New Slosson Intelligence Test norms : implication for ...
Transcript of New Slosson Intelligence Test norms : implication for ...
University of Massachusetts Amherst University of Massachusetts Amherst
ScholarWorks@UMass Amherst ScholarWorks@UMass Amherst
Doctoral Dissertations 1896 - February 2014
1-1-1983
New Slosson Intelligence Test norms : implication for special New Slosson Intelligence Test norms : implication for special
education. education.
James C. Bradley University of Massachusetts Amherst
Follow this and additional works at: https://scholarworks.umass.edu/dissertations_1
Recommended Citation Recommended Citation Bradley, James C., "New Slosson Intelligence Test norms : implication for special education." (1983). Doctoral Dissertations 1896 - February 2014. 3863. https://scholarworks.umass.edu/dissertations_1/3863
This Open Access Dissertation is brought to you for free and open access by ScholarWorks@UMass Amherst. It has been accepted for inclusion in Doctoral Dissertations 1896 - February 2014 by an authorized administrator of ScholarWorks@UMass Amherst. For more information, please contact [email protected].
NEW SLOSSON INTELLIGENCE TEST NORMS:
IMPLICATION FOR SPECIAL EDUCATION
A Dissertation Presented
by
JAMES C. BRADLEY
Submitted to the Graduate School of the
University of Massachusetts in partial fulfillment
of the requirements for the degree of
DOCTOR OF EDUCATION
September 1983
School of Education
James Charles Bradley
All Rights Reserved
ii
NEW SLOSSON INTELLIGENCE TEST NORMS:
IMPLICATION FOR SPECIAL EDUCATION
A Dissertation Presented
by
JAMES C. BRADLEY
Approved as to style and content by:
Dr. Arthur W. Eve, Chairperson of Conimitte
Dr. Madeleine Wheeler, M^ber
ACKNOWLEDGEMENTS
The author wishes to acknowledge a few of the many people who
contributed to the successful completion of his doctoral program.
First, I would like to thank Dr. Arthur W. Eve who provided me
with the necessary support and faith throughout my graduate program.
Second, I would like to thank Drs. Armstrong and Jenson who provided
me with the technical and statistical assistance required to analyze
the data. In addition to their technical expertise I am indebted to
them for their availa±)ility and willingness to share their knowledge.
Third, I would like to thank Dr. Eve, Dr. Wheeler, Dr. Daniels, and
Dr. Maloy for their valuable suggestions which contributed to the
completion of this dissertation. Finally, I would like to thank the
personnel of the various school systems who administered, scored, and
returned the tests used in this study. Without them it would not have
been possible to complete this dissertation.
IV
ABSTRACT
New Slosson Intelligence Test Norms;Inplication for Special Education
September 1983
James C. Bradley, B.S., Salem State College
M.Ed., Salem State College, Ed.D. , University of Massachusetts
Directed by: Professor Arthur W. Eve
The purpose of this study was to determine if the new Slosson
Intelligence Test (SIT, 1981) scores are comparable to Wechsler Intelli-
gence Test for Children-Revised (WISC-R, 1974) scores (Verbal and Full
Scale)
.
The study sample included 336 students ranging in ages 6-15 from
twenty communities in Massachusetts. Sxibstudies were carried out by
examinee sex and age level, by examiner e^^ertise, by order of adminis-
tration of the instr\iments and by time interval between the two test
administrations. The following is a summary of the findings.
1. The correlations across the entire sample and for each of the
subgroups were positive and significemtly different from zero.
The overall correlations between the revised (1981) SIT and
1974 WISC-R (Verbal and Full Scale) were .89 and .83.
2. The correlations across the entire sample between IQ's
obtained from the revised SIT (1981) and the WISC-R (Verbal
and Full Scale) were significantly larger than the correla-
tions relating the old SIT IQ's (1961) and the WISC-R IQ's
(Verbal cind Full Scale) .
V
In sum, the new SIT (1981) scores are comparable to WISC-R
(1974) scores (Verbal and Full Scale)
.
VI
TABLE OF CONTENTS
ABSTRACT ^
ChapterI. INTRODUCTION TO THE PROBLEM 1
Statement of the Problem 4
Purpose 7
Limitations 9Definition of Terms 9References 17
II. RELATED RESEARCH 18
SIT and the SB 19
SIT and the SB-R 29
SIT-R and the SB-R 33
SIT and the WISC 35
SIT-R and the WISC-R 37
References 43
III. PROCEDURES 46
Restatement of the Purpose 46
Design 47
Hypotheses to be Tested 50
Sample 51
Data Collection 52
Statistical Analyses 53
Summary 55
References 56
IV. RESULTS OF THE STUDY 57
References 68
V. SUMMARY AND RECOMMENDATIONS FOR FUTURE RESEARCH .... 69
SummaryFuture Research
BIBLIOGRAPHY
APPENDIX A
APPENDIX B
vii
LIST OF TABLES
1. 1961 Slosson Concurrent Validity Study 202. 1961 Slosson Concurrent Validity Study,
Independent Experimental Study 213. Armstrong and Jensen 1970 Study 244. Armstrong, Mooney and Jensen 1971 Study 255. Chimera Review of 18 Research Studies
Concerning the SIT and the SB 27
6. Armstrong and Jensen 1981 Study, DifferencesBetween Unrevised Stcinford-Binet andUnrevised Slosson IQ's by Levels ofChronological Age and Intelligence 30
7. Armstrong and Jensen 1981 Study, DifferencesBetween Revised Stanford-Binet amd Unre-vised Slosson IQ's by Levels of Chrono-logical Age and Intelligence 31
8. Armstrong and Jensen 1981 Study, DifferencesBetween Revised Stanford-Binet amd RevisedSlosson IQ's by Levels of ChronologicalAge and Intelligence 34
9. Armstrong, Jensen and Reynolds 1974 Study, Means,
Standard Deviations, and Correlations Among
the Variables {N=198) . Corellations Above
the Diagonal are Corrected for Restriction
of Range 1961 Slosson and 1941 WISC 36
10. Armstrong, Jensen and Reynolds 1974 Study, Mean
Absolute IQ Differences Slosson and WISC 37
11. Chimera Review of the 17 Research Studies Con-
cerning the SIT and WISC
12. Means, Standard Deviations, Correlations and
Mean Absolute IQ Differences (MAD) of the
New SIT and WISC-R Tests 59
13. Correlations of the New SIT with the WISC-R
Tests14. Improvement of the Correlations of the SIT
IQ's with WISC-R IQ's
15. Mean Absolute Differences (MAD) of the New
SIT with the WISC-R Tests
16. Improvement Upon the Mean Absolute IQ Differ-
ences (MAD) Between the SIT and WISC-R 56
viii
CHAPTER I
OVERVIEW AND STATEMENT OF THE PROBLEM
Introduction to the Problem
Results from a recently administered test of intelligence are fre-
quently used as one of the criteria for initial assessment of children
with special needs, as well as for periodic reassessment to determine
the appropriateness of the program for each child. Chapter 766 re-
quires a comprehensive individual psychological examination for stu-
dents who are recommended for a full team evaluation.^ The law also
states that each child placed in a special needs program shall be re-
evaluated no less than every three years , or sooner if it is recom-
mended by any member of the evaluation team. Re-evaluation includes
the same areas as the initial screening assessment.
Chapter 766 does not require the inclusion of intelligence test
scores for all psychological assessments, nor does it require the use
of individual intelligence tests when they are a part of the psycholog-
ical evaluation component.
In September, 1980, the writer conducted an informal survey con-
cerning these two issues. Twenty Directors of Special Education were
asked to respond to three questions: (1) To what extent are intelli-
gence test scores incorporated into the psychological assessment com-
ponent; (2) To what extent are individual tests used, rather than group
tests; and (3) Are group tests or individual tests preferred. The in-
clusion of intelligence test scores in psychological assessments ranged
1
2
from 75% to 95%, with 15 of the 20 directors estimating they are in-
cluded more than 85% of the time. Also, the directors stated that
individual tests are used at least 70% (70% to 100%) of the time, and
that they would prefer to use individual tests for all cases. Appendix
A contains a list of the 20 coitmninities involved in this survey. In
sum, the psychological assessment component usually includes the re-
sults of an individually administered intelligence test such as the
2Wechsler Intelligence Test for Children - Revised (WISC-R) or the
Stanford-Binet Intelligence Scale (SB) , Form L-M.
^
Unfortunately, individual intelligence tests such as the WISC-R
and SB require specialized training in their administration, scoring
and interpretation—usually amounting to a full semester course for
each test. They are also time consuming in that they require an aver-
age administration time of one hour. Thus, there is a need for a valid
individual test of mental ability which can be quickly administered
and easily scored. Such a test could be used as a screening and/or
re-testing instrument.
In 1961, Richard Slosson published the Slosson Intelligence Test
(SIT) The author's purpose was to develop a "mini" Stanford-Binet
that is—an individual intelligence test which: (1) did not require
specialized training of its administrators; (2) took only 15 to 20
minutes to administer and score; and (3) would yield IQ scores compar-
able to the SB. The SB (1960) was used by Slosson as the criterion for
establishing the validity of the SIT.^ The norming sample included 701
4-18+ (adult) . Pearson-Product Moment correlations werecases , ages
3
confuted separately for ages 4—18+ (adult) , resulting in concurrent
validity correlations ranging from .90 to .98. Slosson concluded that
the SIT correlated with the SB as well as the SB correlated with
itself.
During the years 1964-1970, numerous concurrent validity studies
were conducted, foremost of which were the Armstrong and Jensen studies.
In their 1970 study (1) , involving 490 students, they reported:
(1) an overall correlation between the SB and the SIT of .93; and
(2) correlations ranging from .89 to .96 for various sub-categories
such as age, sex, order of administration, time between testing and
examiner training. A particularly interesting finding was a correla-
tion coefficient of .94 when the SB was administered by a certified SB
examiner or WISC examiner, and the SIT by a non-certified examiner.
The researchers also found an overall mean absolute IQ difference (MAD)
of 5.64, which is approximately the same as the standard error of
measurement of five IQ score points for the SB.
Armstrong and Jensen concluded that: (1) the SIT could be used as
a valid screening and retesting substitute for the SB; and (2) non-
certified SB and WISC examiners could be easily trained (in approxi-
mately 30 minutes) to be competent SIT examiners. They also noted that
the average administration time for the SB was 64 minutes, compared
with 17 minutes for the SIT.
In a 1974 study, Armstrong, Jensen and Reynolds reported similar
findings for WISC and SIT scores.^ The correlations between the SIT
and WISC Verbal (W-V) , and the SIT and WISC Full Scale (W-FS) scores
were .93 and .90 respectively. Although the correlation between the
4
SIT and WISC Performance (W-P) was considerably lower (.72) , the
correlation between the W-V and W-P was only .75. Mean absolute IQ
differences (MAD) were: 5.03 for W-V and SIT; 5.42 for W-FS and SIT;
4.24 for W-V and W-FS; 8.79 for SIT and W-P; and 8.51 for W-V and
W-P.
Statement of the Problem
Just when it appeared there was sufficient evidence that SIT IQ
scores could be used with as much confidence as SB, W-V and W-FS IQ
scores, both the WISC (1949) and SB (1960) were revised. The SB was
revised in 1972 and the WISC (now WISC-R) in 1974.^
The 1972 SB revision did not include the addition or deletion of
test items. Item revision was limited to two items. At the two year
old level, children are shown the picture of a child and asked to
identify (point or touch) various parts of the body, e.g., hair, eyes,
mouth, etc. On this item, the clothing worn by the child in the pic-
ture was updated. At the seven year old level, examinees are asked to
state the similarity between the words "wood" and "coal." In this
instance, the word "charcoal" can be substituted for coal. Thus, the
1972 SB revision was restricted to score renorming, or changing the 10
values for obtained scores. On the other hand, the 1974 WISC revision
(now WISC-R) included major item changes; (1) addition of new items;
(2) deletion of items; and (3) modifications in wording and/or the
interpretation of some items which were kept.
Studies conducted by Armstrong and Jensen, Sattler, and Thorndike
show that the differences between the 1960 SB scores and the new 1972
5
. 8SB scores are considerable. In almost all cases, the SB scores have
been lowered, which means that the current SIT scores are inflated.
The score differences range generally from 0 to 15 IQ points, with
about half the scores revealing changes of greater than 9 points. The
changes in IQ score differences varied greatly from age group to age
group and for IQ levels within age groups. As a result, no simple for-
mula could be used to align SIT scores with the new SB scores.
Until recently little or nothing had been done to revise or re-
align the SIT with the new SB. This is unfortunate, since many people
are still using the SIT but not obtaining comparable SB (1972) scores.
As stated previously, the differences are considerable in most cases.
Fortxinately , Armstrong and Jensen have recently completed a SIT norm
revision, using calibration techniques similar to those used by Thom-
9dike for the SB in 1972. The new SIT norms, which were published in
1981, once again ensure that the SIT and SB scores are comparable.
Unfortunately, no extensive research has been conducted to deter-
mine the relationship between WISC and WISC-R scores, or between WISC-
R and SIT scores. The few studies that have been conducted to determine
the relationship of WISC-R and 1961 SIT scores have shown major depar-
tures from the original findings between the WISC and the 1961 SIT.
In addition, these studies have been extremely limited in terms of:
(1) the sample size—usually between 25 to 40; and (2) the restricted
composition of the samples (e.g., low IQ scores, high IQ scores, spe-
cific handicaps, specific age or grade levels, etc.). In a small
study, Wechsler reported that the correlation between the WISC-R and
the new SB (1972) was approximately the same as the correlation between
6
the Wise and the 1960 SB.^^ Simply stated, the comparability of 1961
SIT scores with WISC-R scores no longer holds true.
The research findings cited previously, as well as other major
studies, reveal the following concerning the SIT, SB (1960), SB (1972),
Wise, and WISe-R.
1. There is a need for a valid individual intelligence test that:
(1) requires no specialized training of its administrators;
(2) can be quickly administered and scored; and (3) yields IQ
scores comparable to the Stanford-Binet and Wechsler tests.
Such a test would provide testing specialists an opportunity
to administer more individual tests themselves, as well as
provide an invaluable source of additional test administrators.
2. In the early 1970 's there was sufficient evidence that 1961
SIT IQ scores could be used with as much confidence as 1960
SB and WISC (V and FS) scores. Unfortunately, the WISC and
SB were revised. There is sufficient evidence that 1961 SIT
scores are no longer comparable to the new WISC-R or the new
SB (1972) scores.
3. Until recently no norm revision of the 1961 SIT has been
attempted.
4. The new 1981 SIT norms once again align SIT scores with the
1972 SB scores.
5. There is evidence that the correlation between the WISC-R and
the 1972 SB is approximately the same as that reported for
the WISC and the 1960 SB.
7
6. To date, no extensive research has been conducted concerning
the WISC-R scores with the 1971 SIT scores. The results from
the limited research that has been conducted show major de-
partures from the findings between the Wise and the 1961 SIT,
similar to those between the 1972 SB and the 1961 SIT.
7. There is a need to determine if the relationship between the
WISC-R and the new SIT is comparable to the previous relation-
ship of the Wise to the 1961 SIT.
Purpose
The purpose of this study is to determine if the new SIT scores
are comparable to WISC-R scores (Verbal and Full Scale) . In other
words, can the new SIT scores be used with as much confidence as WISC-R
scores—to the extent that the old SIT (1961) scores could be used with
the unrevised WISC scores.
The results of this study will provide valuable information to
WISC-R administrators, especially Directors of Special Education, con-
cerning the use of the 1961 SIT and new SIT. If the results are favor-
able, Directors of Special Education will be able to:
1. administer an individual test of intelligence rather than a
group test, for screening or retesting even when an individual
test is not required;
2. use the SIT instead of the WISC-R for screening and retesting
when an individual test is required;
8
3. increase their test administration staff at minimal cost,
since the SIT does not require specialized training.
Each of the following areas will be examined in terms of correla-
tion and mean absolute IQ difference.
I. Overall - for all examinees;
2.-4. Age levels - 6-8, 9-11, 12-15;
5. Female - all female examinees;
6. Male - all male examinees;
7. Same Administrator - for both tests;
8. Different Administrator - for each test;
9. Certified Both WISC-R - Both test administrators were WISC-R
certified, but not necessarily the same administrator;
10.
Certified WISC-R Only - certified administrator for the
WISC-R and a non-certified WISC-R administrator for the
SIT;
II. Order WISC-R First;
12. Order SIT First;
13. Same Day - when both tests were administered on the same day;
14.-16. Different Day - When the administration of the second test
was completed within the following time intervals: 1-14 days,
15-60 days ; 2 months to 18 months
.
It should be noted here that at the time the study was implemented,
the new SIT norms were completed, but not on the market. However, the
writer obtained permission from the authors (Armstrong and Jensen) to
use an advanced copy of the new SIT norm tables for scoring purposes in
this study. (Note: The new SIT was published in November, 1982.)
9
Limitations
1. The sample on which the study was based could not be drawn
at random from the country as a whole or even from New England.
Nevertheless, the participating communities were selected to provide a
balanced variety of urban and suburban settings, and as heterogeneous
a sample as possible. However, results here reported may differ from
those obtained in other samples drawn from other parts of the country.
Technically, the results of the study will apply only to Massachusetts.
2. Subjects for the study were those for whom a WISC-R test was
administered as part of a core evaluation, with many of them being
actual recipients of special education services. In any event, they
will represent only potential or actual recipients of special educa-
tion services.
Definition of Terins
General Terms
1. Director of Special Education ; Refers to the administrator
responsible for all programs concerning children with special
needs
.
2. IQ High ! For purposes of this study, it refers to scores
eJDOve 110.
