higher intellectual processes.
173
Relative absence of former disability cases by the third
year could also be a factor.
There was little or no observed relationship between
the VFT and factorial tests of visualization or of speeded
perception of visual forms, respectively the S and P
factors in the Primary Mental Abilities Test. This is
somewhat puzzling in view of the above correlations of the
VFT with readiness subtests involving unspeeded perception
of visual forms.
Consistency of an individual’s ten VFT responses, as
reflected in the size of their standard deviation, did not
appear significantly related to other variables. The more
consistent half or quartile did show slightly more desirable
means on some variables.
The reliability coefficient obtained between the two
VFT tests on 92 available cases tested one year apart was
quite low (r “ .174) and reached only about the .09 level
of confidence. Presumably this low reliability is not
entirely due to imprecision of testing. Some changes must
be occurring in the organisms. This may be the first such
year interval reliability reported in the VFT literature,
hence comparison with other results is lacking.
Other correlations of some interest between first and
third grade measures are as follows: the correlation co
efficient obtained between first grade reading readiness
(Lee-Clark) and third grade reading achievement (California
174
Achievement) was approximately .45 for total scores, and
coefficients ranged from about ,2& to .3# for various
subtests. For the factor-analyzed test (Primary Mental
Abilities) the P factor in first grade yielded a co
efficient of about .39 with third grade total reading.
The S factor yielded a coefficient of about ,2& with third
grade reading vocabulary, and about .41 with third grade
reading comprehension. The corresponding V factor co
efficients were .47 and .39 respectively, while those for
the total FMA score were .$7 and .55 respectively.
Of interest also was the accuracy and predictive value
of teachers* ratings. The group rated high in reading
(and general ability) surpassed the medium group by one-
third of a year on the standardized reading test at the
end of the first grade (Table 4*5-B). By the beginning
of the third grade, sixteen months later, they were almost
a full year and a half ahead of the medium group in
reading (Table 4*5-0). The difference in these means
is significant above the .001 level. This is valuable
assessment and prediction.
The difference between the low group and the medium
group was similarly significant. The low group was about
one-third of a year behind the medium group at the end of
first grade and about two-thirds of a year behind at the
beginning of the third grade. Pupils in the low group
moved away or missed subsequent testing to a far greater
175
extent than the others* If all of the low group could
have been followed and re-tested, the teachers* ratings
might have shown even greater predictive value. Ratings
deserve a place in future batteries attacking the problem.
Conclusions
The problem-attack formulated for this study called
for an attempt to answer the following five questions
listed in Chapter I.
1. Can the VFT test be administered successfully to
children prior to or concurrent with the ages
or grades for beginning reading, that is, as
early as kindergarten or first grade and ages
five and one-half or six?
Conclusion: Contrary to impressions from a few
earlier studies such as Hartmann’s (1934), it
is feasible to measure the VFT in children of
these ages, and to secure measurements that are
sufficiently accurate to be considered valid and
meaningful. Individual and class means and
standard deviations are consistent with each
other, and similar to results found in adults.
Test-retest reliability over a one year
interval under the conditions of this experiment
appears seriously low however. Reliability data
for intervals this long or longer apparently
have never been studied or reported in the
literature•
Can such VFT testing be administered at school?
Conclusion: Such testing can be carried on
successfully at school in various locations such
as offices, empty classrooms, storage rooms or,
if necessary, even in screened or protected hall
ways* However, it is accomplished more easily,
rapidly and accurately by utilizing a dismount-
able testing booth, set up in the rear of the
child's own regular classroom, as was done in
the second year of this study. The cost of com
ponents and materials for the booth and testing
apparatus described is approximately three hundred
dollars* The testing time needed by an ex
perienced administrator to test each individual
in the class is about two or three hours per
class.
Are any adaptions of previous procedures necessary
or desirable, and feasible?
Conclusion: Since flickering light at usual or
high intensities is irritating, in testing young
children it is desirable and feasible to use low
intensities (approximately two foot-candles)
