Iowa City, Iowa
recently Dr. Nancy Bayley has presented a
review article in Children (5: 123,
July-August, 1958) entitled Value all(l Limita-tiolls of Infant Testing where she quotes
studies made by her and also by others
wilich would not support the statement
made by Knobloch. Bayley states, “In
gell-eral, tests are of great value in judging and
diagnosing a child’s current status, but they are of very little use in predicting what the
child’s IQ will be a few years later.
How-ever, persons who use the developmental tests have been slow to accept the fact that
tests given in infancy do not predict later
intelligence.”
As one who has used developmental tests
in many clinical situations, I would agree
with the statements made by Bayley. It
would seem to be that pediatricians and
psychologists both should become more
sophisticated concerning developmental
testing in infancy. There is now consider-able evidence to point out the sharp
limita-tions in the predictive sense of these tests.
This evidence should become a part of the
information that pediatricians have
avail-able. Readers wishing to acquaint
them-selves with this information are referred to
tile excellent bibliography in Bayley’s
an-tide.
However much one might like to accept
the conclusions of Knobloch et al., referred to in their article, it would seem unwise in the light of what others think about the use
of developmental testing. It would certainly
be true that many people have observed
clinically the vast differences between the
amount of apparent cortical damage as
re-vealed by the pneumoencephalogram and
the level of intellectual functioning.
Un-fortunately, in my opinion, we still do not
have available instruments which
differ-entiate the discrepancies that exist between structure and function in the central nerv-oils system.
ROBERT B. KUGEL, M.D.
Department of Pediatrics State University of Iowa
To THE EDITOR:
Thank you for giving me this opportunity
to reply to Doctor Kugel’s letter in criticism
of our paper, “Pneumoencephalograms and
Clinical Behavior.” I, too, feel that it is high
time pediatricians should know the nature
of “developmental testing” and what its po-tentialities and lmlitations are. Although it is not possible to present all of the evidence in a letter, I welcome this opportunity to
correct some false impressions.
Before proceeding to some comments on
the validity of infant evaluations, it is
nec-essary to point out that such a discussion
is not directly pertinent to the article under
consideration. Since the developmental
ex-amination was often done several years
after the pneumoencephalogram, the mean
age of the children at the time of the
clini-cal examination was 22 months, no longer
infants, and there was no age difference
be-tween the normal and abnormal groups.
Therefore almost half of the children were
over 2 years of age and a good number over
3 years. Psychologists, of course, feel fairly
secure in the evaluation of a 2-year-old
child, particularly one who can speak.
Con-relations between examinations at 2 years
and 10 years are about .45 and they rise to
about .55 at 3 years.1
The “opinion” about the lack of validity
of infant tests is not recent. We have
re-viewed in several papers the poor
correla-tions reported in the literature, and Nancy
Bayley’s first report on “Mental Growth
during the First Three Years” was published
in 1933. However, the evaluations used in
the past have not been the Gesell
Develop-mental Examination described in
Develop-mental Diagnosis on which our evaluations
are based. Often they have not even been
standardized : usually lumping together
motor, adaptive, language and
personal-social functions; not used with a neurologic background; and frequently given to highly selected groups.
The statements we have made about the
predictive value of developmental tests are
sup-176 LETTERS TO THE EDITOR ported by data. Doctor Kugel must certainly
be aware of this, since the first paper on the
subject was published from the Johns
Hop-kins School of Hygiene and Public Health
when he was a staff member of the
Ma-tennal and Child Health Division, and a
reference to it has been given in the paper under discussion.
To my knowledge there are four major
groups of infants on whom the Gesell
De-velopmental Examination was used and
con-relations made with re-examinations. The
results are considerably higher than those
previously reported with the use of other
methods, which range from .2 down to
small negative correlations.2 The first is a
group of 50 New Haven Negro infants who
were re-examined at 7 years of age by
differ-ent examiners unaware of the initial
find-4 Correlation in this group was .5
be-tween the first and last examinations. The
second is a group of approximately 100
in-fants examined at several ages between 16
weeks and 18 months of age.5 Again the
correlations ranged between .5 and .75, but results of the previous examination were
re-called in most instances. The third, and
probably largest, group comprises 300 out
of 1,000 infants examined in the Study of
Prematures at Johns Hopkins, previously
re-ferred to.’ In this study the examiner had no
knowledge at the initial examination done
at 40 weeks of age, whether the infant was
a premature on a full-term control in the great majority of instances, and no knowl-edge at 3 years of age of the results of the 40-week examination. In this study the cor-relation for the total group of infants was .5; if neurologic or intellectual impairment
was present, the correlation was .75. The
fourth group of infants was examined in
New Haven prior to the completion of
adop-tion placements by my colleagues, the late
Dr. Catherine Amatruda and Dr. Benjamin
Pasamanick; possibly a few were done by myself. These infants were re-examined
be-tween 5 and 9 years of age and the results form the basis for the Wittenborn report on
adoption,7 which has been so widely
dis-seminated, and favorably reviewed by the
Chairman of the Committee on Adoptions
of the American Academy of Pediatrics.