3. IQ Low ; For purposes of this study, it refers to scores
below 90.
4. Public Law 94-142 ; Refers to a comprehensive federal law
regarding children with special needs.
10
5- SB: Refers to the Stanford-Binet Intelligence Test, Form
B~M . SB 1960 and SB 1972 refer specifically to the 1960 and
1972 editions of this test.
6. SIT ; Refers to the Slosson Intelligence Test . The terms
"old" or "original" SIT imply the 1961 edition, and the term
"new" SIT refers to the 1981 norms.
7. WISC ; Refers to the 1949 edition of the Wechsler Intelligence
Test for Children , whereas WISC-R implies the 1974 edition of
the same test.
Statistical terms
1. Arithmetic Mean ; A common measure of the average of a set of
numbers. Obtained by summing the scores and dividing by the
number of scores. Frequently referred to simply as mean.
2. Average ; One of several measures of central tendency. The
most common measures are the median, the arithmetic mean, and
the mode. The term average is often used as a synonym for
arithmetic mean.
3. Central Tendency ; A measure of average, the typical score,
or the score around which the scores in a set tend to cluster.
4. Coefficient of Correlation ; A measure of the degree of rela-
tionship or "going-togetherness" between two sets of measures
for the same group of individuals. The correlation coefficient
roost frequently used in test development and educational re-
search is that known as the Pearson or product-moment r.
11
Correlation coefficients range from .00, denoting a complete
absence of relation, to +1.00 and to -1.00, indicating per-
fect positive or perfect negative correspondence, respectively.
5. Concurrent Validity : A type of empirical (criterion related)
validity in which a test is validated by correlating scores
on it with scores from a test generally accepted or known to
be valid.
6. Construct Validity ; A form of logical validity established
by showing the relationship between test scores and a con-
struct, a logical criterion, established by analysis of rela-
tionships expected on logical grounds. If one or more fac-
tors, such as age, sex, or scholastic achievement, were logi-
cally related to the development of a skill, evidence of con-
struct validity of a test might include data and reasoning
about the relationship of these factors to the test scores.
7. Criterion ; A standard by which a test may be judged or eval-
uated; a set of scores, ratings, etc., that a test is designed
to measure, to predict, or to correlate with. See Validity.
8. Deviation IQ ; A standard score on an intelligence test based
on the deviation between an individual's score for persons of
that age. Deviation IQ's have a normal distribution with a
mean of 100 and usually a standard deviation of 15 or 16.
9. Dispersion; The extent to which scores in a set tend to be
spread out or to vary about an average.
12
Group Test ; A test that can be administered by one person
to more than one examinee at a time.
Test ; A test that can be administered to only one
person at a time because of the nature of the test and/or
the maturity level of the examinees.
12- Intelligence Quotient (IQ) : Originally the ratio of mental
age to chronological age, multiplied by 100. The average IQ
is 100. Developed as a means of ej^ressing rate of mental
development or brightness. (See also Deviation IQ)
13. Intelligence Test ; A test designed to measure general mental
ability. Now frequently comprised of svibtests of different
types of materials (for example, verbal and nonverbal).
14. Item ; A single question or exercise on a test.
15. Mean Absolute IQ Difference (MAD) ; Refers to what an exam-
iner may ejq^ect in terms of the deviation in scores yielded
by both instruments, that is, by how many points on the aver-
age an examinee's scores will differ having been given both
instruments
.
16. Mental Age (MA) : The age for which a given score on a mental
ability test is average or normal. If the average score made
by an unselected group of children 6 years, 10 months of age
is 55, then a child making a score of 55 is said to have a
mental age of 6-10. Since the mental age unit shrinks with
increasing (chronological) age, MA's do not have a uniform
interpretation throughout all ages. They are therefore most
13
appropriately used at the early age levels where mental growth
is relatively rapid.
17. N; The symbol commonly used to represent the niimber of cases
in a group.
18. Norms ; Statistics that supply a frame of reference by which
meaning may be given to obtain test scores. Norms are based
upon the actual performance of pupils of various grades or
ages in the standardization group for the test. Since they
represent average or typical performance, they should not be
regarded as standards or as universally desirable levels of
attainment. The most common types of norms are deviation IQ,
percentile rank, grade equivalent, and stanine. Reference
groups are usually those of specified age or grade.
19. Norm-Referenced Test ; A test that is interpreted by compar-
ing an individual's score with the scores of other individu-
als in a group defined by some characteristic such as age or
grade in school.
20. Performance Test ; A test involving some motor or manual
response on the examinee's part, generally a manipulation of
concrete equipment or materials. Usually not a paper-and-
pencil test. For example, a "performance" test of mental
ability is one in which the role of language is excluded or
minimized, and ability is assessed by what the examinee does
rather than by what he says (or writes). Mazes, form boards,
picture completion, and other types of items may be used.
14
Examples include certain Stanford-Binet tasks, the Perfor-
mance Scale of Wechsler Intelligence Scale for Children,
Arthur Point Scale of Performance Tests, Raven's Progressive
Matrices.
21- Random Sample ; A sample drawn in such a way that every mem-
ber of the population from which it is drawn has an equal
and independent chance of being included, thus eliminating
bias from methods of selection.
22. Range ; For some specified group, the difference between the
highest and the lowest obtained score on a test; thus a very
rough measure of spread or variability, since it is based
upon only two extreme scores. Range is also used in refer-
ence to the possible spread of measurement a test provides,
which in most instances is the number of items in the test.
23. Ratio IQ : See Intelligence Quotient.
24. Reliability ; The consistency of measurement. The accuracy
with which a test measures whatever it measures . Expressed
in the form of a coefficient of correlation or the standard
error of measurement.
25. Representative Sample ; A sample drawn in such a way as to
ensure that the population is represented with regard to
certain characteristics. Often, a sample stratified with
respect to particular criteria such as geographic region or
socioeconomic level.
15
26- Standard Deviation : A measure of the dispersion of scores
around the average of the set. Based on the difference be-
tween each score and the mean. The greater the dispersion of
scores, the larger the standard deviation, and vice versa.
27. Standard Error of Measurement (SEM) ; An estimate of the
magnitude of errors likely to be present in scores obtained
from a test. It can be estimated from the standard deviation
of the test and its reliability coefficient. In approximate-
ly two-thirds of the cases, the obtained scores would not
differ from the true scores by more than one standard error
of measurement.
28. Standard Score : A derived score expressed on a known scale.
Standard scores are expressed in terms of the deviation of
the raw scores from the mean in units of the standard devia-
tion of the distribution, hence, the name standard score.
Examples: Z-scores, sigma scores, and T-scores.
29. Standardized Test : A test designed to provide a systematic
sample of individual performance, administered according to
prescribed directions, scored in conformance with definite
rules, and interpreted in reference to certain normative in-
formation. Some would further restrict the usage of the term
"standardized" to those tests for which the items have been
chosen on the basis of experimental evaluation, and for which
data on reliability and validity are provided.
16
30. Validity ; The extent to which a test measures all of what it
is intended to measure and nothing else. The value of a test
for a specific purpose—selecting students who will succeed
in a vocational program or the extent to which a test consti-
tutes an adequate measure of the objectives and content of
a course.
31. Variability : The dispersion of scores around the average of
the set. The heterogeneity of a set of measures.
32. Variance : A measure of the dispersion of a set of scores
about the mean of the set that is equal to the square of the
standard deviation. A technical term not to be confused
with variability.
FOOTNOTES
Massachusetts Department of Education. Chapter 766 Regulations,July 1981.
2Wechsler, D. , "Wechsler Intelligence Scale for Children -
Revised." Psychological Corporation, 1974.
3Thorndike, R. , Stanford-Binet Intelligence Scale, 1972 Norms
Tcibles . Houghton Mifflin Company, 1972.
Slosson, R. , "Slosson Intelligence Test." Slosson EducationalPublications, 1963.
^Terman, L.M. , and Merrill, M.A., Stanford-Binet IntelligenceScale . Boston: Houghton Mifflin Company, 1960; Slosson, R. , "SlossonIntelligence Test." Slosson Educational Publications, 1963.
gArmstrong, R. ; Jensen, J.; and Reynolds, J., "A Comparison of
Individual Intelligence Tests Used for Special Class Placement,"
Slosson Intelligence Test Manual , Spring, 1974.
7Terman, L. , and Merrill, M. , Stanford-Binet Intelligence Scale,
Manual . Houghton Mifflin Company, 1972; Thorndike, R. , Stanford-
Binet Intelligence Scale, 1972 Norms Tables . Houghton Mifflin Com-
pany, 1972; Wechsler, D. , "Wechsler Intelligence Scale for Children -
Revised." Psychological Corporation, 1974.
^Armstrong, R. , and Jensen J., "The Slosson Intelligence Test
1981 Norms Tables: Application and Development." Chelmsford, MA:
Metrics Press, Inc., Fall, 1981; Sattler, J. ,"Assessment of Child-
ren's Intelligence." W.B. Saunders Company, 1979; Thorndike, R.
,
Stanford-Binet Intelligence Scale, 1972 Noms Tables . Houghton
Mifflin Company, 1972.
^Armstrong, R. , and Jensen, J., "The Slosson Intelligence Test
1981 Norms Tables: Application and Development." Chelmsford, MA:
Metrics Press, Inc., Fall, 1981; Thorndike, R. ,Stanford-Binet
Intelligence Scale, 1972 Norms Tables . Houghton Mifflin Company,
1972.
^*^Wechsler, D. ,Wechsler Intelligence Scale for Children:
Manual. New York: Psychological Corporation, 1949.
17
CHAPTER II
RELATED RESEARCH
The purpose of Chapter II is to report the findings of related
research studies concerning the Slosson Intelligence Test with the
Stanford-Binet and the Wechsler Intelligence Tests. The following
abbreviations will be used to identify the specific version or edition
of the three tests.
SIT; 1961
SB: 1960
WISC: 1949
SIT-R: 1981
SB-R; 1972
WISC-R: 1974
Slosson Intelligence Test
Stanford-Binet Intelligence Test
Wechsler Intelligence Scale for Children
Slosson Intelligence Test
Stanford-Binet Intelligence Scale
Wechsler Intelligence Scale for Children -
Revised.
The research studies are presented under the following five
headings
;
1. SIT and the SB
2. SIT and the SB-R
3. SIT--R and the SB-R
4. SIT and the WISC
5. SIT and the WISC-R
18
19
1. SIT and the SB
The SIT was published by Richard L, Slosson in 1961.^ His stated
purpose for constructing the SIT was to provide a test which would re-
quire minimal training to administrate, took approximately 20 minutes
to administer and score, and yielded IQ scores which were close approx-
imations to the SB IQ.
Slosson used the SB as the criterion in establishing the validity
of the SIT. Table 1 reports the concurrent validity data obtained by
Slosson in his sample of 701 persons ranging in age from four years to
18+ years. It can be seen in Table 1 that the concurrent validity
. . 2coefficients ranged from .90 to .98. Slosson concluded that the SIT
correlated with its criterion, the SB, as well as the SB correlated
with itself.
Table 1 also shows that the mean absolute IQ difference (MAD)
ranged from 4.4 to 6.7, approximately the same as the Standard Error of
Measurement (SEM) of the SB. When the distributions for the variables
in question are normal and have equal means and standard deviations,
the value of MAD is approximately 1.1284 times the SEM.^
Slosson administered both tests to 560 of the 701 examinees in the
original validity study shown in Table 1. In order to avoid any bias,
an independent testing experiment was utilized for the remaining 141
examinees. For this experiment, Slosson administered the SB and the
SIT tests were administered by a group of untrained SB examiners. It
can be seen in Table 2 that the results were comparable; .92 concurrent
validity coefficient, and a 6.1 MAD.
20
TABLE 1
1961 SLOSSON CONCURRENT VALIDITY STUDY
Means, Standard Deviations, Correlations and Mean AbsoluteIQ Score Point Differences (MAD) of the S-B and
SIT Intelligence Tests
Means Standard Deviations
Age N r SB SIT SB SIT MAD
Overall 701 .96 98.9 98.9 21.7 22.0 5.7
4 27 .90 116.6 114.6 19.7 18.7 6.7
5 23 .93 102.1 101.5 20.7 18.0 5.6
6 61 .98 100.7 101.3 20.7 20.2 4.4
7 71 .98 98.9 98.4 23.5 20.9 5.9
8 44 .94 95.5 95.5 17.6 17.0 5.3
9 45 .97 100.7 100.6 25.1 23.7 5.1
10 40 .94 96.1 97.2 23.9 24.6 6.1
11 51 .96 93.1 92.6 21.4 22.0 4.9
12 36 .97 94.0 94.1 22.4 24.6 4.6
13 57 .96 96.3 97.0 23.4 24.9 5.0
14 66 .97 92.7 92.4 20.4 21.5 4.4
15 56 .94 92.7 91.7 18.8 18.2 5.1
16 39 .96 97.6 97.5 23.7 24.0 4.7
17 23 .94 106.0 106.6 16.9 16.7 5.0
18+ 62 .97 101.7 102.5 31.8 31.2 5.9
21
TABLE 2
1961 SLOSSON CONCURRENT VALIDITY STUDYINDEPENDENT EXPERIMENTAL STUDY
Means, Standard Deviations, Correlations and Mean AbsoluteIQ Score Point Differences (MAD) of the S-B and
SIT Intelligence Tests
Means Standard Deviations
Age N r SB SIT SB SIT MAD
4-19 141 .92 107.7 107.2 20.2 19.9 6.1
Studies conducted in the late 1960 's and early 1970 's largely con-
firmed what Slosson had concluded in 1961. The two foremost studies
4were undertaken by Armstrong and Jensen in 1970 and Armstrong, Mooney
5and Jensen in 1971.
The 1970 Armstrong and Jensen study used for its testing sample
490 students (ages 6 to 14) enrolled in 10 public school systems in
Northeastern Massachusetts. Each student was administered both a SIT
and SB within a two week period of time. In order to obtain compara-
tive results and also to avoid bias, the tests were administered using
various combinations of personnel. The personnel or administrators
were classified in three ways, namely, professionals, trainees and
teachers. For the purpose of the study: (1) professionals were defined
as highly trained and experienced personnel in the field of testing
(limited to three in the study); (2) trainees were defined as part-
time graduate students (mostly teachers) who were enrolled in a course
concerned with administering, scoring and interpreting the SB and who
had administered numerous SB's under supervision before administering
22
any tests in this study; and (3) teachers were defined as personnel
with no knowledge concerning the administration of the SB.
Using the Pearson-Product Moment formula, a coefficient of correla
tion was computed separately for each of the following 21 categories.
1. Overall - for all 490 testees;
2.-10. Age Levels - for each age level separately (6-14)
;
11. Male - all male testees;
12. Female - all female testees;
13. Same Administrator - for both tests;
14. Different Administrator - for both tests, but not
necessarily the same administrator;
15. Professional Administrator - for both tests, but not
necessarily the same administrator;
16. Trainee Administrator - for both tests , but not
necessarily the same administrator;
17. Professional and Teacher Administrators - SB by a profes-
sional and the SIT by a teacher, neither was aware the other
test was given or was going to be given;
18. SB Administered First;
19. SIT Administered First;
20. Same Day - when both tests were administered on the same
day;
21. Different Day - when the administration of the second test
was completed between one and fourteen days after the first
administration.
23
Additionally, the mean absolute difference (MAD) between IQ scores
from the two scales was computed for each of the preceding categories.
Table 3 reveals that the overall study Pearson-Product Moment cor-
relation between the SB and the SIT was .93. Also, the range of corre-
lation for the 20 subcategories was from .89 (ages 8 and 9) to .96
(age 13). All correlations were significant beyond the .001 probabili-
ty level. Table 3 also reveals that the mean absolute IQ score point
difference (MAD) between the two tests was 5.64 which is approximately
the same as the standard error of measurement of 5 IQ score points of
the SB. The range of average IQ score point differences for the 20
subcategories was from 4.03 (professional and teacher) to 6.15 (age
10) . These average score point differences present further evidence of
the comparability of the scores of these two tests.
The researchers concluded that for this study, consisting of 490
testees, the SIT appears to be measuring the same thing as the SB,
that is, SIT and SB scores are comparable. No test could correlate so
closely with another, and not be a good measure of it. Thus, the find-
ings suggest that the SIT can be used as a valid screening and retest-
ing substitute for the SB and provide: (1) an opportunity for more in-
dividual intelligence tests to be given; (2) a source of additional
test administrators; and (3) specialized personnel with more time for
their other responsibilities.