provided by a test-patch assembly such as the one
constructed for this study. Dark-adapted vision
should then be used. The necessary dark-
adaptation can be secured by means of the test-
booth mentioned above. An eye-piece and viewing
tube assembly, constructed as described in this
study, are also desirable to quickly position
the child*s head, limit his motion, attention and
vision, and yet leave his breathing unhampered.
Are the VFTs of children in these grades related
to reading retardation at those times, that is,
to lack of reading readiness in kindergarten,
or to retardation of beginning reading in first
grade?
Conclusion: In this study the VFT scores are
correlated to a low but significant degree with
several reading readiness tasks. Namely, these
include correctly identifying drawings of fami
liar subjects (pictorial vocabulary), and cor
rectly discriminating between identical versus
similar "meaningless1 * shapes, identical versus
similar letters, and identical versus similar
words. These tasks appear to have in common
visual perceptual efficiency or the lack of it.
There is also evidence of a negative
correlation of VFT sensitivity with a more com
plex and intellectual first grade reading skill
which calls for comprehension of a sentence and
its completion by selecting one of three words.
It is concluded therefore that the VFT is
maximally related to those simpler visual per
ceptual processes which are the required but
rudimentary phases (neuro-physiological processes)
of reading, and minimally or inversely related
to the intellectual phases (psychological pro
cesses) of beginning reading.
Is such VFT testing related to, or will it pre
dict later reading retardation?
Conclusion: The conclusion with respect to this
question is guarded. Direct evidence in this
study is negative. The correlation of kinder
garten or first grade VFTs with third grade
reading appears negligible. Yet this may be due
to a number of occluding factors such as the
apparent low long-term reliability in the VFT,
differences in functions sampled by reading tests
at beginning and later levels, and so forth.
These occluding factors may be resolved or con
trolled in further research. It might then be
possible to predict in kindergarten and first
179
grade partly by repeated VFT testing and partly
by other measures, those individuals who will
experience reading retardation at later educa
tional levels.
The way seems open for much further VFT
testing both in and outside of school and with
children of all school grades, and perhaps
younger. Many refinements of technique and much
new basic and applied knowledge should result.
Recommendations
The following brief recommendations are offered for
future investigations of the VFT and reading retardation
in young school children.
Matched groups of disabled and normal readers should
be compared at the earliest possible grades and ages. It
now seems clearly possible to replicate Tait's (1956)
study of pupils in grades four through six, but to extend
it downward into lower grades which are more critical for
prediction, diagnosis, early treatment and prevention.
Replication could be made as early as the end of the
second grade and still meet Tait's criteria of reading
age being two years lower than mental age. At very
early reading levels the usual criterion for reading
retardation should probably be adjusted to make use of
smaller decrements. For example, a discrepancy of one-
160
half year between mental age and reading age might be
used as the criterion for reading retardation in kinder
garten and first grade.
The obtained negative correlation between the VFT
and reading comprehension should be investigated and
clarified.
Longitudinal as well as cross-sectional age studies
should be done to determine if possible the curve of
development which may take place in the VFT function in
common with many other human functions.
Sex differences, if any, should be investigated at
early and later ages.
Norms and standard deviations under standard test
ing conditions should be developed. Such knowledge could
be very valuable in detecting mild or masked pathology.
Norms could be achieved by systematic re-testing
of the same population of school children at semester
intervals from kindergarten through secondary school.
This has never been done. It is now seen to be feasible,
and not extremely difficult or time-consuming.
Development of group methods or at least small group
methods of VFT testing should be attempted for testing
children of all ages at school. This is primarily a
matter of designing and developing apparatus which can
expose the same or identical stimuli simultaneously to
1S1
a number of subjects, and which can record each one’s
responses automatically and silently so as not to
influence the responses of others. Such apparatus,
suitable for adults and housed in a trailer, has been
described by Doering, Ward and Hixson (1956). If it can
be adapted for use with young children and fitted into a