I would particularly like to discuss tilis
report because of my assumption that, at
least in part, Doctor Kugel’s statements are
based on it, since he participated in the
study. A group of infants placed in
adop-tion are probably the poorest group to pick
to test validity of an infant evaluation by
correlations between examinations in
in-fancy and later. These infants come from
all kinds of early care backgrounds,
includ-ing institutions and good as well as poor
foster homes; one would expect variation
from early examination after placement in
homes of a different quality. Wittenborn has
given no information that he excluded
insti-tutionalized children, who had low scores,
from this group. He certainly had access to
the records of these children, many of
whom we know had low scores during their
stay in an institution and who after
place-ment in foster homes rose to the full average range.3 Nevertheless, in spite of Witten-born’s statistical manipulations, which are
very impressive but of questionable validity
in the minds of some statisticians, in at least
one group he finds correlations as high as
ours. In his initial analysis he indicates that
the findings on those infants who were
ex-amined at the Clinic of Child Development before placement, and for whom placement
was recommended because tiley were found
to be free from neurologic impairment,
showed a correlation of .55 with the results of their re-examinations at 5 years of age.
Nowhere, however, does Wittenborn
mdi-date the real purpose of a developmental
examination and in no place does he
mdi-cate that there were any poor placements
made on the basis of failure to diagnose
neurologic or intellectual deficit.
Let us turn now to what is to be expected
from an infant evaluation and what it is
not able to do. Its most important function is to make a diagnosis of the normality or
ab-normality of neurologic status. This tends to
in the lay press for ilis unwarranted disturb-ing of the parents of normal children.
Sec-ondly, it silOuld detect those infants with a
developmental potential that is below
nor-mal, usually on an organic basis. The
ex-amination is not designed to detect
supe-riority nor to predict later I.Q. scores
pre-cisely. Although a small percentage of
in-fants might be considered superior, whether
they remain so depends upon their later
ex-periences. The I.Q. determined at school
age is a reflection of learning and
expeni-ence; a high I.Q. is a result of opportunity
and stimulation, and one would not expect
a high correlation later in a group of infants
from an above average background. The
in-dications from the Study of Prematures are
that at 3 years there is already a sharp di-vergence on a racial and socio-economic
basis. In both white and non-white groups,
those infants from the better backgrounds
had an increase in their quotients while
those from homes with restricted
oppor-tunities declined.5 Environmental factors
were evidently already beginning to modify
the behavior and could well explain the fact
that the correlation is not closer to 1.0 than
it is. This shift with increase in age does not
indicate a defect in the infant evaluation but rather the fact that the later examinations measure experience rather than intellectual potential. The infant examination is done at
a period of very rapid development; the
first 2 years of life are probably equivalent
to 7 to 10 years at a later period. The high
correlations that are obtained, therefore, are even more significant and are of the same order as those secured in school-age tests.
The studies that we have done indicate
that when care is taken to eliminate bias
and the infant examination is used as a
din-ical neurologic tool by a physician
ade-quately trained in its use, good correlations
are obtained. These studies have not been
challenged by the critics of infant
evalua-tions; they have merely been ignored. Special training, however, is necessary and there are certain basic qualifications to
be met. As the evaluation is primarily a
neurologic examination, the examiner must
be a physician, preferably a pediatrician
(
because he has knowledge of diseaseprod-esses in children) who is equipped to take
an adequate history with special empilasis
on early behavior and development. He, or
she, must have received adequate training
in the techniques of the examination. We
believe a minimum of 1 year is necessary
before the physician obtains sufficient back-ground for the satisfactory interpretation of
the Gesell Developmental Examination; our
Fellows- and Physicians-in-Tnaining get this
minimum. These trainees can reproduce tile
results we have reported, and there is a
high intercorrelation between independent
judgments. Finally, the physician must have the ability to interpret clinical findings, be-cause the evaluation is not a matter of add-ing pluses and minuses on a score sheet.