Most studies concerning the validity of the SIT used heterogeneous
populations. It was the purpose of the 1971 Armstrong, Mooney, and
Jensen study^ to determine the applicability of earlier findings to a
restricted, homogeneous population of special class or low IQ
24
n
as
Ehcn
oCTi
zuw2W1-3
+J
c*Ho
0)
>H
oO 03
W) 4J03
O! 03
H Eh
0) 03
P OP CiH 03
O O'03 -H
cm03
sT3C(0
03
co•Hp
03+J
CHEhHWTJC(0
nI
(0 w
«
w
cn
ocr>
03p(0
u
TfncO'O’vOmooooocNtNrorocNvo^ncN^cxsrHvDiHvD<Ti03rHroiniDHmro<Nr'^OOir)OOrno3invOLni/3ir)vDLnin^Lnmi/3Lnir)mvD^ir)Lninm
cn O rH 03 03 rH n CN CD cn rH cn CVJ CN cn rH •cr CN CN cn CNCTi 03 03 00 00 03 03 03 CJ3 03 03 03 03 03 03 <33 03 <33 03 03 03
CN r' 00 'O' m rH 'O' m n CN cn n- 00 00 cn rH m 00 C" <33
CD rn CN 03 03 rH CD CN CD 00 03 CN in 03 03 CD iH 'T '3'
o r' r' C" Ch 00 o o rH o 00 03 00 03 00 00 r' 00 O 00 03CN rH rH rH rH rH CN CN CN CN rH rH rH rH rH rH rH rH CN iH rH
O rH 03 CD 'O' CD n 'O' CD O'- rH in cn "O' rH 03 03 in m 'TCD 00 CN in O 03 03 in 03 03 03 00 CN m fH r' '3' o '3'
<T> 00 CD CD 00 03 00 03 03 r' 00 00 CD in 03 00 r' <33 00 00rH iH rH rH iH rH iH rH rH rH rH iH fH rH rH rH rH rH rH rH rH
o 00 rH "O' cn iH 03 o o o cn cn o o in iH r' 00 rHCD o 03 00 o 'O' CD o 'O' CN C" r' cn "O' CD cn rr cn o
ID o CN CD CD CD 00 00 CD 00 r' 03 cn rH rH 00 00 O'- CD '3'
o o iH rH O O O O o O o o o o o rH o o o o fHrH iH rH iH rH rH rH rH fH iH rH rH rH rH rH rH rH rH rH rH rH
o O in CD n rH CD 00 o CD 03 rH CN CD o CD rH 'f CDCD m CN o 03 rH CD in 'O' 'O' cn CD 03 CD m CD cn 'T CD
CD 03 o m in in C" CD CD 00 cn o 03 03 CD p" m fH
O o o iH o o o o O O o O O O o O O O o o rH
rH (H rH rH rH rH iH fH iH rH rH fH rH rH rH rH iH rH rH rH rH
o o 03 00 m 00 00 o CO o cn 'O' CD CN CD CN m in o o(J3 in in in in m in m m m m in "O' 'O' in 00 in 03 03 00 rH
CN CN cn rH iH CN CN iH CN CN
c cn <u
•p >1 0E (TJ h u-o Q Q
• < cp G CU '3' 00 •H P E rH •pp e 0) <o 1 p<0 -o x: cn rp <0
p sC • u \ \ rH+J c CCJ c c <u
</) rH 'H Q) 0 0 p•H m E Eh •p •p pC c '9 +J p +J +J 0
H o rH (N n ^ 03
rrtvOr^00O3iHrHrHrHrH rH
U 03 203030303030303030303rH£> triCT'CTitT'CT'IT'CJ'tT'Cr'ig 03
oo
03
£•P
'Oco>103
XI
4-3
cmo
•HCtr-H03
03
P(0
03+J
C03
•Ho•HOhip
OO
stvitlPutB (Trtblp 4). Thp ehlMfPU In tl\P Mtu«ty rp«>up»t in pvjp rmw m (o
14 yppru. Tliplr 8U 1Q*b «"flngp«1 fi-om 4!i to MM, with « i«Pan of 72, Thp
Btvhty Mctmpl** OOMMiMtPtl of 147 l4t\lt1PI\tM who WPfP PVplnfliPtl for t^XtHihl**
elrtMH vUpo«?mpnt. Krtch Mtvulpnt w^m nlwtpi'p*! hotlt thp 8TT am*! SM
within two-w»*Pk t'wri'irt of timp. tn or*1pr to ohtpln o»w\t'flrrthlP rPHultw
rtiul to rtvoiii bipM, thp oi‘<lpr of <4*1mlniBtrrttion of thp InMtrntnpntH wpb
,1ptprml n^vl ^t r«n\<1oi«. Using thp I'p^rson-brodnet M»M«pnt foimils, a eo«
pffleiont of eorrslrttion wss t'orntnit hptwppn thP SM ^inl StT tQ seores.
Also, thp tnpsn shsolutp IQ soors ^>olnt Uiffprsnop (MAU) wss <'t<n\j.intp«1,
TAMt4« 4
AMMJJTKONU, MtkINKY ANU UKNMKN 1V>71 8TUl)Y
Mpsms, Stsiulsrd Usvistions, V'ssrson-l'rodvict MtM«snt
CorrsUtion, sjvd Mpsn Absoluts IQ 8oors Oiffpipnos (MAU)
for thp 8tsnf©rd-Minst {8-M) snd Slosson (8iT)
lntp4ligst\op Tost Seorps
itsnford-Minst8-8
SlossonSIT
N 147 147
X 72.07 73.0M
8U 10. Ml 10.05
r .Ml
w* .oi
Mppi\ Absoluts IQ Uiffprsnos (MAU) 4.4M
*Adiusts(1 for rsstrietion of rsngs.
26
It can be seen in Table 4, the Pearson—Product Moment correlation
between SB and the SIT scores was .85. Although this was slightly lower
than the results reported for validity studies of heterogeneous samples
(approximately .90) , it is impressive considering that the sample from
which these data were drawn showed rather severe restriction in varia-
bility. The adjusted or estimated correlation for a similar sample with-
out restriction of range was .93. The mean absolute IQ score difference
(MAD) between the two instruments was 4.48.
The average administration time for the SB was 64 minutes, com-
pared to 17 minutes for the SIT. Therefore, on an average, the SB took
approximately four times longer to administer.
The researchers concluded, that for these students, scores obtained
from the SIT could have been used with as much confidence as scores ob-
tained from the SB for special class placement. The results suggest that
one of Slosson's purposes in constructing the test—to develop an abbre-
viated test that could be used as a valid screening and retesting in-
strxament—was achieved.
7In a 1980 review of the literature. Chimera reported the findings
of 18 major studies concerning the SIT and the SB. The range of corre-
lations was from .60 to .98 with a median concurrent validity coefficient
of .90. The 18 studies involved 3,253 examinees, with an average of 181
examinees per study. Table 5 contains a summary of the studies.
27
TABLE 5
CHIMERA REVIEW OF 18 RESEARCH STUDIES CONCERNINGTHE SIT AND THE SB
Author ( s
)
& Year r
SBMean SD
SITMean SD
Sample Size& SubjectTypology
Slosson1961
.96 98.90 21.75 98.98 22.00 701 Aged 4 to 18
and over
Slosson1961
.92 107.2 19.9 107.7 20.2 141 Aged 4 to 19
Poissant1967
.89 69.3 9.7 68.0 8.7 30 Aged 4 to 50
DeLapa1968
.60 66.37 8.94 68.86 9.24 57 Educable re-
tarded Aged 8
to 12
DeLapa1968
.90 103.97 13.72 99.92 13.20 60 Students Aged
8 to 12
Keany1968
.84 * 14.4 it 11.1 32 Sixth Grade
Children
Hill1969
.84 68.0 8.64 66.37 7.29 30 Aged 7 to 12
"adjusted curric.
(IQ 50 to 80)
Jongeward1969
.76 71.70 7.68 69.30 8.08 30 Educable Spec-
ial Class Pupils
Aged 7 to 10
Armstrong& Jensen1970
.93 107.6 20.42 106.6 19.60 490 Public School
Students Aged 6
to 14
Carlisle
(6 yearinterval)
.91 19.19 16.78 21.13 15.78 122 Profound &
Severe Retardates
Data not available.
28
Tcible 5 (cont.)
Sample SizeAuthor (s) & SB SIT & Subject
Year r Mean SD Meam SD Typology
Armstrong& Mooney1971
.94 101.30108.65
18.1917.91
100.90109.51
15.2418.58
204 Public SchoolPupils Aged 6 to
14
ArmstrongMooney &
Jensen1971
.85 73.08 10.03 77.07 10.81 147 Special Class
Pupils Aged 8 to14
Johnson &
Johnson1971
.79 85.76 13.80 87.86 13.65 29 Head StartChildren
Stewart,et al.
1971
.94 79.4 19.6 79.1 18.9 81 Special Educa-
tion ReferralsAged 6 to 15
Armstrong& Jensen1972
.92 107.28 19.23 106.44 18.20 724 Students Aged
6 to 14
Lamp &
Traxler1973
.75 91.8 12.7 231 Disadvantaged
Head Start & 1st
Graders
Ritter1973
.92 99.9 15.74 95.3 14.05 44 Students Aged
4 to 13
Stewart . 84
1974
(4 year interval)
60.37 17.58 61.22 16.58 100 Special Class
Students Aged 6
to 16
*Data not availcible.
29
2. SIT and the SB-R
The value of the SIT in accurately estimating the corresponding SB
IQ was firmly established nearly ten years ago. Studies cited in the
previous sections showed that the SIT was exceptionally reliable and
valid as an estimate of a person's SB 10 (Armstrong and Jensen, 1970,
1973) . The creator of the SIT, Richard L. Slosson, managed to capture
the essential qualities of the SB in a much shorter and less expensive
SIT. Since the SIT could be administered accurately by persons who had
only a modest background in intelligence testing, and could be adminis-
tered and scored in about twenty minutes, its popularity as a screening
device was easily understood.
While the bulk of the research which showed the accuracy and utility
of the SIT as an estimate of the SB was being conducted, the SB was it-
self undergoing an extensive renorming. While the SIT had been an ex-
cellent estimate of the 1960 norms for the SB, the extensive changes
introduced in the 1972 SB-R norms meant that the old relationship between
the SIT and the SB no longer held. As a result of the SB-R, differences
of 10 or even 15 IQ points were being yielded by the two instr\aments in
measuring the same individual.^ Prior to the 1972 SB-R, such differen-
ces were virtually unheard of. While the scores for most examinees
given by the SB-R and the SIT were still close, it was clear that the
1972 SB-R norms had distorted the parallelism which had characterized
both instrioments until that time.
Tables 6 and 7 reveal the results of a 1980 study conducted by
Armstrong and Jensen as part of their renorming of the SIT in 1981.
30
TABLE 6
ARMSTRONG AND JENSEN 1981 STUDY
Differences Between Unrevised Stemford-Binet and UnrevisedSlosson IQ's by Levels of Chronological Age and Intelligence
ChronologicalAge Group in
Years-Months Below 84
SB IQ Group
84 - 116 Above 116 Row Total
6-6 and Mean - 2.64 - 1.38 2.22 - 0.67
Below SD 4.82 7.23 8.47 7.38
N 22 69 32 123
6-7 to Mean - 2.95 - 1.48 1.90 - 1.23
10-6 SD 5.40 6.88 8.34 7.02
N 86 283 70 439
10-7 to Mean - 0.03 - 0.82 - 1.81 - 0.79
13-6 SD 4.94 6.89 8.47 6.77
N 73 195 48 316
13-7 and Mean 0.49 - 0.46 1.65 0.20
Above SD 5.92 6.32 7.10 6.32
N 85 112 34 231
Column Mean - 1.02 - 1.10 0.94 - 0.74
Totals SD 5.60 6.83 8.28 6.86
N 266 659 184 1109
31
TABLE 7
ARMSTRONG AND JENSEN 1981 STUDY
Differences Between Revised Stcinford-Binet and UnrevisedSlosson IQ's by Levels of Chronological Age and Intelligence
ChronologicalAge Group inYears-Months Below 84
SB IQ Group
84 - 116 Above 116 Row Total
6-6 and Mean - 8.82 - 8.91 - 6.50 - 8.27Below SD 5.96 7.96 8.41 7.78
N 22 69 32 123
6-7 to Mean - 4.37 - 4.36 - 3.09 - 4.1610-6 SD 5.54 6.86 8.23 6.87
N 86 283 70 439
10-7 to Mean - 1.44 - 3.24 - 5.08 - 3.1013-6 SD 4.98 6.92 8.67 6.90
N 73 195 48 316
13-7 and Mean - 1.34 - 4.17 - 3.82 - 3.08Above SD 6.07 6.52 7.00 6.54
N 85 112 34 231
Column Mean - 2.97 - 4.47 - 4.34 - 4.09
Totals SD 6.01 7.11 8.20 7.08
N 266 659 184 1109
32
These tables illustrate dramatically both the adequacy of the SIT as an
estimate of the corresponding SB IQ prior to the renorming of the SB and
the extent to which the SB-R served to undermine the usefulness of the
SIT. Both tables present N's, means and standard deviations for IQ dif-
ferences observed in combinations of age level and SB 10 level. The
columns of each tcible divide the total group into three levels for IQ.
A low group was formed consisting of all subjects who scored more than
one standard deviation below the mean on the SB. A corresponding high
group included subjects who scored more than one standard deviation above
the mean on the SB. Subjects who scored between 84 and 116 on the SB
constituted a middle, "average" group.
Table 6 shows that overall, the SIT and SB averaged less than one
IQ point difference in the IQ's they yielded prior to the renorming of
the SB. The marginal entries also show good correspondence between the
two instruments for different age levels and for the three IQ groups,
the largest differences observed being on the order of one IQ point.
Since the cell entries are formed by subtracting the SIT IQ from the
corresponding SB IQ, the negative entries mean that the SIT was over-
estimating the SB. It may be seen that there was a tendency for the SIT
to give slightly inflated estimates of the SB IQ's which it was estima-
ting in most of the age and IQ categories, though that tendency was not
marked or apparently serious.
The results which appear in Taible 6 serve to justify the claims
which were made in years past that the SIT did a good job of estimating
SB IQ's. Table 7 presents corresponding results after the SB had been
renormed (SB-R) . These results show quite clearly why the SIT needed to
33
normed as well. Note that the overall IQ difference in Table 7 is
-4.09 points. This means that following the SB's renorming, the SIT
was clearly overestimating the SB-R. Without exception, the marginals
and the cells show negative entries indicating a systematic error which
had become attached to SIT IQ's. Though the SIT was systematically
overestimating corresponding SB-R IQ's, it should be emphasized that
the amount of the overestimation varied as a function both of chrono-
logical age and IQ level. Note that the errors were most serious for
young children and for bright examinees. Least affected were the older
children and adults who were in the low IQ group, yet there was a ten-
dency for their SB-R IQ's to be overestimated as well. Quite clearly,
something had to be done to realign SIT IQ's with the renormed SB-R
IQ's if the SIT were to continue to be used as an estimate of the cor-
responding SB-R IQ's.
3. SIT-R and the SB-R
10In 1981, Armstrong and Jensen published the 1981 SIT-R norms.
Table 8 presents the results which bear upon the success of the norming
procedures. Table 8 shows that, as a result of the norming procedures,
the average difference between IQ's generated by the two instruments
has been reduced from 4.09 points (with the SIT systematically overesti-
mating the SB-R) to four hundredths (.04) of an IQ point. In the same
vein, IQ differences attributable to chronological age have essentially
disappeared, the largest difference noted being on the order of one-half
(.5) an IQ point. These differences are clearly due to minor sampling
34
TABLE 8
ARMSTRONG AND JENSEN 1981 STUDY
Between Revised Stanford—Binet and RevisedSlosson IQ's by Levels of Chronological Age
and Intelligence
ChronologicalAge Group inYears-Months Below 84
SB IQ Group
84 - 116 Above 116 Row Total
6-6 and Mean - 0.45 - 1.52 2.00 - 0.41Below SD 5.45 6.69 8.25 7.04
N 22 69 32 123
6-7 to Mean - 1.21 - 0.12 1.60 - 0.0610-6 SD 5.10 6.63 7.84 6.61
N 86 283 70 439
10-7 to Mean - 1.33 0.66 3.25 0.5913-6 SD 4.29 5.84 7.52 5.97
N 73 195 48 316
13-7 and Mean - 1.41 - 1.10 2.62 - 0.67Above SD 5.39 5.69 6.19 5.80
N 85 112 34 231
Column Mean - 1.24 - 0.20 2.29 - 0.04
Totals SD 5.00 6.29 7.52 6.33
N 266 659 184 1109
35
error. In Siam, it is once again the case that the SIT-R may be used
with confidence in estimating corresponding SB-R IQ's.
4. SIT and the WISC
Although the SIT was developed to seirve as an abbreviated or short
form SB, numerous studies were conducted to detennine the relationship
of the SIT to the WISC. Again, the most comprehensive study was con-
ducted by Arinstrong, Jensen and Reynolds in 1974.^^ As a result, this
section will follow the same reporting format as the "SIT and the SB"
section; a detailed report of the Armstrong and Jensen finding, fol-
lowed by a summary of other pertinent studies.
The purpose of the 1974 Armstrong, Jensen and Reynolds study was
to deterTTiine if the SIT could be used with as much confidence as the
WISC for special class placement. The sample consisted of 198 elemen-
tary school students in a Northeastern Massachusetts community who were
evaluated for special class placement. Each student was administered
both the SIT and the WISC within a two week period of time. Correla-
tions among the variables were coirputed and, since the saunple from
which the data were drawn showed rather severe restriction in variadail-
ity, correlations were also estimated for the corresponding unrestric-
ted sample. In addition, the mean absolute difference (MAD) in IQ
across all 198 subjects was determined for each pair of variables.