portable test booth it would be a great advantage.
An attempt should be made to adapt the “constant
stimuli" method of VFT testing, as described by Ricciuti
and Misiak (1954) for use in schools, and with children
of all ages including kindergarten and first grade.
Investigators interested in the field of VFT
measurement should attempt, perhaps through correspondenoe
or professional conferences, to develop a set of recom
mended specifications for satisfactory equipment and
procedures which if utilized would make their results
directly comparable. This is a long and necessary first
step toward the establishment of much-needed norms.
B I B L I O G R A P H Y
BIBLIOGRAPHY
Bartley, H. S. (195#) "Some factors influencing critical
flicker frequency," J. Psychol. 46: 107-115
Barry, Hortense (1955) "Classes for aphasics," in Frampton,
E. F. & Gall, Elena (editors) Special Education For
The Exceptional, Vol II. Boston: Porter Sargent
362-36?
Battersby, W. S. (1949) "Critical flicker frequency in
patients with cerebral lesions," J. Psychol.
23: 223-233
Bender, M. B. & Teuber, H. L. (1949) "Disorders in visual
perception following cerebral lesions," J. Psychol.
23: 223-233
Blanchard, Phyllis (1946) "Psychoanalytic contributions to
the problem of reading disability," Psychoanal.
Study Child. 2: 163-163
Brown, K. T. (1956) "A simple method of demonstrating
differences in the critical flicker frequency of
foveal and peripheral vision." Amer. J. Psychol.
69: 303
Brozek, J. & Keys, A. (1945) "Changes in flicker-fusion
frequency with age." J. Consult. Psychol.
9: 87-90
Bujas, Z. (1957) "Adaptacija vidnog analizatora na
treperenje svijetla i na isprekidane elektricne
fropene." (Adaptation of the organ of vision to
flickering and intermittent and electrical phos-
phenes.) Arh. Hig. Rada 3: 1-14
Burks, H. F. (1955) A study of the organic basis for be-
havorial deviations in school children. Los
Angeles: Univ. of Squth. Calif., unpub. dissert.
Colgan, C. M. (1954) "Critical flicker frequency, age and
intelligence," Amer. J. Psychol. 67: 711-713
133
134
Coppinger, N. ¥. (1955) "The relationship between critical
flicker frequency and chronologic age for varying
levels of stimulus brightness." J. Geront.
10: 43-52
Dearborn, W. F. (1931) Ocular and manual dominance in
dyslexia. Paper read at Amer. Psych. Assoc.,
Toronto, September 12, 1931.
Doehring, D. G., Ward, W. E., and Hixson, W. C. (1956)
"The development and standardization of a group
test for critical flicker frequency," U. S. Naval
Sch. of Aviat. Medic. Research Reports.
Dondero, A., Hofstaetter, P. R., & 0fconnor, J. P. (1953)
"Critical flicker frequency in light- and dark-
adaptation," J. Gener. Psychol. 53: 11-16
Eames, T. H. (1953) "The blood picture in reading failure,"
J. of Ed. Psychol. 44: 372-375
Enzer, N., and Simonson, E. (1940) "Experiments to deter
mine the effect of circulatory failure on some
simple functions of the central nervous system,"
Pis. Nerv. System. 1: 332
Foley, P. J. (1956) "Effect of background on critical
flicker frequency," Canad. J. Psychol. 10: 200-206
Freud, S. (1924) "Psychogenic visual disturbance according
to psychoanalytic concepts," in E. Jones (editor)
Collected Papers of Sigmund Freud. London: Hogarth.
Friedl, F. P. (1954) "Anxiety and cortical alpha in normal
subjects," Stud. Psychol. & Psychiat. 9: No. 2
Gillingham, Anna, & Stillman, Bessie W. (I960) Remedial
Training For Children With Specific Disability In
heading. Spelling And Penmanship. (Sixth edition)
, Mew York: Anna Gillingham hub., 25 Park Ave.,
*Bronxville 3.
Goldstone, F. (1955) "Flicker fusion measurements and
anxiety level," J. Exp. Psychol. 49: 200-204
Goodenough, Florence L. (1926) Measurement of Intelligence
By Drawings. New York: World Book.
Halpern, S. (1957) "The significance of critical flicker-
fusion thresholds in cerebral palsied children,"
Dissert. Abstr. 17: 1607
185
Halstead, W. C. (1947) The Brain And Intelligence. A Quan
titative Study of the Frontal Lobes. Chicago:
Univ. of Chicago Press.
Hartmann, G. W. (1934) "Comparison of the flicker thresh
olds in children and adults." Child. Developm.
5: 122-126
Head, H. (1926) Aphasia and kindred disorders of speech.
London: Cambridge Univ. fcress.
Hecht, S. (1937) "Rods, cones and the chemical basis of
vision," Physiol. Rev. 17: 239-290
Hinshelwood, J. (1917) Congenital Word Blindness. London:
H. L Lewis & Co.
Jarvis, V. (195#) "Clinical observations on the visual
problem in reading disability." Psvchoanal. Study
Child. 13: 451-470
Keller, J. E. (1958) "Interrelationships among several
measures of visual perception in mentally re
tarded children," Dissert. Abstr. 19: 571
Knox, G. W. (1945) "Investigation of flicker and fusion:
I. The effect of practice under the influence of
various attitudes on the critical flicker-fusion,"
J. Gener. Psychol. 33: 121-129
Krugman, M. (1947) "Flicker fusion frequency as a function
of anxiety reaction; an exploratory study,"
Psychosom. Med. 9: 269-272
Kushner, M. (1955) "The effects of experimentally induced