The most important point about the use of
the infant neurologic developmental
evalua-tion is that the examination does best what we expect of a clinical tool; that is, it identi-fies the abnormal infant and defines his
ab-normality. As clinicians we would be
satis-fled with this statement: “This infant has
no neurologic impairment and his potential
is within the healthy range; depending on
what his life experiences are between now
and 6 years of age, and excluding later
brain damage, he will at that time have a
Stanford Binet I.Q. above 90.” As a matter of fact, the Gesell Developmental
Examina-tion does considerably better than this.
Contrary to other statements in the
litera-tune, that 60% of infants vary more than 20
points in developmental quotient from one
examination to another of the 300 infants in
the Premature Study, 50% varied less than
10 points and 75% less than 15 points
be-tween 40 weeks and 3 years of age. We are
not disappointed that the evaluation does
not necessarily identify in infancy the
su-perior school child who has the benefit of
the enriched cultural academic experiences
that the standard intelligence test is
de-signed to evaluate.
178 LETTERS TO THE EDITOR
in several groups of infants, and when the
experimental design is such that care is
taken to eliminate bias, the burden of proof is on those who say “It can’t be done.” They
must review not only the nature of the
de-velopmental tools which they are using, but
also the adequacy of their training in
ad-ministering and interpreting them. HILDA KNOBLOCH, M.D. Clinic of Child Development Ohio State University
Columbus, Ohio
References
1. Bayley, N., and Jones, H. E. : In Carmichael,
L. : Manual of Child Psychology. New
York, Wiley & Sons, 1946, p. 588.
2. Bayley, N. : Mental growth during the first
three years. A developmental study of 61 children by repeated tests. Genet. Psychol. Monogn., 14:1, 1933.
3. Pasamanick, B. : A comparative study of
the behavioral development of Negro in-fants.
J.
Cenet. Psvchol., 69:3, 1946.4. Nash, E. H., Nash, H., Pasamanick, B., and
Knobloch, H. : Further observations n
the development of Negro children : status at seven years. In preparation.
5. Knobioch, H., and Pasamanick, B. : The
distribution of intellectual potential ill Il infant population, in The Epidemiology
of Mental Disorder: A Symposium in
Celebration of the Centennial of Emil
Kraepelin. Washington, D.C., American Association for the Advancement of Sci-ence. To be published; available in mime-ographed form.
6. Knobloch, H., Rider, R. V., Harper, P. A.,
and Pasamanick, B.: The neuropsychiatnic sequelae of prematurity: a longitudinal study. J.A.M.A., 161:581, 1956. 7. Wittenborn,
J.
R., et a!.: A study of adoptivechildren. II. The predictive validity of the
Yale Developmental Examination of
in-fant behavior. Psvchol. Monogn., 70:59,
1956.
AKTUELLE PROBLEME DER
KINDERTUBERKU-LOSE, Von Prof. Dr. H. Wissler. Stuttgart, Georg Thieme Verlag, 1958, 76 pp., $3.05. This 76-page, paper-bound book by the
lead-ing Swiss student of pulmonary disease in
chil-dren discusses in nine short chapters the new
problems and new horizons opened up by the
effective anti-tubenculous drugs now available. Sparingly illustrated, it does not claim to be in
any sense a textbook, but simply a series of
reports and discussions of recent developments in the bacteriology of tuberculosis (culture media; isoniazid resistance) ; the changing epi-demiologic pattern (including supenmnfection
of infected children) ; the place of adrenal cortical hormones in treatment; the manage-ment of cervical adenitis; the electrolyte dis-tunbances seen in patients with tuberculous meningitis; the changing clinical picture;
prog-nosis and management of bone and joint
le-sions, and attitudes towards BCG vaccination. Each chapter has a very up-to-date
bibliog-raphy. The book (one is tempted to say
book-let) is so short and to-the-point that it is
heart-ily recommended for anyone who reads
Gen-man and is interested in some of the fascinat-ing new problems presented by tuberculosis today.