As shown in Table 9, the means and standard deviations of the SIT
cind Wise-Full Scale (W-FS) scores were very nearly the same (and not
significantly different) in this sample. None of the variables differed
significantly from any of the others with respect to variance, but
b«eflUHo of Hip Urrjp niirtil.ar of siihjpof.s omplf^yofl, ql| ruMni.qr I Hf.rm aJom.j
tnppns wprp Bl.jr.lrloflr.h nt fho . 0 ^ IpvpI, wHIi Mip p|K,vp-not p.1 PKnapMon
of <:hp oofn|.)prl pohi
'rAhi,r. '
AMMJiTMONU, .IKNUfilN ANIJ MI'lYNOUJfl r-»74
Mpptis, Sfpttdprrl iJpvl pti OOP ,
prifl Cottpl qf Ions Awnti'j fhpVptlphlpp (N^l'^M), CortplphloriH Ahovp hhp iJlptjrjfipl pro
Corroetofl for Mophrlefloft of Uprujo
iVfil SloPHon pn*l WtHC
Topt K B BIT W-V W-P W-PB
BIT 0.74 1 . 00 .01 .75 .00
W-V 70. 10 M.20 .75 1 .00 .72 .01
W-P 71. 5M 12.57 .02 .50 1 . 00 .00
W-fB 00 . 20 10.00 .70 .Ml .01 1.00
Corrolpfelonp botwoon flip StT Phfl WfHC^Vorbpl (W“V) pml W-lr'B pooroo
wpfp fpl rly ptiljp tpfihip 1 oohpldof Iny tlio rpphflel o»1 vpf Ipbl 1 tfy In tho
Hpinplp pft*1, wlion pptl/npfcpp wofp nip*lo of Mip oot-folpflonp to bo oxpootorl
lf» Rpmi-'lop liMfoptrletod in vprlpbiltty, Sl'l' peotop woro vory ptrongly
rolptorl to W“V pnti poerop (poo TpJjIo , ^iIjp fjoin|.>pfpbn 1 ty of
poorop on tliP HfT Pnd W-fH wpp diip Iprfjoly to tho olopo flyroomont bo-
Kwpon tho HTT pnd W«V poofop, WlhG-hofforrtipnop (W-h) poorop wofo not
4 P woll pptlmptod by HIT pootop op woro W-V pnd W-f'h poorop.
l»'of tlip roptrlefiod spmpl»» tho mopn pbpohbo dlfforoneo In 10
peofpp ylpipiiod by tfio stT pnd w-V or W-Ph (poo TpJ»)o KJ) wpp only
pllghtly Iprqof (lipn I ho ptpndprd orror of moppuromont for fhopo
instruments. The researchers concluded that the evidence from this
study indicates that the SIT, W-V and W-FS are indeed good estimates
of each other.
37
TABLE 10
ARMSTRONG, JENSEN AND REYNOLDS 1974 STUDY
Mean Absolute IQ DifferencesSlosson and WISC
Test SIT W-V W-P W-FS
SIT 0.00W-V 5.03 0.00W-P 8.79 8.51 0.00W-FS 5.42 4.24 5.77 0.00
12In Qiimera's 1980 review of the literature concerning the SIT
and SB, he also cited the findings of 17 studies concerning the SIT and
WISC. The median correlations of the SIT with the WISC-V, P and FS
were .82, .62 and .75 respectively. The studies included 1,785 examin-
ees with an average of 88 per study. Table 11 presents a summary of
the studies.
5. SIT-R and the WISC-R
Just when it appeared that SIT IQ scores could be used with as much
confidence as WISC-V and WISC-FS IQ scores, the WISC was revised (now
WISC-R). The 1974 WISC-R includes the same subtests, but major changes
were made in the content or items of the various subtests. Unfortunate-
ly, no extensive research has been conducted to determine the relation-
ship between WISC and WISC-R scores, or between WISC-R and SIT scores.
38
TABLE 11
CHIMERA REVIEW OF THE 17 RESEARCH STUDIESCONCERNING THE SIT AND WISC
Wise SIT Sample SizeAuthor (s) and SubjectSt Year Criterion r Meeui SD Mean SD Typology
Houston WISC-FS+ .60 104.60 13.21 104.23 11.03 56 SchoolSi Otto WISC-V .91 Children Aged196813 WISC-P .62 7 to 14, Avg.
to Above Avg.
Intelligence
Jongeward WISC-FS .54 61.20 7.11 64.90 8.16 30 Educable19691"^ WISC-V .85 Special Class
WISC-P Pupils Aged12-15
Meissler WISC-FS .82 80.4 16.0 81.9 15.0 200 Referred
197015 WISC-V .83 Pupils from
WISC-P .70 Grades 1 to 8
Pate WISC-FS .84 79.5 ic 78.0 * 100 Suggested
197o16 WISC-V .87 Mental Retard-
WISC-P .68 ates Aged 6-15
Swanson Si WISC-FS .44 110.0 12.6 106.0 12.1 64 Second
Jacobson WISC-V .64 Graders with
1970^ WISC-P .10 Learning Prob-
lems
Lessler Si WISC-FS .67 All IQ's Below 85 135 Referred
Galinsky WISC-V .68 Pupils Aged
I974 I8 WISC-P .57 7-18
''’fS = Full Scale; V = Verbal; P = Performance.
Data not available.
39
Table 11 (Cont.)
Author (a)
& Year Criterion
Maxwell WISC-FS .881971 WISC-V .93
WISC-P .73
Stewart, WISC-FS .94et al. WISC-V .76197120 WISC-P .51
Jerrolds
,
WISC-FS .74et al. WISC-V .76197221 WISC-P .51
Machen WISC-FS .44197222 WISC-V .52
WISC-P .27
Lanp & WISC-FS .74
Traxler197323
Pikulski WISC-FS .75197324 WISC-V .73
WISC-P .59
Sinco WISC-FS .94197325 WISC-V .94
WISC-P .84
Wise SIT
Meem SD Mean
105.2 21.27 102.2
75.3 11.7 70.2
100.25 15.15 95.29
151.23 11.85 133.33
91.8 12.7 *
104.6 16.3 101.2
89.4 20.2 86.8
Sample Sizeand Subject
SD Typology
17.14 60 6th Gradeand Special Ed-ucation Stu-dents Aged 11-
14
12.7 76 SpecialClass PupilsAged 6-13
13.48 51 Students Re-ferred for Read-ing ProblemsAged 6-14
10.44 75 GiftedChildren Aged9-11
* 231 Disadvan-taged HeadStart & 1stGrade Children
11.9 59 Referred to
Reading ClinicAged 7.5-14.6
18.6 80 ReferredChildren Aged6-15
*
Data not available.
40
Table 11 (Cent.)
Author (s)
& Year Criterion r
WISC SIT Sample Sizeand SubjectTypologyMean SD Mean SD
Armstrong WISC-FS .76 68.95 9.74 69.26 10.09 198 Special& Jensen .90** Class Pupils1974 WISC-V .75
.93**
WISC-P .62.75**
Stewart & WISC-FS .83 69.53 13.46 71.17 13.04 128 SpecialMyers WISC-V .79 (2 1/2 year interval) Class Students197427 WISC-P .72 Aged 6-16
Covin WISC-FS .60 75.44 11.16 70.83 9.23 65 Low Achiev-197728 WISC-V .69 ers
WISC-P .58
Martin & WISC-FS .79 it * * it 33 PublicKidwell School Students197729 Aged 12.3-14.1
*Data not available.
it it
Recalculated for restricted range of scores (Jensen & Armstrong,
1974)
.
The few studies that have been conducted to determine the relationship
of the WISC-R and SIT scores have shown major departures from the orig-
inal findings between WISC and SIT scores. In addition, these studies
have been extremely limited in terms of sample size and the restricted
composition of the sample.
In a small study, Wechsler^^ reported that the correlations between
the WISC-R and SB-R are approximately the same as the correlations be-
tween the WISC and SB. Armstrong and Jensen^^ reported that the corre-
lations between the SIT-R and SB-R is approximately the same as the
41
correlations between the SIT and the SB. Now, there is a need to de-
termine if the SIT-R correlates with the WISC-R the way the SIT cor-
related with the Wise.
The following is a summary of the research findings cited pre-
viously concerning the SIT, SB, WISC, SIT-R and WISC-R:
1. There is a need for a valid individual intelligence test
that (1) requires no specialized training of its adminis-
trators; (2) can be quickly administered and scored; and
(3) yields IQ scores comparable to the SB and Wechsler
tests. Such a test would provide testing specialists an
opportunity to administer more individual tests themselves,
as well as provide an invaluable source of additional test
administrators
.
2. In the early 1970 's there was sufficient evidence that SIT
IQ scores could be used with as much confidence as SB and
WISC (V and FS) scores. Unfortunately, the WISC and SB
were revised. There is sufficient evidence that SIT scores
are no longer comparable to the new WISC-R or the SB-R
scores
.
3. Until recently no norm revision of the SIT had been attempted.
4. The new SIT-R norms have aligned the SIT-R scores with SB-R
scores
.
42
5. There is evidence that the correlation between the WISC-R
and the SB-R is approximately the same as that reported for
the Wise and the SB.
6. To date, no extensive research has been conducted concern-
ing the WISC-R scores with the SIT scores. The results
from the limited research that has been conducted show
major departures from the findings between the WISC and
the SIT, similar to those between the SB-R and the SIT.
7. There is a need to determine if the relationship between
the WISC-R and the SIT-R is comparable to the previous
relationship of the WISC to the SIT.
FOOTNOTES
Slosson, R. , "Slosson Intelligence Test" (East Aurora, NY:Slosson Educational Publications) , 1961.
^Ibid.
3Armstrong, R. , Jensen, J. , and Reynolds, J. , "A Comparison of
Individual Intelligence Tests Used for Special Class Placement,"Slosson Intelligence Test Manual , 1974.
4Armstrong, R. and Jensen J. , "Can Scores Obtained from the
Slosson Intelligence Test be Used with as Much Confidence as ScoresObtained from the Stanford-Binet Intelligence Scale?" (East Aurora,NY: Slosson Educational Publications), 1970.
5Armstrong, R. , Mooney, R. , and Jensen, J. , "Testing Low IQ
Children," Slosson Intelligence Manual (East Aurora, NY; Slosson
Educational Publications) , 1971.
^Ibid.
^Chimera, Joseph, "Findings Based on Literature Research
Studies" (unpublished) , 1980.
^Armstrong, R. and Jensen, J. , "The Slosson Intelligence Test:
1981 Norms Tables: Application and Development" (Chelmsford, MA:
Metrics Press, Inc.), 1981; Sattler, J., Assessment of Children's
Intelligence (W.B. Saunders Company) , 1979.
9Ibid.
^^Armstrong, R. and Jensen, J., 1970.
^^Armstrong, R. , Jensen, J. , and Reynolds, J., 1974.
^^Chimera, Joseph, "Findings Based on Literatxire Research
Studies .
"
^^Houston, C. and Otto, W. , "Poor Reader's Functioning on the
Wechsler Intelligence Scale for Children, Slosson Intelligence Test
and Quick Test," Journal of Educational Research 62 (1968) :157-159.
43
44
Jongeward, P.A., "A Validity Study of the Slosson IntelligenceTest for Use with Educable Mentally Retarded Students" (Doctoraldissertation, University of Oregon, 1969). Dissertation AbstractsInternational, 30, 3323A, University Microfilm No. 70-25, 12.
15 .
Meissler, G.R. , "A Correlation of the Slosson IntelligenceTest and the Wechsler Intelligence Scale when Administered toAtypical Children" (Doctoral dissertation. Catholic University ofAmerica) , 1970.
16Pate, R.H. , Jr., and Nichols, W.R. , "The Slosson Intelligence
Test Used as a Screening Device: A Validity Study" (American Psycho-logical Association, Catalog of Abstracts , Experimental PublicationSystem) , 1970.
17Swanson, M.S. and Jabodson, A., "Evaluation of the Slosson
Intelligence Test for Screening Children with Learning Disabilities,"
Journal of Learning Disabilities 3 (1970) : 318-320.
18Lessler, K. and Galinsky, M.D. , "Relationship Between the
Slosson Intelligence Test and Wechsler Intelligence Scale for Children
Scores in Special Education Candidates ," Psychology in the Schools 8
(1971) :341-344.
19Maxwell, M.T., "The Relationship Between the Wechsler Intelli-
gence Scale for Children and the Slosson Intelligence Scale,"
Child Study Journal 1 (1971) :164-171.
20Stewart, K.D., Wood, D.Z., and Gallman, W.A. ,
"Concurrent
Validity of the Slosson Intelligence Test," Journal of Clinical
Psychology 27 (1971) :218-220.
^^Jerrolds, R.W. , Callaway, R. , and Gwaltney , W.K. ,"Comparison
of the Slosson Intelligence Test and WISC Scores of Subjects Referred
to a Reading Clinic," Psychology in the Schools 9 (1972) :409-410.
^^Machen, L.H., "A Validity and Reliability Study of the Slosson
Intelligence Test with an Atypical Population: Gifted Children"
(Doctoral dissertation. Catholic University) ,1972.
^^Lamp, R.E. and Traxler, A.J., "The Validity of the Slosson
Intelligence Test for Use with Disadvantaged Head Start and First
Grade School Children," Journal of Community Psychology 1 (1973):
27-30.
45
Pikulski, J,J., "Validity of Three Brief Measures of Intelli-gence for Disabled Readers," Journal of Educational Research 67(1973) :67-68.
25 .
Sinco, H.J., "A Concurrent Validity Study of the SlossonIntelligence Test and the Wechsler Intelligence Scale for Children"(Master's thesis, Marywood College, Scranton, PA), 1973.
26Armstrong, R. , Jensen, J. , and Reynolds, J., "A Comparison
of Individual Intelligence Tests Used for Special Class Placement,"Slosson Intelligence Test Manual (East Aurora, NY: Slosson EducationalPublications) , 1974.
27Stewart, K.D. and Myers, D.G., "Long Term Validity of the
Slosson Intelligence Test," Journal of Clinical Psychology 30
(1974) ;180-181.
28Covin, T.M. , "Comparison of SIT and WISC-R IQ's Among Special
Educational Candidates," Psychology in the Schools 14 (1977) :20-23.
29Martin, John D. and Kidwell, Jane, C. ,
"Intercorrelations of
the Wechsler Intelligence Scale for Children-Revised, the Slosson
Intelligence Test, and the National Educational Developmental Test,"
Educational and Psychological Measurement (Winter 1977)
.
^^Wechsler, D. , Wechsler Intelligence Scale for Children-Revised
(Psychological Corporation) , 1974.
^^Armstrong, R. and Jensen, J. , "The Slosson Intelligence Test:
1981 Norms Tables: Application and Development" (Chelmsford, MA:
Metrics Press, Inc.), 1981.
CHAPTER III
PROCEDURES
Restatement of the Purpose
At this point, a restatement of the purpose of the study is in
order before the study procedures are detailed. The overall purpose of
this study was to determine if the new 1981 SIT scores are comparable
to 1974 WISC-R scores (Verbal and Full Scale) . In other words, can
the new SIT scores be used with as much confidence as WISC-R scores
—
to the extent that the old SIT (1961) scores could be used with the
unrevised WISC scores.
The intent of the study was to provide valuable information to
WISC-R administrators, especially Directors of Special Education, con-
cerning the use of the 1961 and new 1981 SIT. In other words, if the
results were favorable. Directors of Special Education would be able
to
;
1. administer an individual test of intelligence, rather than
a group test, for screening or retesting even when an
individual test is not required;
2. use the SIT instead of the WISC-R for screening and re-
testing when an individual test is required;
3. increase their test administration staff, since the SIT
does not require specialized training.
46
47
Design
The study was designed to assess the validity of the new 1981 SIT
in relation to the WISC-R (Verbal and Full Scale) , where the latter
instrument served as the criterion. As has been stated previously,
abundant evidence exists which shows the 1981 SIT to have had high con-
current and construct validity as regards the 1960 SB, Similar evidence
now exists for the revised 1972 SB.^ This evidence was obtained re-
cently in response to the renorming of the SB in 1972.^ Since the WISC
was revised, little fresh evidence of the validity of the 1961 SIT in
relation to the WISC-R has been produced. That which has been pro-
duced was based upon rather small samples, and showed a substantial
decrease in the relationship that the 1961 SIT had with the 1949 WISC.
Hence, the need for a thorough validity study of the new 1981 SIT in
relation to the WISC-R.
Validity may be viewed in several ways depending upon the purpose
for which the test in question is to be used. Where the instrument is
to be validated for use in place of the criterion test, its concurrent
validity is of importance. But since in the present case both instru-
ments also purport to measure the same, or quite similar constructs,
construct validity is also of importance. The principal statistic
ordinarily used to establish the concurrent or construct validity of a
measuring instrument is Pearson's Product Moment correlation coefficient.
The type of validity of principal interest in this study is concurrent,
since the practical outcome of greatest interest was to determine
whether the new 1981 SIT can be used in place of the WISC-R for initial
48
screening or for retesting of persons being evaluated as potential or
continuing recipients of special education services. To whatever ex-
tent the concurrent validity of the new SIT in relation to the WISC-P
can be established, its construct validity will also be established
since both approaches to validation require high degrees of correla-
tion. Thus, Pearson's r calculated for pairs of scores on the two
instruments across a large sample in a variety of settings will give
evidence concerning both types of validity. As a result, the study
relied heavily on correlation techniques.
The methodology used borrows from that used previously in the
Armstrong, Jensen, Mooney and Reynolds studies of .the validity of the
1961 SIT in relation to the 1962 SB and in relation to the xanrevised
WISC.^ These studies used the Pearson r as the statistic of principal
interest. They also employed a simple, but straightforward and easy
to understand statistic—the mean absolute IQ difference (MAD) . This
latter statistic shows what an examiner may expect in terms of the
deviation in scores yielded by both instruments, that is, by how many
points on the average an examinee's score will differ having been given
both instruments. Therefore, the mean absolute IQ difference (MAD)
provides information which appeals to the common sense of the examiner
in the field. Thus, as a technique of proven value, the mean absolute
IQ difference (MAD) was also used in this study.