stress upon critical flicker frequency, palmar
skin resistance and reversible figure fluctua
tion," Dissert. Abstr. 15: 2583-84
Landis, C. (1949) "Psychology," in F. A. Mettler (editor)
Selective Partial Ablation Of The Frontal Cortex.
Uew York: Hoeber ^92-496
. (1951) "Something about flicker fusion." Sci.
Mon. 73: 308-314
. (1953) An Annotated Bibliography Of Flicker
Fusion Phenomena. Ann Arbor, Michigan: Armed
Forces-Uationai Research Council Vision Committee
Secretariat, 3433 Mason Hall, Univ. of Mich.
. (1954) "Determinants of the critical flicker-
fusion threshold," Physiol* Rev. 24: 259-286
________, & Hamwi, Violet (1954) "The effect of certain
physiological determinants on the flicker-fusion
threshold," J. Appl. Physiol. 6: 566-572
________. (1956) "Critical flicker frequency, age, and
intelligence," American J. Psychol. 69: 459-461
Mahneke, A. (1956) "Flicker fusion thresholds," Acta
Opthal. (Kbhvn.) 34: 113-120
Mark, H. J., Meier, P., & Pasamanick, B. (1958) "Varia
bility of critical flicker fusion thresholds in
brain-injured children," AMA Arch. Neurol.
Psychiat. 80: 682-688
Miller, V. L. (1942) "The critical frequency limen for
visual flicker in children between the ages of six
and eighteen," Genet. Psychol. Monogr. 26: 3-55
Misiak, H. (1947) "Age and sex differences in critical
flicker frequency." J. Exper. Psychol. 37: 318-
332
________. (1951) "The decrease of critical flicker
frequency with age," Science. 113: 551-552
Mpnroe, Marion (1932) Children Who Cannot Read. Chicago:
Univ. of Chicago Press.
Mucher, H., & Wendt, H. W. (1951) "Group test for the
determination of binocular critical flicker fre
quency: changes under caffeine and after a day*s
work," Arch. f. Exper. Path. u. Pharmakol. 214:
29-37
Orton, S. T. (1928) "Specific reading disability — stre-
phosymbolia," J. Amer. Med. Assoc. 90: 1095-1099
Peckham, R. H., & Hart, W. M. (1958) "Critical flicker fre
quency, photo-chemical mechanisms, and perceptual
responses," Arch. Opthal. 60: 461-471
Plateau, J. (1835) "Sur un principe de photometrie,"
Bull. Acad. Roy. Sci. Bell.-Let. Bruxelles
2: 52-59
Reuning, H. (1955) "A new flicker apparatus for measuring
individual differences," Acta Psychol. 11: 447—
466 ---
137
Rabinovitch, R. D., Drew, A. L., DeJong, R. N., Ingram,
Winifred, & Withey, Lois (1956) "A research
approach to reading retardation,** Neurol. &
Psychiat. In Childh. 34s 363-396
Ricciuti, H. N. & Misiak, H. (1954) "The application of
the constant method in determining critical flicker
frequency," J. Gener. Psychol. 51: 213-219
Richardson, J. (1950) "A factor analysis of reading ability
in ten-year old primary school children," Brit. J.
Of Educ. Psychol. 20: 200-201
Robinson, Helen .M. (1949) Why Pupils Fail In Reading.
Chicago: Univ. of Chicago Press.
Ronchi, L. & Bittini, M. (1957) "On the influence of the
shape of equal light pulses on the critical
flicker frequency," Atti. Fond. Ronchi. 12:
173-179
Schaefer, K. & Carey, C. R. (1954) "Influence of exposure
to various carbon dioxide concentrations on flick
er fusion frequency and alpha blocking," U. S.
Naval Submar. Medical Research Lab. Reports.
ttm s ------------ —
Simonson, E. & Brozek, J. (1953) "Flicker fusion frequency:
background and applications," U. S. Air Force Sch.
Of Aviat. Med. Pro.i. Reports. T S T S 5
, Enzer, N., & Blankstein, S. (1941) "The influence
of age on the fusion frequency of flicker,"
J. Exp. Psychol. 29: 252-255
________, Fox, M. S., Sc Enzer, N. (1943) "Influence of
vestibular stimulation on the fusion-frequency of
flicker in normal subjects and in patients with
post-concussion syndrome," Arch. Otolaryng.
33: 245-251
Smith, D, E. P. & Carrigan, Patricia (1959) The Nature Of
Reading Disability. New York: Harcourt, Brace.
Statten, T. (1953) "Behavior patterns, reading disability
and the EEG," Amer. J. of Psychiat. 110: 205-206
Stauffacher, J. C., Marks, J. B., Sc Ax, A. F. (1955) "The
effect of pinhole observation on the flicker
fusion frequency in patients with brain pathology,"
J. Consult. Psychol. 19: 150
ids
Strachey, J. (1930) "Some unconscious factors in reading,"
Int. J. Psych-Anal. 11; 322-331
Strauss, A. A., & Lehtinen, Laura E. (1947) Psychopathology
And Education Of The Brain-In.iured Child. Yol 1.
New 'York; &rune & Stratton
Tait, J. F. (1956) Mild neurological impairment as a factor
in reading disability: an experimental study. Los
Angeles: Univ. of South. Calif., unpub. dissert.
i-ias
Talbot, H. F. (1&34) "Experiments on light," Phil. Mag.
(Third series) 5: 321-334
Tanner, W. P., Jr. (1950) "A preliminary investigation of
the relationship between fusion of intermittent
light and intelligence," Science. 112: 201-203
Teuber, H., & Bender, M. B. (1948) "Changes in visual per
ception of flicker, apparent motion and real motion
after cerebral lesion," Amer. Psychol. 3: 246-247
Thurstone, L. L. (1944) A Factorial Study Of Perception.
Chicago: Univ. of 6hicago Press.
Vier, W. (1956) "Satiation, supersatiation and fatigue in