The study also emulates the previously cited studies of the 1961
SIT in relation to the 1960 SB in that it brealcs down the overall sample
to study the relative validity of the new SIT in relation to the WISC-
R for different demographic groups and for varying administrative
49
arrangements under which the two instruments were given . Substudies
were carried out by examinee sex and age level, by exciminer expertise,
by order of administration of the instruments, by time interval between
the two test administrations, etc. The following is a breakdown of
the sxibstudies conducted;
1. Overall - for all examinees;
2.-4. Age Levels - 6-8, 9-11, 12-15;
5. Female - all female examinees;
6. Male - all male examinees;
7. Same Administrator - for both tests;
8. Different Administrator - for each test;
9. Certified Both WISC-R - both test administrators were WISC-R
certified, but not necessarily the same administrator;
10. Certified WISC-R Only - certified administrator for the WISC-R
and a non-certified WISC-R administrator for the SIT;
11. Order WISC-R First;
12. Order SIT First;
13. Same Day - when both tests were administered on the same day;
14.-16. Different Day - when the administration of the second test
was completed within the following time intervals: 1-14 days,
15-60 days, 2 months to 18 months.
An additional aspect of the proposed study did not borrow from pre
vious research. The 1961 SIT was developed to be a short but accurate
estimate of the corresponding 1960 SB IQ, not necessarily of the corres
ponding WISC (unrevised) IQ. Research previously cited has shown that
the lanrevised SIT (1961) was an excellent estimate of the unrevised SB
50
(1960) , and more recently that the revised SIT is an excellent estimate
of the revised SB (1972) . Indeed, the revised SIT has now been ex-
pl^citly equated to the 1972 SB, its old ratio IQ having been abandoned.
Also, the unrevised SIT (1961) was a good, if not excellent estimate of
the unrevised WISC. However, the WISC-R (1974) represents a much more
extensive revision of the instrument itself than was true for the re-
normed SB (1972) which remains virtually unchanged, it only having been
renormed. Therefore, it was unclear whether the old SIT IQ or the new
SIT IQ would provide the better estimate of the corresponding WISC-R
IQ. To be sure, it was expected that the new SIT IQ's would do a bet-
ter job of estimating WISC-R IQ's. Thus, an additional aspect of the
study was directed at obtaining an explicit answer to this question.
Hypotheses to be Tested
As a validity study, it was e^qaected that the observed correlations
between the revised SIT and the WISC-R would be sizeable and certainly
significantly different from zero. Therefore, to state hypotheses con-
cerning whether observed correlations will depart significantly from
zero is probably trivial. Nevertheless, they are given in the interests
of completeness.
Part A
Al.Ho: The correlations observed across the entire sample and
for each of its several subgroups between IQ's obtained
from the revised SIT and the WISC-R will not differ
significantly from zero.
51
A2.nai The correlations obsorvod across the entire sairijilo and
for each of its sovoral subgroups between IQ's obtained
from the revised SIT and the WISC-R will bo positive
and significantly different from zero.
A second sot of hypotheses wore offered to test whether the re-
vised SIT IQ's (1981) are a better estimate of the WISC-R than wore
the unrevised SIT IQ's (1961).
Part D
Bl.Ho: The correlations obsorvod across the entire sample be-
tween IQ's obtained from the revised SIT and the WISC-R
(V & FS) will not differ significantly from corresponding
correlations relating the unrevisod SIT IQ's (1961) and
WISC-R IQ's.
B2.Ha: The correlations observed across the entire sample be-
tween IQ's obtained from the revised SIT and the WISC-R
will bo significantly larger than correlations relating
the unrevisod SIT IQ's (1961) and WISC-R IQ's.
Sample
Given that the proposed study will attemi’t to provide stable es-
timates of tho correlation between the SIT and the WISC-R for several
demographic groups and administrative arrangements, the sample upon
which it is based must bo quite large and representative of the popula
tion to which generalizations will be made. The actual Bami>lo size
included 336 students ranging in ages 6-15. For the most part, those
52
students were administered the WISC-R as part of a core evaluation,
with many of them being actual recipients of special education services.
They were drawn from several commimities in Massachusetts to ensure
that all relevant demographic characteristics were adequately repre-
sented.
The sample could not be drawn at random from the country as a
whole or even from New England. Nevertheless, the participating com-
munities were selected to provide a balanced variety of urban and
suburban settings, and as heterogeneous a sample as possible.
Data Collection
Each child who was a participant in the study was administered a
SIT and a WISC-R. Additional data concerning the conditions of admin-
istration for both testing sessions was obtained as well to provide
bases upon which to classify svibjects into various subgroups for
analysis
.
Consistent with the requirements of confidentiality, demographic
information concerning the subject's sex, age, grade level, community
setting, etc. , were gathered as well. These data were obtained via a
data recording instrument completed by the person (s) examining each
subject. A copy of the data collection instrument and cover letter is
included in Appendix B.
53
Statistical Analyses
Data for each subject were keyed to a machine readable medium.
The Statistical Package for the Social Sciences (SPSS) was used for
ij^itial file construction and for all subseguent analyses.
Following initial file construction, N's, means, standard devia-
tions and frequency distributions for each variable were obtained
using subprogram FREQUENCIES utilizing the entire data set. This was
done so that errant data could be identified and either corrected or
discarded as unusable.
When data cleaning was accomplished, the proposed analyses were
carried out. Hypotheses A1 through B2 required that the data set be
broken down into subgroups for separate analyses. The s\abgroups were
formed and the analyses carried out using subprograuns BREAKDOV'/N and
PEARSON CORR.
Tests of statistical significance between correlation coefficients
were performed by hand calculation based upon the descriptive statistics
output in prior stages of analysis since SPSS does not directly provide
significance tests for differences observed between correlation coeffic-
ients. The significance of difference from zero for all correlations
produced were tested using subprogram PEARSON CORR.
Hypotheses stated in Part B required that differences in correla-
tion coefficients be tested for statistical significance. The signifi-
cance tests implied in hypotheses Bl were performed as recommended by
Guilford.
54
The mean absolute IQ difference (MAD) is a statistic which has
been shown to be of use in similar validity studies. In this con-
text, it is simply the average of the absolute differences between IQ's
generated for all persons tested.
Examiners in the field need to know what the "margin for error"
is if one instrximent is to be used in place of another. The MAD sta-
tistic provides an examiner the needed information. For example, if
MAD for a given subgroup equals 5 IQ points , then an examiner knows
that the average amount by which IQ's gotten from either instriament
will differ in a like sample of individuals is 5 IQ points. Some will
differ more, some less, but the average "margin for error" will be 5
points for persons in the subgroup.
The standard error of measurement (SEM) indicates much of the same
thing as the MAD statistic, though it often appears to be more esoteric.
In fact, it has been shown that the mean absolute difference between
all possible pairs of a person's scores on an infinite number of par-
allel forms of the same test is equal to roughly 1.1 times the standard
error of measurement. Though this relationship holds for parallel
forms of an instrument, the SIT and the WISC-R may be parallel, but as
entities they are probably not parallel. Thus, MAD and SEM in the con-
text of the proposed study should not be regarded as directly or simply
related.
However, the MAD statistic does speak directly to the concerns of
practitioners in the field. For this reason it was reported along with
the other statistics previously described. The relative size of MAD
for different subgroups of the data set provides an easily understood
55
"benchmark" for interpretation of anticipated differences between
measxores from the two instruments in question.
Summary
Chapter III described the overall procedural steps of this study,
including; design, sample, hypotheses to be tested, data collection,
and statistical analyses. The study results are reported in Chapter
IV.
FOOTNOTES
Armstrong, R. and Jensen, J. , "The Slosson Intelligence Test1981 Norms Tables: Application and Development" (Chelmsford, MA:Metrics Press, Inc.), 1981.
^Ibid.
3Armstrong, R. and Jensen, J. , "Can Scores Obtained from the
Slosson Intelligence Test be Used with as Much Confidence as ScoresObtained from the Stanford-Binet Intelligence Scale," Slosson Intelli-gence Manual (East Aurora, NY: Slosson Educational Publications),1970; Armstrong, R. and Jensen, J. , "Testing Low IQ Children," SlossonIntelligence Manual (East Aurora, NY: Slosson Educational Publications),1971; Armstrong, R. and Jensen, J. , "An Abbreviated Stanford-Binet:Is It Valid," Journal of Educational Psychology and Measurement(June 1972); Armstrong, R. , Jensen, J., and Reynolds, J. , "A Comparisonof Individual Intelligence Tests Used for Special Class Placement,"
Slosson Intelligence Test Manual (East Aurora, NY: Slosson Educa-
tional Publications) , 1974; Armstrong, R. and Mooney, R. , "The Slosson
Intelligence Test: Implications for Reading Specialists," The Reading
Teacher (January 1971); Armstrong, R. and Mooney, R. , "A Short,
Reliable, Easy to Administer Individual Intelligence Test for Special
Class Placement," Child Study Journal (Spring 1971); Armstrong, R. and
Mooney R. , "The Use of Individual Intelligence Tests for Special Class
Placement," Slosson Intelligence Test Manual (East Aurora, NY: Slosson
Educational Publications) , 1974.
"^Nie, Hull, Jenkins, Steinbrenner , Berel, Statistical Package
for the Social Sciences, 2nd edition (McGraw-Hill), 1976.
^Guilford, J.P., Fundamental Statistics in Psychology and
Education , 4th edition (McGraw-Hill) ,1965.
^Armstrong, R. , Jensen, J. , and Reynolds, J., "A Comparison of
Individual Intelligence Tests Used for Special Class Placement,"
Slosson Intelligence Test Manual (East Aurora, NY: Slosson Educational
Publications) ,1974.
56
CHAPTER IV
RESULTS OF THE STUDY
The pxorpose of this chapter is to report the findings of this
study, as prescribed in Chapter III. The format of the chapter will be
to restate each of the hypotheses presented in Chapter III, followed by
the study findings.
Since many of the categories in the tables are coded, a decription
of each is in order.
1. Overall - for all examinees
;
2.-4. Age Levels - 6-8, 9-11, 12-15;
5. Female - all female examinees;
6
.
Male - all male examinees;
7. Same Administrator - for both tests;
8. Different Administrator - for each test;
9. Certified Both WISC-R - both test administrators were WISC-R
certified, but not necessarily the same administrator;
10. Certified WISC-R Only - certified administrator for the WISC-R
and a non-certified WISC-R administrator for the SIT:
11. Order WISC-R First;
12. Order SIT First;
13. Same Day - when both tests were administered on the same day;
14.-16. Different Day - when the administration of the second test
was completed within the following time intervals: 1-14 days,
15-60 days, 2 months to 18 months.
57
58
The first set of hypotheses addresses the question, are the corre-
1 2lations between the new/revised SIT and WISC-R IQ's positive and
significantly different from zero?
Part A
Al.Ho; The correlations observed across the entire sample and for
each of its several subgroups between IQ's obtained from
the revised SIT (1981) and the WISC-R (Verbal and Full
Scale) will not differ significantly from zero.
A2.Ha; The correlations observed across the entire sample and
for each of its several subgroups between IQ's obtained
from the revised SIT (1981) and the WISC-R (Verbal and
Full Scale) will be positive and significantly different
from zero.
Table 12 reports means, standard deviations, correlations and mean
absolute IQ differences (MAD) for the new/revised (1981) SIT and the
WISC-R tests. The design, as stated in Chapter III, delimited the study
to a comparison of the SIT and the WISC-R Verbal subtest and Full Scale
scores. However, the WISC-R Performance subtest was included in the
interest of completeness.
Using the Pearson-Product Moment method, a coefficient of correla-
tion was computed separately between the new/revised SIT and the WISC-R
(Verbal, Performance, Full Scale) IQ's for sixteen subgroups of the
data. Since it was expected that the correlations would be positive and
significantly different from zero, a one tailed test of significance
Table 12 reveals that all correlations between the SITwas utilized.
MEANS,
STANDARD
DEVIATIONS,
CORRELATIONS
AND
MEAN
ABSOLUTE
IQ
DIFFERENCES
(MAD)
OF
THE
NEW
SIT
AND
WISC-R
TESTS
59
*u
I
uwHS
EhH
8T* .27 .32 .32CD00 ,76 47 57
CO16 16 13 17, 16. 15. 16. 14.
C4 o cn 00 <p1 CD 00 cn oc cnU •
CO o r—
1
iH ro iH cnM o o o O o os nH iH rH rH iH rH
EhHCO
.02 .62 .54 60* .63 00 o .22 .20 .39 .38 .06
ino ,99 49
00 03 50 11 01 56 45 64 18
CD11 r' r'
12 00 CD11 m 11, r' cd’ 10. in 12. r' CD* 12. CD 11. CO
SSS£SgSS?S!3g§§gSS§gC;SSSS
00 (P roCN O CD
• • • .
o CD 00 00o O (P CP1—
1
rH
00
0^CT'
CP CP OCN ro rH
rH CN 'ej*
o O CPrH rH
UCO
0)
to
oS CO
>co co CO CO CO CO COfa>CL,CL,>(l,Pn>(i,Cn>(i<P4>(iifa>pjPL,> CO
a> h
CD 00 CD o CD CD (P 00ro 00 O ro 00 CNro iH rH rH rH CN rH
oCP<upto
u
00rHrH
-15
rH 1 1 CNrH CD CP rHtcj
P in in cn
Cl) 0) 0) cu
> tp Cp CPO < < rt;
uo4Jto
pp +j
0 014J •Hto cp •HV ein
*H <C'H +j
e c-o 0)
0) < prH 0)
to 0) CD >4H
e rH e 14H
0) to (0 •iH
s in QV
He
.001
(all
correlation
coefficients
are
significant
beyoncJ
the
.001
level).
Means
Standard
Deviations
60
*
.26
00.19 9T* .62 .85 .57 .62 .55 .62 .22 .22 ,37
CM16 96 54 80 65 89 33 CM
00 03 30
in11 r' CD
11 VD11 in 11, r' in 11. CO CD 12. 00 D 11. r' in 11. r'
.90 .57 .83 .88 .58 .82
0000 .55 .81 06* .65 .85 .91 .61 .85 .83 .43 ,75 ,87 59 D 54 79
I
UCOHs
'^v^cocNvOvoounomror^cNCr'CDiHvO^rHCOiHTj'rHOOCOCyi
'^ftninvoinvOLnininininminrHrHr-liHiHr-IrHrHrHiHiHi—li—
I
KO Oin <ji
(N rH iH ro o CO 00CTi 00 CM rH C^ rH
minoo^^voinm^00 iHCM o
EhHCO in
VO CM rH G\VD CD 0 CM ro 0vO in in VOrH rH rH iH rH rH rH
oiI
uCOHs
o;®OrHrH^inocNcriiH'<a'i£>cMvDoorHrocricy>r^[^'5i<v£)
n3
Eh
0 CM <y\ 0Eh CO in 00 00H • • • •
CO 00 iH 00 Oc m<Ti 0 OC 0 0
rH rH rH
CD m r'CD CM oc
n rn m0 0 ocrH rH
KI
uCO
0 rH 00 0 'd' in 00rH 0 in CD c^ rHrH CM CM rH
cn
x: cn
P >1 cn
>1 C nj rH(0 0 Q (0
Q s Q000 CD 'a'
6 rH 1 rHC4 rH (0 1 in 1
1 rH CO CM rHU aCO 0 4J c c c c
>1 H m 0 0 0 0!h s Pi •rl •rl •rl •H0 1 •iH 4J P P) P PCn D! u b cn nj (0 fO (0
0 4J CO iH P P p p+J 0 H Pi •rl +J -P p 4JrO m s 1 Ph cn cn cn cn
U u H •H •H •HTJ CO Eh c c C c
0 0) H H •P •rl
•H •r) S CO e EMH Mh rQ '3 'U•H H u U < < < <4J 4J 0) d)
73 TJ 'C3
0) (U C c c cU u 0 0 CM CN CM CN
V*
.001
(all
correlation
coefficients
are
significant
beyond
the
.001
level).
61
and WISC-R IQ's were significant beyond the .001 probability level. As
a result, the null hypothesis that the correlations observed across the
entire sample and for each of its several subgroups between IQ's ob-
tained from the new/revised SIT (1981) and the WISC-R would not differ
significantly from zero was rejected. Given this, the alternative
hypotheses that the correlations observed across the entire sample and
for each of its several sx±)groups between IQ's obtained from the new/
revised SIT (1981) and the WISC-R would be positive and significantly
different from zero was accepted.
Table 13 summarizes the correlations reported in Table 12 in a
more manageable way. The overall correlation between the SIT and Verbal
subtest was .89. The range of SIT/Verbal correlations was from .83
(2 months to 18 months) to .93 (ages 12-15) , with five correlations of
.90 and above. Although these correlations were impressive, they were
not xanexpected since the SIT is a highly verbal test.
The overall correlation between the SIT and the Performance sub-
test was .58, with the range of correlations from .48 (2 months to 18
months) to .68 (female) . Although these correlations were substantially
lower than those for the Verbal siibtest, they were not unexpected since
verbal tests usually do not correlate as highly with performance tests
as do verbal tests with verbal tests. It should also be noted that
these correlations were not part of the original design. They were in-
cluded since the WISC-R has a performance component.
The overall correlation between the SIT and the WISC-R Full Scale
IQ's was .83. The range of correlations was from .75 (2 months to 18
months) to .88 (ages 12-15), with 12 of the 16 correlations above .80.
62
TABLE 13
CORRELATIONS OF THE NEW SIT WITH THE WISC-R TESTS
Category WVWISC-R*
WP WFS
Overall .89 .58 .83Ages 6-8 .87 .63 .82Ages 9-12 .85 .53 .79Ages 12-15 .93 .63 .88Female .90 .68 .86Male .89 .51 .80Same Administrator .89 .57 .83Different Administrator .85 .54 .78Certified Both WISC-R .90 .57 .83Certified WISC-R Only .88 .58 .82Order WISC-R First .88 .53 .81Order SIT First .90 .65 .852nd Administration Same Day .91 .61 .852nd Administration 1-14 Days .87 .59 .84
2nd Administration 15-60 Days .86 .54 .792nd Administration 2-18 Months .83 .48 .75
*.001 (all correlations are significant beyond the .001 level).
These correlations are particularly iinpressive considering that the
WISC-R ^ull Scale score is the combination of an individual's Verbal
and Performance scores. In other words, a certain amount of deteriora-
tion from the SIT/WISC-R Verbal correlations was expected when compared
with the correlation between SIT and the WISC-R Full Scale scores.
Table 14 contains the results which pertain to the second set of
hypotheses. Are the new/revised SIT IQ's (1981) a better estimate of
3
the WISC-R IQ's than were the old/unrevised SIT IQ's (1961).
63
TABLE 14
IMPROVEMENT OF THE CORRELATION OF THESIT IQ'S WITH WISC-R IQ’S
N = 336
WISC-R
SIT WV WP WFS
Old/Unrevised SITRatio IQ
.875 .569 .815
New/Revised SITDeviation IQ
.890 .577 .825
< .01 NS <.05
Part B
Bl.Ho: The correlations observed across the entire sairple be-
tween IQ's obtained from the revised SIT (1981) and
the WISC-R (Verbal and Full Scale) will not differ
significantly from corresponding correlations relating
the unrevised SIT IQ's (1961) and WISC-R IQ's (Verbal
and Full Scale)
.
B2.Ha: The correlations observed across the entire sample be-
tween IQ's obtained from the revised SIT (1981) and
the WISC-R (Verbal and Full Scale) will be significantly
larger than correlations relating the unrevised SIT IQ s
(1961) and WISC-R IQ's (Verbal and Full Scale).
64
Although the original design was delimited to the WISC-R Verbal
and Full Scale, the Performance subtest was included in the interest of
completeness. In order to test the null hypothesis, SIT ratio IQ's,
SIT deviation IQ's and WISC-R IQ's were determined for all members of
the sample. Then Pearson correlations were calculated for both SIT
IQ's in relation to WISC-R IQ's. A two tailed test of significance
was utilized.
Taible 14 reveals that the correlation between the new/revised SIT
IQ's and WISC-R IQ's increased in all three areas, and that the differ-
ences were significant for the Verbal subtest (<.01) and the overall
test or Full Scale score («c-. 05) . Although the correlation between the
new SIT and WISC-R Performance subtest increased, the difference was
not significant. Thus, the null hypotheses that the correlations
across the entire sample between IQ's obtained from the revised SIT
(1981) and the WISC-R (Verbal and Full Scale) would not differ signifi-
cantly from corresponding correlations relating the unrevised SIT IQ's
(1961) and the WISC-R IQ's (Verbal and Full Scale) was rejected. In-
stead, the alternative hypotheses that the correlations observed across
the entire sample between IQ's obtained from the revised SIT (1981)
and the WISC-R (Verbal and Full Scale) would be significantly larger
than correlations relating the unrevised SIT IQ's (1961) and WISC-R
IQ's (Verbal and Full Scale) was accepted.
The mean absolute IQ difference (MAD) is a statistic which has been
shown to be of use in similar validity studies. In this context, it is
siitply the average of the absolute differences between IQ's generated for
all persons tested. For example, if MAD for a given subgroup equals 7 IQ
65
points, then an examiner knows that the average amount by which IQ's
gotten from either instrument will differ in a like sample of individu-
als is 7 IQ points. Some will differ more, some less, but the average
"margin of error" will be 7 points for persons in the subgroup.
Table 12 contains the MAD statistics for each of the groups in-
volved in the study. Table 15 summarizes the MAD statistics in a more
manageable way. The overall MAD statistics were 6.02 (Verbal) , 11.62
(Performance) and 7.54 (Full Scale). The results are similar to those
reported for correlations in Table 13. That is, the best results were
for the Verbal, the worst for Performance, with Full Scale showing a
small degree of deterioration from the Verbal.
The MAD ranges were: Verbal - 5.37 (2nd Administration, Same Day)
to 7.09 (Ages 6-8); Performance - 10.99 (Female) to 12.63 (Ages 6-8);
Full Scale - 7.06 (Ages 12-15) to 8.84 (Ages 6-8)
.
Examination of
Table 12 shows very little fluctuation in any of the subgroups from the
overall MAD statistics.
Table 16 reveals an additional comparison of the SIT with the
WISC-R. Did the now SIT norms improve upon the mean absolute IQ differ-
ences (MAD) between the SIT and the WISC-R? In order to determine this,
SIT ratio, ^ SIT deviation, and WISC-R IQ's were found for all members of
the sample. Then, the mean absolute IQ differences (MAD) were calcu-
lated for both SIT IQ's with WISC-R IQ's.
66
TABLE 15
MEAN ABSOLUTE DIFFERENCESNEW SIT WITH THE WISC-
(MAD) OFR TESTS
THE
MAD
Category WV WP WFS
Overall 6.02 11.62 7.54Ages 6-8 7.09 12.63 8.84Ages 9-12 6.07 11.22 7.20Ages 12-15 5.39 11.38 7.06Female 6.05 10.99 7.49Male 5.87 12.03 7.50Same Administrator 6.11 12.01 7.56Different Administrator 6.45 11.64 8.18Certified Both WISC-R 5.26 11.78 7.19Certified WISC-R Only 6.16 11.62 7.85Order WISC-R First 6.57 11.62 7.55Order SIT First 5.62 11.22 7.222nd Administration Same Day 5.37 11.12 6.912nd Administration 1-14 Days 5.82 11.03 7.302nd Administration 15-60 Days 6.65 11.89 7.332nd Administration 2-18 Months 6.96 12.54 8.80
TABLE 16
IMPROVEMENT UPON THEBETWEEN
MEAN ABSOLUTE IQ DIFFERENCES (MAD)
THE SIT AND WISC-RN = 336
WISC-R
SIT WV WP WFS
Old/Unrevised SIT
Ratio IQ
7.05 12.15 8.29
New/Revised SIT
Deviation of IQ
6.02 11.62 7.54
67
It can be seen in Table 16 that the mean eibsolute IQ differences
(MAD) between the new SIT and the WISC-R were lower in all three areas.
The MAD decreases or improvements were 1.03 (WV) , .52 (WP) , and .75
(WFS) . In other words, the new SIT norms did improve upon the mean
absolute differences (MAD) between the SIT and the WISC-R.
FOOTNOTES
Armstrong, R.J. and Jensen, J.A., "The Slosson Intelligence Tost1981 Norms Tables; Application and Development" (Chelmsford, MAi MetricsPress, Inc. )
,
1981.
2Wechsler, D. , Manual for the Wechsler Intelligence Scale for
Children - Revised (New York; Psychological Corporation), 1974.
^Slosson, R.L., Slosson Intelligence Test for Children and
Adults (East Aurora, NY; Slosson Educational Publications), 1961.
4Armstrong, R.J. and Jensen, J.A., "The Slosson Intelligence Test
1981 Norms Tables; Application and Development" (Chelmsford, MA;
Metrics Press, Inc.), 1981; Slosson, R.L., Slosson Intelligence Test
for Children and Adults (East Aurora, NY; Slosson Educational Publica-
tions) , 1961; Slosson, R.L., Slosson Intelligence Test for Children
and Adults , 2nd edition (East Aurora, NY; Slosson Educational Publica-
tions) , 1981.
68
CHAPTER V
SUMMARY AND RECOMMENDATIONSFOR FUTURE RESEARCH
Summary
Results from a recently ac3ministered test of intelligence are fre-
quently used as one of the criteria for initial assessment of children
with special needs, as well as for periodic re-assessment to determine
the appropriateness of the program for each child. The psychological
assessment component usxially includes the results of an individually
administered intelligence test such as the Wechsler Intelligence Test
for Children - Revised (WISC-R) or the Stanford-Binet Intelligence Scale
(SB) , Form L-M .
Unfortunately, individual intelligence tests such as the WISC-R
and SB require specialized training in their administration, scoring
and interpretation—usually amounting to a full semester course for each
test. They are also time consuming in that they require an average ad-
ministration time of one hour. Thus, there is a need for a valid indi-
vidual test of mental ability which can be quickly administered and
easily scored. Such a test could be used as a screening and/or retest-
ing instr\iment.
In 1961, Richard Slosson published the Slosson Intelligence Test
(SIT). The author's purpose was to develop a "mini" Stanford-Binet,
that is, an individual intelligence test which: (1) did not require
specialized training of its administrators; (2) took only 15 to 20
69
70
minutes to administer and score; and (3) would yield IQ scores compar-
able to the SB.
Just when it appeared there was sufficient evidence that SIT IQ
scores could be used with as much confidence as SB, W-v and W-FS IQ
scores, both the WISC (1949) and SB (1960) were revised. The SB was
revised in 1972 and the WISC (now WISC-R) in 1974.
The following is a summary of the research findings concerning the
SIT (1961), SIT (1981), SB (1960), SB (1972), WISC and WISC-R.
1. There is a need for valid individual intelligence tests that:
(1) requires no specialized training of its administrators;
(2) can be quickly administered and scored; and (3) yields
IQ scores comparable to the Stanford-Binet and Wechsler
tests. Such a test would provide testing specialists with
an opportunity to administer more individual tests them-
selves, as well as provide an invaluable source of addition-
al test administrators.
2. In the early 1970' s, there was sufficient evidence that 1961
SIT IQ scores could be used with as much confidence as 1960
SB and WISC (V and FS) scores. Unfortunately, the WISC and
SB were revised. During the late 1970' s, there was sufficient
evidence that 1961 SIT scores were no longer comparable to
the new WISC-R or the new SB (1972) scores.
3. The new SIT norms (1981) once again aligned SIT scores with
the 1972 SB scores.
71
4. There is evidence that the correlation between the WISC-R
and the 1972 SB is approximately the same as that reported
for the Wise and the 1960 SB.
5. To date, no extensive research has been conducted concerning
the WISC-R scores with the 1981 SIT scores. The results from
the limited research that has been conducted show major de-
partures from the findings between the WISC and the 1961
SIT, similar findings between the WISC and the 1961 SIT,
similar to those between the 1972 SB and the 1961 SIT.
The purpose of this study was to determine if the new SIT (1981)
scores are comparable to WISC-R scores (Verbal and Full Scale) . In
other words, can the new SIT (1981) scores be used with as much confi-
dence as WISC-R scores—to the extent that the old SIT (1961) scores
could be used with the unrevised WISC scores.
The study sample included 336 students ranging in ages 6-15 years
from twenty communities in Massachusetts. The participating communities
were selected to provide a balanced variety of urban and suburban set-
tings, and as heterogeneous a Scimple as possible. For the most part,
these students were administered the WISC-R as part of a team evalua-
tion, with many of them being actual recipients of special education
services.
Substudies were carried out by examinee sex and age level, by ex-
aminer expertise, by order of administration of the instruments, by
time interval between the two test administrations, etc. The following
is a breakdown of the siibstudies conducted.
72
1. Overall - for all examinees;
2.-4. Age Levels - 6-8, 9-11. 12-15;
5. Female - all female examinees;
6. Male - all male examinees;
7. Same Administrator - for both tests;
8. Different Administrator - for each test;
9. Certified Both WISC-R - both test administrators were WISC-R
certified, but not necessarily the same administrator;
10. Certified WISC-R Only - certified administrator for the
WISC-R and a non-certified WISC-R administrator for the
SIT;
11. Order WISC-R First;
12. Order SIT First;
13. Same Day - When both tests were administered on the same day;
14.-16. Different Day - when the administration of the second test
was completed within the following time intervals: 1-14
days, 15-60 days, 2 months to 18 months.
Data analysis addressed the following three questions:
1. Across the entire sample and for each of its several subgroups,
are IQ's obtained from the revised SIT (1981) and the WISC-R
(Verbal and Full Scale) positive and significantly different
from zero?
2. Are the correlations across the entire sample between IQ'sAre the correlations across the entire sample between IQ's
obtained from the revised SIT (1981) and the WISC-R (Verbal
and Full Scale) significantly larger than the correlations
73
relating the unrevised SIT IQ's (1961) and the WISC-R IQ's
(Verbal and Full Scale)
?
3. Did the new SIT norms (1981) improve upon or decrease the
mean absolute difference (MAD) between the SIT and the
WISC-R (SIT 1961 versus SIT 1981)?
The design, as stated in Chapter III, delimited the study to a
comparison of the SIT and the WISC-R Verbal subtest and Full Scale
scores. However, the Performance subtest was included in the interest
of completeness.
The following is a summary of the study findings reported in
Chapter IV
:
1. The correlations across the entire seimple and for each of
the subgroups were positive and significantly different
from zero. The overall correlations between the SIT and
WISC-R Verbal, Performance and Full Scale were .89, .58,
and .83 (Tables 12 and 13).
2. The correlations across the entire sample between IQ's
obtained from the revised SIT (1981) and the WISC-R
(Verbal, Performance and Full Scale) were larger than the
correlations relating the unrevised SIT IQ's (1961) and
the WISC-R IQ's (Verbal, Performance and Full Scale). Also,
the score differences were significant for the Verbal sub-
test and the overall or Full Scale (Table 14)
.
3. The new SIT norms (1981) improved upon or decreased the
mean absolute IQ difference (MAD) between the SIT and the
WISC-R (SIT 1961 versus SIT 1981) (Tables 15 and 16) . The
74
overall MAD statistics were 6.02 (Verbal), 11.62 (Performance)
and 7.54 (Full Scale) . The MAD decreases or improvements
were 1.03 (WV) , .52 (WP) and .75 (WFS)
.
It was also noted in Chapter IV that the difference between the
correlations and MAD statistics for the numerous subgroup comparisons
were minimal. For example, the SIT and WISC-R Verbal correlations for
female and male were .90 and .89, with mean absolute 10 differences
(MAD) of 6.05 and 5.87 respectively. However, there were two important
findings that need to be discussed.
The first area concerns the time between test administrations.
In this study there were four time interval categories; same day, 1-14
days, 15-60 days and 2-18 months. One would expect that same day test-
retest results would be better than test-retest results with a time in-
terval between them, such as 2 weeks, 2 months or 1 year.
The deterioration from Scime day test-retest is usually caused by a
combination of factors, such as: (1) a different environment; (2) a
different administrator; (3) the attitudinal, emotional and physical
characteristics of the examinee (and/or examiners) on two different
occasions; and (4) the efforts of the learning process (etc.) when an
extended period of time occurs between two test administrations, e.g.,
one year.
The results reported below for this study are no different than
would be expected for WISC-R Verbal and Full Scale scores. Perhaps the
most important question of this study is whether the results of the SIT,
in a given situation, are comparable to the WISC-R Verbal and Full
Scale results, had they been administered. The same day findings of
75
this study provide sufficient evidence that SIT scores can be used as
estimates of the WISC~R Verbal and Full Scale scores, had they been
administered.
WV FS
r MZU) r MAD
Same Day .91 5.37 .85 6.91
1-14 Days .87 5.82 .84 7.30
15-60 Days .86 6.65 .75 7.33
2-18 Months .83 6.96 .75 8.80
The second area of siibgroups that needs to be discussed is the
certification status of the test examiners. It can be seen below that
there was very little difference in the results when the SIT was admin-
istered by a non-certified WISC-R examiner or a certified WISC-R exam-
iner. In other words. Directors of Special Education can use the re-
sults of a SIT administration when the SIT is administered by a non-
certified WISC-R examiner with as much confidence as when the SIT is
administered by a certified WISC-R examiner.
WV FS
r MAD r MAD
Both CertifiedWISC-R Administrators .90 5.26 .83 7.19
SIT AdministratorNot WISC-R Certified .88 6.16 .82 7.85
76
In svun, the new SIT (1981) scores are coitparable to WISC-R scores
(Verbal and Full Scale). In other words, the new SIT (1981) scores can
be used with as much confidence as WISC-R scores—as the old SIT (1981)
scores could be used with the unrevised WISC scores. In fact, the
findings show an improvement in the relationship between the two tests
.
Directors of Special Education should feel confident to: (1) use
the new SIT (1981) instead of the WISC-R for screening and retesting,
and (2) use the results of the SIT administered by a non-certified WISC-
R examiner. In many cases, the use of the SIT will result in an in-
crease in the test administration staff, since the SIT does not require
specialized training.
Future Research
The following are examples of needed research concerning the
Slosson Intelligence Test.
1.-3. The current study was delimited to students who were
administered the WISC-R as part of a core evaluation.
Thus, there is a need to determine the relationship
between the SIT and the WISC-R in terms of:
1) Low 10;
2) Gifted;
3) Full Range.
4.-7. What is the relationship between the new SIT and the
Wechsler Pre-School Primary Scale (WPPSI)?
4) Low IQ;
5) Gifted;
77
6) Full Range;
7) Special Needs.
8.-11. What is the relationship between the new SIT and the
Wechsler Adult Intelligence Scale-Revised (WAIS-R)?
8) Low IQ;
9) Gifted;
10) Full Range;
11) Special Needs.
12.-22. Repetition of studies 1-11 but also including the 1972
Stanford-Binet (SB)
.
23.-25. What is the relationship between the new SIT and
various college entrance tests such as:
23) Scholastic Aptitude Test (SAT)
;
24) Preliminary Scholastic Aptitude Test (PSAT)
;
25) American College Test (ACT)
.
26. What is the relationship between the new SIT and various
group tests of intelligence such as the Otis-Lennon
School Ability Test?
27. What is the relationship between the new SIT and the
various standardized achievement tests , such as the
Metropolitan?
28. What is the relationship between the new SIT and the
Differential Aptitude Battery Test (DAT)
?
29. Is the new SIT a better predicter of standardized
achievement test scores (e.g.. Metropolitan) than
group intelligence tests (e.g., Otis-Lennon)?