their effect on the optical fusion frequency,"
Zeit. Psychol. 159: 246-279
Wagenheim, Lillian E. (1949) The relationship of measles
contracted in infancy or early childhood to later
disability in reading. Los Angeles: Univ. of
South. Calif., unpub. dissert.
Walters, R. H., Van Loan, M., & Crofts, Irene (1961) "A
study of reading disability," J. of Consult.
Psychol. 25: 277-283
Wechsler, D. (1949) The Wechsler Intelligence Scale For
Children. New York: Psychological Corpor.
Werner, H., & Thuma, B. (1942) "Critical flicker frequency
in children with brain damage," Amer. J. Psychol.
55: 394-399
Young, K. M. (1949) "Critical flicker frequency," in F. A.
Mettler (editor), Selective Partial Ablation Of
The Frontal Cortex"! New 1fork: Hoeber 2i>7-263
A P P E N D I X
t 1 "T— 1
1 9 0
CORRELATION MATRICES:
All Variables Except First Grade Reading Test (Harsh-Soeberg SPRD) and Subtests of Intelligence Test (Primary Mental Abilities)
Run 1 1 ( t J = 31 *)
Intelligence
Test
( P r im a r y
Mental
A b ilities)
Kindergarten Reading Readiness Test
(Lee-Clark)
Third Grade Reading Test (California Achievement)
Reading Vocabulary
VFT Test
Mental (Matching Cross* Picture Multiple Total
Letters out Vocabu- Choice Score
In Different lary Matching
Months Letters Letters
and
Words
Word Opposite Vocabu- Vocabu* Follow* Reference Inter- Sub-
Recog* Words lary lary ing Skills pretation total
nition Sub Grade Direc- of Para* Compre-
Total Level tions graphs hension
Kinder* First
garten Grade
1 2 3
1 .
2. .289
3. .318 *95
1. .153 . 21*3 .021
5. .286 *3l *10
8. .390 .880 .811
7. .536
8. .586
9. .606
10. M
U. .506
12. .5II
13.
11. .619
15,-159
16 . * . o l l
10 11 12
13
ll 15 16
.169
.32? .902
*1(0
.398 .192 .178
•355
*50 *05 .278
•355
*89
•737
*76 *29
.273 .311 *71 .881
.987
*8?
.367 ,272
•335 *79 .811
•959 .978
*79 *69 ,231 .209 .381 .898
•777 .798 .700
.189 .210 .280 .181 .251
*53 •579 •570 .592 .188
.32I *58
.273
.321 *18
.727 .830 .818 .602 .718 .601
•379
*58 .299 .231 *38 .711 .880
.873 .821
.855 .757 •952
,029
.095
.199
.265
.168
.ott .131
.079 .155
-.155 *.192 *.191 -225 -.195 ■«* -001
..038 .120 .069 .058 .113 .238 -0®
-.053
.13?
.210
191
CORRELATION MATRICES:
All Variables Except Third Grade Reading Test
Run #2 (I = 1 *2)
First Grade Reading Test
(Harsh-Soeberg S.P.R.D.)
First Grade Intelligence Test
(SRA Primary Mental Abilities)
First Grade Reading
Readiness (Lee-Clark)
VFT
K
Test
1
1. 2.
3.
4.
5.
6. 7. 8.
9.
10. 11. 12.
13.
14.
15-
16.
17.
18,
19.
a
Form lord Word Sen Sen Story Total MA M m MA IMA PMA Match- Cross Pic Multi Total Kin- First
Com Com Rec tence tence Com- Score Ver Per Quanti Motor Spatial Men ing out ture ple Score der- Grade
pari pari ogni Recog Com- pre bal cep tative tal Letters Dif Voca Choice gar- VFT
son son tion nition pre- hen tual
Age ferent bulary Match ten
hen- sion Letters ing VFT
sion Letters
and
Words
1. -
2. -091
3. 120
■215
4. 030 -004 464
5. -004 -101
354 296
6. 232 023 286 269 320
7. 446 096 650 526 640 736
8.
525 179 293 073 106 266 472
9- 232 184
323
284
185 134 383
489
10. 502
137 477 313 154 211
549 577 578
11. 125 -l66 122 031 277 059 176 071 434
155
12.
569 033
180 -124
097
-009
273
440 470
530 195
13. 557
162 42$ 211 186 232 560 819
715
892 210 642
14.
383 124 204 250
343 259
491
357
268
369
118
197
384
15.
-023 082 304 449 160 101
279 331
742 278 233
060 388 328
16. 243
117
068 150 151 028 219 172 174 187 -051 115 222
334 161
17. 303 ■093
482 382 342 292 544
325
382
354 179 154 378 451 484 189
18.
CVJ
-016 489 431 355
304
577 407 532 430 143 193
488 604 65I 396 920
19. 025 295 -177 -038 -268 -Utl -11(7 135 073 Oil -085 -025 021 06! 342 254 -036 091
20, 115 020 -097 -020 212 018 Q91 107 133 023 -026 -067 047 104 339 069 # 119 260
1 9 2
CORRELATION MATRICES:
All Variables Except Flicker Fusion Scores
Run # 3 (S = 66)
First Grade Reading Test
(larsh-Soeberg S.P.R.D.)
First Grade Intelligence Test
(SRA Primary Mental Abilities)
First Grade Reading
Readiness (Lee-Clark)
Third Grade Reading Test
(California Achievement Test)
1 2 3 * 5 6
1 8 9 10 11 12 13 1* 15 16 11 18 19 20 21 22 23 2* 25 26
1.
2. 3*0
3- 158
*. 251
5. 198
6, 151
7. 529
l i t . * * 2
15. 085
16. 300
11. 186
18. 212
311 ltl9
116 355 3*8
226 3l t 3 1** 3lt5
511 151 51* 603 650
1.
2.
3.
I t .
5.
6.
1.
8. ill 1A9 1+08 228 212 35* *29
9. 286 119 219 2*0 181 259 *0* *20
10. 355 153 *26 318 205 319 *90 *50 *9*
U. 069 260 086 115 205 202 251 016 305 06*
12. 5*6 l** 250 181 209 081 319 29* 393 *16 230
13. 385 202 *86 380 29* 390 588 166 666 862 12* 596
38* 229 3*1 316 231 *52 195 22l 1*2 08* 216
383 316 *13 158 225 *01 120 398 211 300 161
325 06* 225 115 1*0 262 161 123 015 -051 198
3*3 333 295 301 335 *11 29* 228 11* 1** 17*
50* 385 *23 312 361 515 30* 313 229 155 26l
239
258 269
180 280
211 265
350 *69
19. 121 295 556 *9* 33* 288 56* 322 291 261 233 185
20. Ill 2*3 519 *91 *65 *01 623 *19 *03 *03 18* 30*
21. 166 269 560 553 *56 385 6*3 **1 *09 *01 181 216
22. 19* 2** 566 528 *82 **1 666 *10 390 *18 1*1 219