78
30. What is the relationship between the new SIT and other
individual pre-school/primary/elementary tests of ability
such as the McCarthy Scales, Peabody Picture Vocabulary,
ITPA, etc.?
31. What is the relationship between the new SIT and the
various tests used in special education?
32. Can the new SIT be used (or modified to be used) with
people who have special handicaps (e.g.
,
blind, deaf,
etc. )
?
33. What is the relationship between the new SIT and the
various graduate level aptitude tests such as the GRE
and LSAT?
BIBLIOGRAPHY
Ames, L.B. "Academic Progress in Negro Schools." Journal of LearningDisabilities 1 (1968) : 570-77
.
Armstrong, R. , and Jensen, J. "Can scores obtained from the SlossonIntelligence Test be used with as much confidence as scores ob-tained from the Stanford-Binet Intelligence Scale?" East Aurora,NY: Slosson Educational Publications, 1970.
Armstrong, R. ; Mooney, R. ; and Jensen, J. "Testing Low IQ Children."Slosson Intelligence Manual . Slosson Educational Publication,September, 1971.
Armstrong, R. ; Mooney, R. ; and Jensen, J. "The Slosson IntelligenceTest: Implications for Reading Specialists." The Reading Teacher24 (1971) :336-40.
Armstrong, R. , and Jensen, J. "Validity of an Abbreviated Form of theStanford-Binet Intelligence Scale, Form L-M." Educational andPsychological Measurements 32 (1972) :463-67.
Armstrong, R. ; Jensen, J. ; and Reynolds J. "A Comparison of IndividualIntelligence Tests Used for Special Class Placement." SlossonIntelligence Test Manual , Spring, 1974.
Armstrong, R.; Mooney, R. ; and Jensen, J. "The Use of Individual
Intelligence Tests for Special Class Placement." East Aurora,
NY: Slosson Educational Publications, 1972.
Armstrong, R. , and Jensen J. "The Slosson Intelligence Test: 1981
Norms Tables: Application and Development." Chelmsford, MA:
Metrics Press, Inc., Fall, 1981.
Armstrong, R. ; Mooney, R. ; and Jensen, J. "A Short, Reliable, Easy to
Administer Individual Intelligence Test for Special Class Place-
ment." Child Study Journal 1 (1971) : 156-63
.
Baum, Dale D. , and Kelly, Thomas J. "The Validity of the Slosson
Intelligence Test with Learning Disabled Kindergartners.
"
Journal of Learning Disabilities (April 1979)
.
Bloom, A.S.; Raskin, L.M. ; and Resse, A.H. "A Comparison of the WISC-R
and Stanford-Binet Intelligence Classifications of Developmentally
Disabled Children." Psychology in the Schools 13 (1976) :288-90.
Bonfield, J.R. "Predictors of Achievement for Educable Mentally Re-
tarded Children." Unpublished doctoral dissertation, Pennsylvania
State University, 1968.
79
80
Boyd, J.E. "Use of the Slosson Intelligence Test in Reading Diagnosis."Academic Therapy 9 (1974) :441-44.
Burns, L.M. "The Slosson Intelligence Test as a Pre-screening Instru-ment with Potentially Gifted Children." Unpublished master'sthesis, Sacramento State College, 1968.
Carlisle, A.; Shinedling, M. ; and Weaver, R. "Note on the Use of theSlosson Intelligence Test with Mentally Retarded Residents."Psychological Reports 26 (1970) :865-66.
Carney, F.L., and Karfgin, L. "Using the Slosson Intelligence Testwith a Criminal Population." Personal Communication, 1973.
Charlton, N.W. "An Investigation of Selected Visual-Perceptual andMotor Parameters of Young Trainable Mentally Retarded Children."Unpublished doctoral dissertation. University of Houston, 1970.
Chimera, Joseph. "Findings Based on Literature Research Studies."
Unpublished, 1980.
Covin, T.M. "Comparison of SIT and WISC-R IQ's Among Special Educa-
tion Candidates." Psychology in the Schools 14 (1977) : 20-23
.
Covin, T.M. "Relationship of the SIT and PPVT to the WISC-R."
Journal of School Psychology 9 (1977) ;259-60.
Datta, Lois-Ellen. "Review of Slosson Intelligence Test." In
Pediatric Screening Tests, edited by William K. Frankenburg and
Bonnie Camp. Springfield, IL: Charles C. Thomas, Publisher,
1975.
DeLapa, G. "Correlates of Slosson Intelligence Test, Binet, Form L-M,
and Achievement Indices." Unpublished doctoral dissertation. West
Virginia University, 1967.
DeLapa, G. "The Slosson Intelligence Test: A Screening and Retesting
Technique for Slow Learners." Jom^tiQ’l of School Psychology 6
(1968) :224-25.
Duggan, M.D. "A Study of the Relation Between the Slosson Reading and
Intelligence Tests and Other Standardized Tests at the Second
Grade Level." Unpublished master's thesis. Cardinal Stritch
College, 1968.
Eichorn, Dorothy H. "Review of Slosson Intelligence Test. In
Pediatric Screening Tests, edited by William K. Frankenburg and
Springfield, IL: Charles Thomas, Publisher, 1975.Bonnie Camp.
81
Fagert, C.M. "The Relationship Between the Slosson Intelligence Testand the Wechsler Intelligence Scale for Children. " Unpublishedmaster's thesis, Kent State University, 1968.
Fudala, Janet B. "Differential Evaluation of Students with the SIT."Academic Therapy 15 (1979)
.
Garrison, M. , Jr., and Hammil, D. "Who Are the Retarded?" ExceptionalChildren 38 (1971) :13-20.
Gellespie, P.H. "A Study of the Performance of Dyslexic and NormalReaders on the Slosson Intelligence Test for Children and Adults."Unpublished doctoral dissertation. West Virginia University, 1970.
Guilford, J.P. Fundamental Statistics in Psychology and Education .
4th ed. McGraw-Hill, 1965.
Hammill, D. "The Slosson Intelligence Test as a Quick Estimate of
Mental Ability." Journal of School Psychology 7 (1968-69) : 33-37.
Hammill, D. ; Crandel, J.M. , Jr.; and Colarusso, R. "The SlossonIntelligence Test Adapted for Visually Limited Children."Exceptional Children 36 (1970) :535-536.
Hatcher, Catherine W. , et al. "A Comparison of Various Measures of
Intelligence with the WISC Among Disabled Readers." Paper pre-
sented at the annual meeting of the International Reading Associ-
ation, New York City, NY: May 13-16, 1975.
Hedl, J.J., Jr. "An Evaluation of a Computer-based Intelligence Test."
Office of Naval Research, Washington, D.C. Personnel and Training
Research Programs Office, 1971.
Hedl, J. J. ,Jr. "Computer-based Intelligence Testing." Paper presented
at annual meeting of the American Education Research Association,
New York, NY, 1971.
Hedl, J.J., Jr.; O'Neil, H.F., Jr. ; and Hansen, D.N. "Affective
Nature of Computer-based Testing Procedures." Abstract, Pro-
ceedings of the 79th Annual Convention of the American Psychologi-
cal Association 6 (1971) : 535-36.
Hill, C. "Prediction of Slosson Intelligence Test Scores from Stanford-
Binet Form L-M Test Scores." Unpublished master's thesis. Univer-
sity of Toledo, 1969.
Houston, C., and Otto, W. "Poor Reader's Functioning on the Wechsler
Intelligence Scale for Children, Slosson Intelligence Test and
Quick Test." Journal of Educational Research 62 (1968) :157-59.
82
Hunt, J.V. "Review of Slosson Intelligence Test (SIT)." In TheSeventh Mental Measurements Yearboolc . Edited by Oscar yT.
Buros. Highland Park, NY: The Gryphon Press, 1972.
Hutton, J.B. "Practice Effects on Intelligence and School ReadinessTests for Pre-school Children." Training School Bulletin 65(1969) :130-34.
Hutton, J.B. "Relationships Between Preschool Screening Test Data andFirst Grade Academic Performance for Head Start Children."Unpublished doctoral dissertation. University of Houston, 1970.
Hutton, J.B. "Relationships Between Teacher Judgement, ScreeningTest Data and Academic Performance for Disadvantaged Children."Training School Bulletin 68 (1972) : 197-201
.
Jerrolds, R.W.; Callaway, R.; and Gwaltney, W.K. "Comparison of the
Slosson Intelligence Test and WISC Scores of Subjects Referredto a Reading Clinic." Psychology in the Schools 9 (1972):
409-410.
John, J.L. "An Informal Item Analysis for the Slosson Intelligence
Test." Minnesota Reading Quarterly 19 (1975) : 164-67
.
Johnson, D.L., and Johnson, C.A. "Comparison of Four Intelligence
Tests Used with Culturally Disadvantaged Children." Psychological
Reports 28 (1971) :209-10.
Johnson, D.L. ,and Shinedling, M.M. "Comparison of Columbia Mental
Maturity Scale, Peabody Picture Vocabulary Test and Slosson
Intelligence Test with Mentally Retarded Children." Psychological
Reports 34 (1974) : 367-70.
Johnson, D.L. "A Comparison of Three Intelligence Tests in Assessing
the Mental Abilities of Mentally Retarded Children." Unpublished
doctoral dissertation, Brigham Young University, 1972.
Jongeward, P.A. "A Validity Study of the Slosson Intelligence Test for
Use with Educable Mentally Retarded Students." Doctoral disserta-
tion, University of Oregon, 1969. Dissertation Abstracts Inter-
national , 30, 3323A (University Microfilm No. 70-25, 12).
Kapel, M.B., and Kapel, D.E. "Hebrew English Reading Achievement in a
Jewish Day School: A Comparison Among Reading Achievements and
Attitudes Toward the Two Languages." Jewish Education 40 (1970).
23-32.
Karnes, Frances A., and Brown, K. Eliot. "Comparison of the SIT with
the WISC-R for Gifted Students." Psychology in the Schools_ 16
(1979)
.
83
^ H* ^ 3.nd Ivsnoff ^ J . '*Th0 Slosson Int0 llig0ncG T0st 0s ci
Scr00ning Instrum0nt with a R0habilitation Population."Excaptional Childron 35 (1969) :745.
Koany, M. "A Study of tha Ralation Batwaan tha Slosson Raading andIntalliganca Tasts and Othar Standardizad Tasts at tha SixthGrada Laval." Unpublishad mastar's thasis, Cardinal StritchCollaga, 1968.
Kilduff, C.T. "A Study of tha Ralation Batwaan tha Slosson Raadingand Intalliganca Tasts and Othar Standardizad Tasts at thaFourth Grada Laval." Unpublishad mastar's thasis, CardinalStritch Collaga, 1969.
Klain, Alica E. "Tha Raliability and Pradictiva Validity of thaSlosson Intalliganca Tast for Pra-Kindargartan Pupils."Education and Psychological Maasuramant 38 (1978) : 1211-17
.
Lamp, R.E., and Traxlar, A.J. "Tha Validity of tha Slosson Intalli-ganca Tast for Usa with Disadvantagad Haad Start and First GradaSchool Children. " Journal of Community Psychology 1 (1973):
27-30.
Lamp, R.E.; Traxlar, A. J. ; and Gustafson, P.P. "Predicting AcademicAchievement of Disadvantaged Fourth Grade Children Using the
Slosson Intelligence Test." Journal of Community Psychology 1
(1973) :339-44.
Lessler, K. , and Galinsky, M.D. "Relationship Between the Slosson
Intelligence Test and Wechsler Intelligence Scale for Children
Scores in Special Education Candidates." Psychology in the
Schools 8 (1971) :341-44.
Lowrance, D. ,and Anderson, H.N. "A Comparison of the Slosson
Intelligence Test and the WISC-R with Elementary School
Children." Psychology in the Schools 16 (1979).
Machen, L.H. "A Validity and Reliability Study of the Slosson
Intelligence Test with an Atypical Population; Gifted Children."
Unpublished doctoral dissertation. Catholic University, 1972.
MacKinnon, R.C. ,and Elliott, C. "A Comparison of an Achievement
Battery with Two Tests of Ability and Educable Mental Retardates."
Final Report. Office of Education (DHEW) ,Washington, D.C.,
Bureau of Education for the Handicapped, 1969.
Martin, John D. ,and Kiewell, Jane C. "Intercorrelations of the
Wechsler Intelligence Scale for Children - Revised, the Slosson
Intelligence Test and the National Educational Development Test."
Educational and Psychological Measurement (Winter 1977).
84
Massachusetts Department of Education. Chapter 766 ReoulationRJuly 1981. '
Maxwell, M.T. "The Relationship Between the Wechsler IntelligenceScale for Children and the Slosson Intelligence Test." ChildStudy Journal 1 (1971) : 164-71.
'
McRae, J. "A Comparison of the Wechsler Preschool and Primary Scaleof Intelligence with the Slosson Intelligence Test." Unpublishedmaster's thesis. Eastern Washington State College, 1968.
Meissler, G.R. "A Correlation of the Slosson Intelligence Test andthe Wechsler Intelligence Scale when Administered to AtypicalChildren.” Unpublished doctoral dissertation. Catholic Univer-sity of America, 1970.
Mesinger, J.F. "An Intelligence Test to Assess Compensatory EducationPrograms for Disadvantaged Youth." University of VirginiaResearch 7 (1969) :52-58.
Miller, C.K. "Conservation in Blind Children." Education of theVisually Handicapped 1 (1969) :101-5.
Miller, W.D. "Usefulness of the Peabody Picture Vocabulary Test andthe Slosson Intelligence Test in a Teacher Training Progrcim."Southern Journal of Educational Research 6 (1972) :71-78.
Miller, W.D. "Predicting Reading and Interpreting IQ Differences."Southern Journal of Educational Research 71 (1973) : 31-39.
Mize, John M. ; Callaway, Bryon; and Smith, Janie Wooten. "Comparisonof Reading Disabled Children's Scores on the WISC-R, PeabodyPicture Vocabulary Test and the Slosson Intelligence Test."
Psychology in the Schools 16 (1979) :356-58.
Nash, M.S. "The Development of Depth Perception in Intermediate Age
Children." Unpublished doctoral dissertation. University of
Kentucky, 1969.
Nicholson, C.L. "The Use of Four Screening Instruments. In selected
Papers on Learning Disabilities of the Sixth Annual International
Conference of the Association for Children with Learning Disabili-
ties, 1969, San Rafael, California." Academic Therapy (1969):
101-7.
Nie; Hull; Jenkins; Steinbrenner ; and Berel. Statistical Package for
the Social Sciences. 2nd Edition. McGraw-Hill, 1976.
85
Nicholson, C.L. "Analysis of the Functions of the Slosson IntelligenceTest." Perceptual and Motor Skills 31 (1970) ;627-31.
O'Keefe, S.L, "An Inquiry into the Use of the Slosson IntelligenceTest with Fast Learners." Unpublished master's thesis, OhioState University, 1967.
O Neill, H.D. Partial Validation of the Slosson Intelligence Test."Unpublished doctoral dissertation. University of Oklahoma, 1969.
Pate, R.H., Jr., and Nichols, W.R, "The Slosson Intelligence TestUsed as a Screening Device: A Validity Study." American Psycho-logical Association. Catalog of Abstracts . Experimental Publica-tion System, 1970:9.
Pikulski, J.J. "Validity of Three Brief Measures of Intelligence forDisabled Readers." Journal of Educational Research 67 (1973):67-68.
Poissant, P.A. "A Product Moment Correlation Between the Stanford-Binet and Slosson Intelligence Tests with Slow Learners."School Psychological Newsletter 12 (1) (1967)
.
Poole, M.J. "The Development of a Spanish Version of the SlossonIntelligence Test." Unpublished doctoral dissertation.University of Rochester, 1974.
Raskin, L.M.; Offenbach, S.I.; and Black, K.N. "Relationship Between
the Peabody Picture Vocabulary Test and the Slosson Intelligence
Test in Preschool and Third Grade Children." Psychology in the
Schools 11 (1974) :66-68.
Reynolds, William M. "A Caution Against the Use of the Slosson Intelli-
gence Test in the Diagnosis of Mental Retardation." Psychology
in the Schools 16 (1979) :77-79.
Ritter, David; Duffey, James; and Fischman, Ronald. "Comparability of
the Slosson and S-B Estimates of Intelligence." Journal of
School Psychology 11 (1973) : 224-27.
Sattler, J. Assessment of Children's Intelligence . W.B. Saunders
Company, 1979.
Schooler, Douglas L. ,and Anderson, Robert L. "Race Differences on
the Developmental Test of Visual Motor Integration, the Slosson
Intelligence Test, and the ABC Inventory." Psychology in the
Schools 16 (1979) :453-56.
Schwarting, F.G. "A Comparison of the WISC and WISC-R." Psychology
in the Schools 13 (1976) :139-41.
86
Shepher, C.W., Jr. "Childhood Chronic Illness and Visual PerceptualDevelopment." Exceptional Children 36 (1969) ; 39-42
.
Sinco, H.J. "A Concurrent Validity Study of the Slosson IntelligenceTest and the Wechsler Intelligence Scale for Children." Master’sthesis, Marywood College, Scranton, PA, 1973.
Slosson, R. "Slosson Intelligence Test." East Aurora, NY: SlossonEducational Publications, 1961.
Slosson, R.L. Slosson Intelligence Test for Children and Adults . 2ndedition. East Aurora, NY: Slosson Educational Publications, 1981.
Smith, L.L. "Comparison of Reading Expectancy Sets as Determined fromSele-ted Intelligence Measures." Unpublished doctoral disserta-tion, University of Illinois, 1972.
Stewart, Kenneth D. , and Jones, Elivis C. "Validity of the Slosson
Intelligence Test: A Ten Yecir Review." Psychology in the Schools
(October 1976)
.
Stewart, K.D., and Myers, D.G. "Long Term Validity of the Slosson
Intelligence Test." Journal of Clinical Psychology 30 (1974):
180-81.
Stewart, K.D.; Wood, D.Z.; and Gallman, W.A. "Concurrent Validity of
the Slosson Intelligence Test." Journal of Clinical Psychology
27 (1971) :218-20.
Stone, Mark. "An Interpretive Profile for the Slosson Intelligence
Test." Psychology in the Schools (July 1975) : 330-33.
Stuhler, A.M. "An Experimental Study of the Relation Between the
Slosson Reading and Intelligence Tests and Other Standardized
Tests at the First Grade Level." Unpxiblished Master's thesis.
Cardinal Stritch College, 1970.
Swanson, M.S., and Jabodson, A. "Evaluation of the Slosson Intelligence
Test for Screening Children with Learning Disabilities. Journal
of Learning Disabilities 3 (1970) :318-20.
Terman, L.M. , and Merrill, M.A. Stanford-Binet Intelligence Scale.
Boston: Houghton-Mifflin , 1960.
Terman, L.M. , and Merrill, M. Stanford-Binet Intelligence Scale, Manual
Boston, Houghton-Mifflin, 1973.
Thorndike , R. Stanford-Binet Intelligence Scale, 1972 Norms Tables.
Boston: Houghton—Mifflin Company, 1973.
87
Trivedi, A. "A Comparison of Three Intelligence Tests for the Assess-ment of Mental Retardates." Journal of Mental Deficiency Re-search (December 1977)
.
Watson, C.G., and Klett, W.G. "The Henmon-Nelson, Cardall-Miles
,
Slosson and Quick Tests as Predictors of WAIS IQ." Journal ofClinical Psychology 31 (1975) : 310-13.
Wechsler, D. Wechsler Intelligence Scale for Children: Manual .
New York; Psychological Corporation, 1949.
Wechsler, D. Wechsler Intelligence Scale for Children; Revised .
New York; Psychological Corporation, 1974.
Whitacre, R.L. "Use of the Slosson Intelligence Test with the
Stanford-Binet, L-M as a Screening Device for Slow Learner
Placement." Unpublished master's thesis. Bowling Green
State University, 1968.
APPENDIX A
89
SCHOOL SYSTEMS SURVEYEDTo Determine the Feasability of Conducting ResearchP®^taining to the Use of the Slosson Intelligence
Test (Revised) in Place of the WechslerIntelligence Test for Children-Revised
Ayer Public SchoolsAntietman StreetAyer, MA 01432
Methuen Public Schools75 Pleasant StreetMethuen, MA 01844
Beverly Public Schools4 Colon StreetBeverly, MA 01915
N. Middlesex School DistrictP.O. Box 501Townsend, MA 01469
Chelsea Public SchoolsWalnut StreetChelsea, MA 02150
North Reading Public SchoolsSherman RoadNorth Reading, MA 01864
Dracut Public Schools1540 Lakeview AvenueDracut, MA 01826
Peabody Public Schools210 Washington StreetPeabody, MA 01960
Ipswich Public SchoolsOne Lord's SquareIpswich, MA 01938
Revere Public Schools101 School StreetRevere, MA 02151
Lawrence Public Schools58 Lawrence StreetLawrence , MA 01840
Saugus Public Schools25 Main StreetSaugus , MA 01906
Lowell Pxablic Schools57 Tenth StreetLowell, MA 01854
Stoneham Public Schools101 Central StreetStoneham, MA 01280
Lynn Public Schools42 Franklin StreetLynn, MA 01905
Wakefield Public Schools525 Main StreetWakefield, MA 01880
Marblehead Pioblic Schools
2 Humphrey StreetMarblehead, MA 01945
Woburn Public Schools
Elm StreetWoburn, MA 01801
Marlboro Pioblic Schools
Bolton StreetMarlboro, MA 01752
Worcester Public Schools
20 Irving Street
Worcester, MA 01609
90
EducationalSpecialistsa//ociate/ 160 Pleo/ont Street ftortft RncJover fTlfl 0IW5 617-685*0111
Dear Director of Special Education:
In anticipation of completing my doctorate this comingyear, I need some direct assistance from you in gatheringdata. I have spoken to some of you individually about thisproject, bat not to all the Directors I had hoped to. Mydissertation hopefully will benefit all of us so that anyimposition at this time will ultimately be rewarded in thefuture
.
Specifically, I am attempting to validate the SlossonIntelligence Test in relation to the WISC-R. Doctors Armstrongand Jensen have re-normed the Slosson Intelligence Test inrelationship to the revised Stanford-Binet (1972) . Sincethe V7ISC was revised, little evidence of the validity ofthe SIT in relation to the WISC-R has been produced. Therefore,I am attempting to provide the validity between the re-normedSIT and the WISC-R.
Individual intelligence tests such as the WISC-R andS-B require specialized training in their administration,scoring and interpretation. They are also time consuming inthat they require an average administration time of one hour.Thus, there is a need for a valid individual test of mentalability which is quickly administered and easily scored. Sucha test could be used as a screening and/or retesting instrument.
I have taken the liberty of enclosing materials that areneeded to gather the required data. Naturally the largerthe sample the more accurate the results, so there is not a
specific number of student data requested from you. I willgratefully accept any information you can provide.
I am asking that the Slosson Intelligence Test beadministered to any student who has had, or will have, a
WISC-R administered between March 1, 1980 and June 26, 1981.
The sample will include ages 5 through 16.
To be returned to me are (without identity of student)
:
1. The Slosson score sheet
2. A copy of the front page WISC-R
3. The questionnaire for each administration of theSlosson
.
If you need more copies, please make additional copies orcontact me at 683-0111 and I will provide what you need.
The results of this study have national significancebesides my personal needs.
Drs. Armstrong and Jensen have agreed to let me reportback to you, if you desire, any student's IQ as indicated onthe re-normed Slosson Intelligence Test. Just use your owncode number on the yellow sheet and I will give you the IQfor the code number.
Thank you in advance for your help and participation inthis project. If there is any question please call me at683-0111.
Sincerely
,
James C. Bradley
JCB/ct
Enclosure
IMPORTANT; Each individual student's material should bestapled together.
92
GducQtionQlSpecialistso//ociQte/ 160 PteoAant Street OortfS findover fTlfi OIM5 617-683-ail
February 27, 1981
1
To Directors of Special Education:
This is a follow-up to the packaqe that was sent to youon Dccen)l)er 1, 1980 containing data sheets, etc. for mydissertation - Slosson Intelligence Test compared to theWISC-R, The response to date has been positive, however I
still need more data sets to complete my study. If you havestarted tl:e test, please send me whatever you have at thistime, and any further tests at a later date.
I would like to suggest that you find students who havehad the WISC-R since a year ago (March 1, 1980) and administerthe Slosson to some of them. Also, many of the initial responsesreceived were found in student files where both tests had beenpreviously administered. Re-evaluations art taking place andthis may be the time you choose to complete my request.
I would appreciate any help you can give me, even thoughit is an imposition on you and your staff. If I could get an
average of 15 per system, enough data would be generated to
start data analysis.
Thank you in advance for your cooperation. If you have
misplaced the material, have any questions, or need more material
please call me at 683-0111.
I plan to he at the March 4 Directors meeting if you have
any results or questions.
1
Sincerely
,
y
:)CD/ct
93
REQUEST FOR PARENTAL PERMISSION
I am conducting a doctoral study relating to the comparison of twotests generally given to students who are referred to Special Needs.They are the WISC-R and the Slosson Intelligence Test.
Your child has already been administered the WISC-R in the pastby school personnel. I would like your permission to have themadminister the Slosson Test. The Slosson Test is very short and easilygiven .
I require names for identifying information about you or yourchild, just the two test results.
This study will be compiled with results from many other cities
and towns and will have national significance in the identifying and
subsequent helping of special needs students.
If you give permission to have your child participate in this
study, thank you. Please fill out the bottom part of this permission
slip.
Thank you.
James C. Bradley
I give permission for my child(name of child)
to have the Slosson Intelligence Test administered and the results
only given to Mr. James Bradley.
APPENDIX B
95
SLOSSON/WECHSLER TEST DATA FORM
TEST ADMINISTRATOR INFORMATION
Wechsler Administrator Slosson Administrator (if different)
Last Name First Last Name ^First
Position/Title Position/Title
School System School System
SLOSSON (SIT) ADMINISTRATOR
Yes No Have you administered Slosson Intelligence Tests priorto participating in this research study?
Yes No Do you feel you are a competent Slosson IntelligenceTest administrator? If no, please specify reason:
Yes No Do you consider the results from this Slosson IntelligenceTest administration to be valid? If no, please specifyreason (e.g. student was restless, too many interruptions)
Yes No Are you certified to administer the Wechsler?
Yes No Are you certified to administer the Stanford-Binet?
WECHSLER (WISC-R) ADMINISTRATOR
Yes No Did you also administer the Slosson to these students as
part of this research study?
Yes No Do you consider the results from this administration of
the Wechsler to be valid? If no, please specify reason
(e.g. student was restless, too many interruptions, etc.)
Yes No Are you certified to administer the Stanford-Binet?
96
STUDENT INFORMATION
SEX : Female Male Grade Level Student PrototypeNumber (if applicable)
Please specify any student handicaps:
TEST ADMINISTRATION DATA
Slosson Data
Date of Test: yr.
Chronological Age: yr.
Mental Age :yr
.
Ratio IQ: yr.
Wechsler Data
Date of Test: yr.
Chronological Age; yr.
Verbal IQ:
Performance IQ:
Full Scale IQ:
Administration Order:
mo. daymo. daymo. day
mo. day
mo.
day
mo.
^day
Wechsler First Slosson First
Time Between Test
^same day
^1-7 days8-14 days
Administrations
15-29 days30-44 days45-60 days
If more than 60 days
please specify days
Testing Time: Administration and Scoring Only—DO NOT Include Initial^
Rapport Establishment Time as Part of Administration Time
Test Administration Time Scoring Time
Wechsler Hr(s) mins. Hr(s) mins.
Slosson Hr(s) mins. Hr (s) mins.
Total Time
Hr(s) mins.
Hr(s) mins.
PLEASE BE SURE TO ATTACH A COPY OF THE FRONT PAGE OF THE WECHSLER RECORD
form and the SLOSSON SCORE SHEET. WE ALSO ASK THAT STUDENT NAMES BE
INKED OUT. THANK YOU FOR YOUR COOPERATION.
97
SI.OSSON INTEI.LI CllNCE TEST (SIT)
Fon
CHIIDBEN AND ADULTS
Cupyrlqiit © 1963. Richard L. Sloison. M.A.
scouK siii:i:t
CA
IQ s MA X 100 MA
CAIQ
V r Alt MON IMt
Niimo
Add rnss
LAST FIRST Ml OOL C
School Grade
tlxainiiinr
Present time
Dirth elate
Age or CA
TTB “Tnnrrn
—
~nnv
VCAn MON IM ~B.V
•<6n I
Plndlng the MABasal ace
IJUl
Added months
Finding the IQ
CA
MA
IB
* Thr CA never
exceed! tS yeert
or 102 month!.
f 00 (HA in monthsi
CONVERTING 1-12 . 4 - 48 7 - 84 lQ-120 13-156 16-192 19-228 22-264 25- 300
YEARS 2-21 5-60 8-96 11-132 14-168 17-204 20 - 240 23-276 26-312TO MONTHS: 3-30 0-72 9-108 1 2-144 15-180 18-216 21-252 24- 288 27 - 324
1/2 MONTH' S CREDIT 1 MOUTH'S CREDIT 2 MONTH'S CREDIT 3 MONTH' S CREDIT
Years and mon tlis Years and months Yoars and months Years and months
0-0. 5 1-0.0 0 1 4^ ho 9-0 13-0 10-8 2M0-1.0 1-0.5 2-1 4-1 5-2 9-2 13-2 16-3 U-JL0-1. S 1-1.0 4-2 5-4 9-4 13-4 10-6 22-6
o o 1-1. 5 2-3 4-3 5-6 9-6 13-6 16-9 22-0
0-2.5 1-2.0 ird Jti 9-8 13-8 17-0 23-0
0-3.1 1-2.5 2-5 4-5 5-10 9-10 13- 10 17-3 23-3
0-3. 5 1-3.0 2-6 4-0 6-0 10-0 14-0 17-6 23-6
0-A. 0 1-3. 5 2-7 4-7 6-2 10-2 14-2 17-9 23-9
0-4.5 1-4.0 _2;8 ±1 ±Ji 10-4 H-4 LSlO ilia
0 -r.. 0 1-4. 5 2-9 1*9 6-0 10-0 14-6 10-3 itJ0-5. 5 1-5.0 2- 10 4-10 6-8 10-8 14-8 18-6 24-6
0-f.. 0 1-5. 5 2-11 vn 6- 10 10-10 M-iq 18-9 ilia
0-B. 5 1-C.O 3-0 7-0 n-9 ISlS 19-0 25-0
0-7. 0 1-0. 5 3-1 7-2 11-2 15-2 19-3 25-3
0-7. 5 1-7.0 :i-2 7-4 11-4 15-4 19-6 25-0
O-P.O 1-7. .5 3-3 7-6 11-6 15-g 19-9 2.5-9
0-H. 5 1-8.0 3-4 7-8 11-8 15-0 20-0 26-0
0-9.0 1-8. .5 3-5 7- 10 11-10 15-10 20-3 26-3
0-9. 5 1-9.0 iJi 8-0 bi 20-6 20-6
0- 10.0 1-9.5 3-7 8-2 12-2 20-9 3(5:1
0-10.
5
1- 10.0 3-8 8-4 .121.121-0 27-0
0-11.0 1-10. !) 3-9 8-6 ll!-0 21-3
OJi-S. 1-11.0 2-_ia 8-8 12-0 21-8
1-11.5 ;L'JLi 1
8- 10 12- 10 21-9
1/2 X 1 X - 2 X _ - 3 X_ -
r H I II 1
1
fi I U . 5 . A .
Sl.l)SS()N •MlfJNAt. ft mi ICA I IONS. INf
f.t). Il.i> .’so. f:;!.! Ntw V.iik IJ:l.l2
98
3-'J, l-O. •l-jl.
3- 1 i\ cook In for me
like this."
3'
6
“Which of these squares
is smaUer?"
5~ 2 "Draw a block for me
like this."
fi-O (Dnielns AiiplOH)
Triel I. Trial 2.
Trial I. Trial 2.
Trial I. Trial 2.
(llnclrcle the nuaibers of all
arltliantlc pmliluas os you co«e
to thcoi so that the personbelnc tested con look at then
while fonaiilatini; on answer.)
9-8 2a
10-
8 45V U
11
-
10 36
12-2 s< 654
12-8 504
13-0 13
14-0 404 104
14-0 804 2 04(lUviMd repdai 19T0
14-6 12 20 6:00
14-10 504 Mi 6
15-6 9 12
16-0 S5.00 754 304
16-0 $10.00 SI. SO(RavlMd reprlat 1977)
604
16-9 300 5 1
19-0 3/5 2/3 5/8
19-9 1 40
20-3 $40,000 5 3
20-9 5 40 8
21-9 10 2 3
22-3 216
22-9 1000 107. 5%
23-3 $1000 8 12
24-0 2, 4, 12. 48.
24-3
99
t,::
w 'j y li'u
1 RECORDFORM
Wo<h>l«r Inlolligunco Scalo
for Chitdron—Rovutd
NAME ACE SEX
ADDRESS
PARENT'S NAME
SCHOOL GRADE
PLACE OF TESTING TESTED BY
REFERRED BY
WISC-R PROFILE
Ciioicirint who wirh lo drow o prolilo iKoijId firtt tronrfbr Iho child't %coUtl icurof to tho row
briow. Then inork on X on the doi corresponding lo iho icoirrd Koro for eoch Icit, and dio
connecting Iho X'l.'
of boaoiw o tine
VERBAL TESTS
Srolerl
Scorn
19
IB
17
16
IS
14
13
12
1
1
10
9
e
7
6
5
4
3
2
1
PERFORMANCE TESTS
si s r“
2 t 3 e.X o t Oaana it:
*S«:« CScpIcr i in Iho monuol for o ditev
19
18
17
16
IS
14
13
12
11
10
9
8
7
6
5
4
3
2
1
on cf tho irgnificonc renett bolwoon icorot on I
Senind
krur.
19
18
17
16
IS
14
13
12
11
10
9
8
7
6
S
4
3
2
1
NOTES
Yoor Monlli Day
Dole Tbtled
Dole of Birth
Ao«
Row Scoled
Score Score
VERBAL TESTS
Infnrmnlinn _
Similoriliei —Arilhmelic
Vocobulory —rninprrhcntiAfi —(Digil SponI ( ) 1 1
Verbol Score—PERFORMANCE TESTS
Picture Completion
Block Design
Object Assembly _
Coding —(MozesI (
'
1 )
Performonco Score
Scoled
Score lO
Verbol Score•
Performonce Score•
Full Seole Score
‘Prorolod Irorn 4 losll. If nocoiiory.
D
Coorrithi 41 1071. I07C k» Iho r.cchninolrol tornororl.n.
«l,l. .I...C4 N. po.l ol Ihit rn.nrO lor.. ...» ^ Mll'’n’.r!l,«l I"’o!.li€.rl.n In'
rr::' ‘-r"*-U« Cw^****!!***. Ilp-w \ iFfc. M.r. 100I7
••v »**P*f« **4
74-1 0.1Ar. 0 000334