23. 11* 181 511 5** *09 268 5*5 363 319 311 113 3*1
2*. 260 230 35* 182 *0* 290 *5* 36* 236 391 083 **8
25. 116 155 526 **6 *93 392 59* 329 310 310 198 3*5
26. 220 20* 5*1 *61 510 380 622 391 381 *21 188 *1*
38* 211
5** 301
550 290
513 285
506 211
502 128
*19 150
555 202
Form Comparison
Word Comparison
Word Recognition
Sentence Recognition
Sentence Comprehension
Story Comprehension
Total Score
8. H4A Verbal
9. IMA Perceptual
10, IMA Quantitative
11. IMA Motor
12. IMA Spatial
13. IMA Mental Age
1*. Matching Letters
15. Cross-Out Different Letters
16. Picture Vocabulary
11. Multiple Choice Matching Letters and Words
18. Total Score
19. Word Recognition
20. Opposite Words
21. (Sub-total) Vocabulary
22. Vocabulary Grade Level
23. Following Directions
915 2*. Reference Skills
25. Interpretation of Paragraphs
26. (Sub-total) Comprehension
366 091 313 391
291 329 382 *6y 68*
299 308 363 *50 812 969
293 218 383 *56 801 9** 913
352 265 299 390 6*1 199 821 158
258 308 288 3*1 *56 515 56* 583 *93
313 360 3*6 *22 690 825 839 812 1*0 618
351 366 363 **9 101 862 815 8** 851 151 951
016
530 301
696 *59
1 2 3 * 5 6 1 9 10 11 12 13 1* 15 16 11 18 19 20 21 22 23 2* 25
GLOSSARY
GLOSSARY OF VFT TERMS, SYMBOLS, AND LAWS
brightness
CFF
compensation
cps
F
FFF
fc
Ferry-Porter
Law
subjective estimate of the dimness or brilli
ance of a perceived light source.
symbols or abbreviations usually denoting the
VFT, though with certain misleading connota
tions. Landis (1954) prefers this notation
and defines it clearly as the "critical
threshold of flicker-fusion" (page 260).
But the important term threshold is not
suggested by the letters and it is often in
accurately translated as ’critical flicker
frequency* or "flicker-fusion frequency"
(See FFF). It is misleading to emphasize
frequency instead of threshold, since the
same threshold capacity of the same individ
ual can be reflected by many different fre
quencies depending upon other factors such
as luminance, visual angle, LDR, etc. (See
VFT).
(for Talbot’s Law). When a light beam is
interrupted by a sector disk there is an
apparent reduction in brightness of the light
even at fusional frequencies. Compensation
is increasing the luminance in order to off
set this apparent loss of brightness in an
interrupted or intermittent light, as com
pared with the brightness of a light of the
same intensity that is not interrupted.
cycles per second
frequency of intermittent stimulation.
Usually given in cps (but sometimes in fps
or rpm).
flicker-fusion frequency. Same as CFF, hence
also commonly but inaccurately used as a
synonym for the VFT (See VFT).
foot candles. A measure of luminance. See
ml (millilamberts). One fc = 1.076 ml.
FVFT “ a • Log * + where Fy^ = the fre
quency for the VFT, a & b = constants for the
individual, and log I *= logarithm of the
luminance. The law states that for all in
dividuals the frequency F at which the VFT
194
195
flicker
F
max
fps
fusion
Granit-Harper
Law
I
LDR
log I
luminance
will be experienced rises or falls in pro
portion to the logarithm of the luminance.
(This is an application of the more general
Veber-Fechner Law that linear (arithmetic)
differences in sensory judgments are pro
portional to logarithmic (geometric) differ
ences in the stimuli.
perceptions of unsteadiness or pulsations in
an intermittentTight source, as contrasted
with at least two other possible perceptions:
fusion; and impressions of rhythmic light-
and-dark periods (slow enough to be recogniz
ed as such).
the maximum frequency at which flicker can
still be perceived when all other variables
affecting the VFT are manipulated to elevate
F to its highest.
flashes per second.
perception of steadiness in an intermittent
light source (as contrasted with flicker, or
with impressions of rhythmic light-and-dark).
Fvft “ c 1°S A + d. The "law" expressing the
fact that the F of the VFT rises or falls in
linear (arithmetic) units in relation to
logarithmic (geometric) increases or decreas
es in the size of the area (A) of the retina
stimulated. Constants for each individual
are represented by c and d. (Note the
similarity to the Ferry-Porter Law).
intensity of the illumination at its source.
Usage now favors synonym: luminance.
light-dark ratio. Ratio of the duration of
the light flash to the duration of the dark
period, e.g. 10:1, 1:1, 1:10. Similar (but
not identical) to PCF.
logarithm of the luminance units.
luminous flux from the light source per unit
of solid angle; usually given in terms of
millilamberts or foot-candles. Same as older
term, intensity (I).
196
ml
PCF
rod-cone
break
rpm
sector-disk
stationary-
state
equation
Talbot’s Law
millilambert. A unit of luminance. One ml =
.929 fc.
pulse-cycle fraction. The duration of the
light period as compared to the duration of
the entire cycle, e. g. .1, .5, .9; similar
(but not identical) to LDR.
the range of frequencies (from about 16 - 22
cps depending on the individual) in which
there is little increase in F in spite of
large increases in luminance; this is pre
sumably due to a change-over from predominant
usage of rod-receptors to cone-receptors as
higher levels of luminance are reached.
revolutions per minute. Sometimes used in
stead of cps because stroboscope dials are
calibrated in rpm.
a disk of cardboard or other material with
alternate sectors which are either painted
in different colors, or removed entirely;
when rotated rapidly in the path of a
light beam, the light is either passed by
the open sectors and interrupted by the
solid sectors, or reflected unequally by
the painted sectors.
K 1/2 = xn/(a - x)m. Formulation by Hecht
(1937) of the photochemical processes in the
retina which take account of many VFT deter
minants, but not those affecting the optic
nerve and brain. K, m and n are constants;
a = the initial concentration of sensitive
material; and x - the concentration of
photoproducts. I = luminance.
(Also known as the Talbot-Plateau Law.)
The reduction in apparent brightness of an
object viewed through a sector-disk (see
compensation) is proportional to the size of
the open sector if the disk’s speed is high
enough to eliminate flicker. The luminance
must be changed logarithmically (geometri
cally) to produce a corresponding linear
(arithmetic) change in the perception of
brightness.
197
surround
test-patch
va
VFT
the area surrounding and bordering the test-
patch.
a source of intermittent light, projected or
reflected, used as the stimulus for testing
the VFT.
visual angle; the angle subtended within the
eye (at the nodal point) by light from any
visual object* The size of the va deter
mines the size of the retinal area stimulat
ed.
all anatomic, physiologic, metabolic, psycho
logic or other internal conditions of the
moment belonging to an individual which
determine in an intermittent light stimulus
the various combinations of features (such
as frequency, luminance, visual angle, color,
LDR, and so forth) under which he perceives
a subjective transition from flicker to
fusion or vice versa.
ADDENDUM
ADDENDUM
Discussions following completion of this report have
convinced the investigator of the need for reconciling
terms expressing the sensitivity of the visual fusion
threshold with similar terms for most other thresholds.
When speaking of most other thresholds, it is common
practice to say that the threshold is lowered when sensi
tivity is increased, and is raised when sensitivity is
decreased. With respect to the visual fusion threshold,
however, the custom is just the reverse. The visual
fusion threshold is said to be lowered when sensitivity
is decreased, and raised when sensitivity is increased.
This reversal causes confusion and difficulty in dis
cussing the visual fusion threshold and the factors which
may affect it.
An example may make the situation clearer. Brain-
injury, anxiety, and lack of oxygen, are conditions which
reduce sensitivity to flicker and are said to lower the
visual fusion threshold. On the contrary, aspirin and
other analgesics which reduce sensitivity to pain are said
to raise the pain threshold. Thus, when referring to the
visual fusion threshold, that which lowers sensitivity
is said to lower the threshold; whereas when referring to
other thresholds, that which lowers sensitivity is said
to raise the threshold.
199
200
The author suspects that the reason for this situa
tion is that the visual fusion threshold has been and still
is most often measured by the frequency of the intermittent
light. This usage is reflected in such popular symbols as
CFF and FFF. Researchers using these symbols tend to use
"frequency" and "threshold" interchangeably and synonymously.
But this is confusing as shown above, and unjustified.
The lack of justification is easily seen if one recalls
that either dark duration or intensity of the light
flashes may also be used to measure the visual fusion
threshold. For example, frequency may be held constant
at 30 cps while the duration of the dark period in each
cycle is lengthened or shortened, or the luminance is made
stronger or weaker. The smaller the dark duration that can
be noted as flicker, or the lesser the luminance level at
which flicker can be seen, the more sensitive is the or
ganism. In these measurements lesser (lower) numbers
mean greater sensitivity, which is the opposite of the
case with frequency.
In an already complicated situation it would ordinar
ily be well to refrain from suggesting use of another set
of symbols. However, in this instance, since symbols
stressing frequency are misleading and likely to result in
the reversals mentioned above, it is recommended that the
symbols VFT be adopted to denote the visual fusional
201
threshold.
As stated elsewhere in this report, this would carry
the added advantage that fusion in other modalities could
be represented by logically parallel symbols. For in
stance the auditory ("flutter") fusion threshold could be
represented by AFT, the tactile by TFT, and so on.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

An Evaluation Of Orientation Programs In Seven Selected Small Liberal Arts Colleges, Using Student Satisfaction Criteria

PDF

Development And Analysis Of Some Tactual Measures Of Intelligence For Adolescent And Adult Blind

PDF

Some Effects Of Lessons In Human Relations In The Sixth Grade

PDF

Attitudinal, Perceptual And Cognitional Rigidity In Boys With Reading Disabilities

PDF

The Appropriateness Of Field And Level Of Vocational Choice As Related Toself-Concepts, Intelligence, School Achievement, And Socioeconomic Status

PDF

Reversal And Nonreversal Shift Performance Of Retardates Under Various Motivational And Stimulus Conditions

PDF

Self Concept And Social Class Correlates Of Contrasting Behavioral Subcultures Among Ninth Grade Boys

PDF

Color Vision Deficiency In Third And Sixth Grade Boys In Association To Academic Achievement And Descriptive Behavioral Patterns

PDF

An Empirical Study Comparing Various Methods Of Achievement Expectancy Bymeans Of Mental Ability

PDF

An Assessment Of Effects Of A Large, County Wide, Public School Program Of Special Training Classes On Trainable Retarded Children

PDF

Attitudes Of Student Ethnic Groups On Issues Related To Education

PDF

Reward Expectancy Strength As Related To The Magnitude Of Frustration In Children

PDF

An Instrument For The Measurement Of Cognation: The Cognition Of Social Interaction

PDF

The Relationship Of Creative Thinking Abilities To School Achievement

PDF

The Effects Of Pretraining In Auditory And Visual Discrimination On Texting In First Grade Boys

PDF

An Analysis Of Data From School Records Desired By Employers In Selectionof Beginning Workers

PDF

An Investigation Of Two Language, Two Memory, And Two Perceptual Abilities In Retardates Of Mental Age Four

PDF

Social System And Psychological Models In The Labeling Of Children As Educable Mentally Retarded

PDF

Utilization Of The Earliest Childhood Recollection In Detecting Maladjustment Among Junior College Students

PDF

Frustration Versus Secondary Reinforcement In Children Under Continuous And Partial Primary Reinforcement

Asset Metadata

Creator Howe, John Wesley (author)

Core Title The Visual Fusion Threshold (Vft) Test As A Measure Of Perceptual Efficiency In Kindergarten And First Grade, And As A Possible Predictor Of Later Reading Retardation

Degree Doctor of Philosophy

Degree Program Educational Psychology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag education, educational psychology,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Meyers, Charles Edward (committee chair), Carnes, Earl F. (committee member), Longstreth, Langdon E. (committee member), Thorpe, Louis P. (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-285911

Unique identifier UC11359005

Identifier 6305054.pdf (filename),usctheses-c18-285911 (legacy record id)

Legacy Identifier 6305054.pdf

Dmrecord 285911

Document Type Dissertation

Rights Howe, John Wesley

Type texts

Source 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 au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA