|
|
 |
Test
Review: Slosson Intelligence Test - Revised (SIT-R)
Jennifer A. McKechnie, M.Ed. & Bradley T. Erford, Ph.D.
Loyola College in Maryland
- General Information
- Title:
Slosson Intelligence Test (SIT-R)
- Authors:
Richard Slosson; revised by Charles L. Nicholson and Terry Hibpshman.
- Publisher:
Slosson Educational Publications, Inc., P.O. Box 280, East Aurora,
NY 14052.
- Forms;
groups to which applicable: Ages 4 through adult.
- General
type: Mental (verbal) ability of children and adults.
- Date of
publication: 1990; 1998 calibrated norms revision.
- Practical
features: The SIT-R must be administered individually and hand-scored.
No computerized administration, scoring and interpretive software
is available. Total Standard Score (M = 100, SD = 16), percentile
rank, and mean age equivalent (MAE) interpretive statistics are
available.
- Cost:
$79.00 per test kit.
- Time required
to administer: The time required to administer and score this
test varies from about 15 to 20 minutes for the average person (Kunen,
Overstreet, & Salles, 1996) or 20 to 30 minutes for the slow,
the very gifted, or the person who is deficient in certain areas
while normal or high in other areas (Nicholson & Hibpshman,
1998).
- Purpose and
Nature of the Instrument
- Stated
purpose: “The SIT-R is designed to be an individual test
for use in screening or estimating the cognitive ability of an individual,
public school student, college student, mental patient or mentally
handicapped person. Because the SIT-R is a screening instrument
it alone should not be used in final placement decisions” (Nicholson
& Hibpshman, 1990, p. 1). The original Slosson (SIT; Slosson,
1963) was frequently used as a quick screener for mental retardation
(McCormick, Campbell, Pasnak, & Perry, 1990; McCuller, Salzberg,
& Linnuris Kraft, 1987) and giftedness (Clark, McCallum, Edwards,
& Hildman, 1987; Karnes, Whorton, & Currie, 1986). Others
have raised concerns over its usefulness for intelligence screening
(Harris & Reid, 1991; Salvia & Ysseldyke, 1995).
- Description
of test, items and scores: Items on the SIT-R are not presented
in a subtest format, because the scale is comprised of 187 questions
presented in a unidimensional arrangement with age appropriate starting
markers. Items were derived from six cognitive domains (Nicholson
& Hibpshman, 1990): Vocabulary, General Information, Similarities
and Differences, Comprehension, Quantitative, and Auditory Memory.
- Use in
counseling: The SIT-R is used primarily for assessment of client
verbal skills and ability.
- Practical
Evaluation:
- Usefulness
of manual: The manual for the SIT-R is user-friendly, albeit
not comprehensive. Instructions are easy to follow and straight
forward. The test is administered orally and has been scripted.
Words to be repeated verbatim by the administrator are highlighted
in blue.
- Adequacy
of directions for administering the instrument: Administration
and scoring procedures are clearly explained in the manual. The
SIT-R uses a small number of verbal item types and a dichotomous
item scoring system making the testing process intuitive to use
(Kemphaus, 1994). A concise summary of administration and scoring
procedures is located on the inside of the manual's front cover.
Mean Age Equivalent (MAE) and Total Standard Score (TSS) are easily
obtained from the norm table in the technical manual (Nicholson
& Hibpshman, 1998).
- Qualifications
of examiners: According to Nicholson and Hibpshman
(1990), to effectively administer, score and interpret the SIT-R
the examiner should have Level B qualifications (graduate degree
in psychology, education or a related field, and a course in testing
and measurement).
- Scoring
provisions: A basal score is defined as the highest level at
which
the examinee obtains ten consecutive correct items. The ceiling
item is the last correct item before the examinee misses ten consecutive
items. The total raw score is a simple sum of scores and is the
total number of correct items above the basal added to the basal
item (which gives credit for all the items below the basal item).
- Technical Considerations
- Normative
sample: Children and adolescents aged 4 - 18 (51% female,
49% male) comprised the norm group. The authors suggested that the
18-year-old norms can be generalized to interpret the scores of
adults. The 1998 technical manual involved a recalibration of the
1990 norms to reflect recent research and correct for apparent discrepancies
between the WISC-III (Wechsler, 1991) and SIT-R. Protocols were
collected on more than 2,400 individuals and a sample of 1,854 was
selected to match, as close as possible, U.S. population demographics
(Nicholson & Hibpshman, 1998). Geographically, the sample was
comprised of participants from the Northeast (20%), North Central
(22%), South (38%), and West (20%) United States. Racially, the
sample was comprised of Black (14%), White (83%), and Other (3%)
participants. The Other category included Native, Asian, Hispanic,
Pacific Island Americans and other similar groups. The noted disproportion
of these individuals was reportedly due to the testing protocol
requiring the subject to speak English. The primary language for
many of the individuals comprising minority populations was not
English (Nicholson & Hibpshman, 1998). Distribution by size
of residential populations included: 1,000,000 or more (1%); 500,000
- 999,999 (3%); 100,000 - 499,999 (8%), 50,000 - 99,999 (13%), 25,000
- 49,999 (14%), 10,000 - 24,999 (18%), 5,000 - 9,999 (14%), 2,500
- 4,999 (14%), 0 - 2,499 (14%) (Nicholson & Hibpshman, 1998).
Percentages of occupational groups reported by fathers of children
comprising the standardization sample included: Professional (27%),
Technical/Office/Sales (28%), Service (20%), Production/Craft/Repair
(8%), Operators/Fabricators (11%), and Farm/Fishing/Forestry (7%)
(Nicholson & Hibpshman, 1998). Percentages of educational levels
reported by fathers of children comprising the standardization sample
included: less than high school (19%), high school (39%), some post
secondary (24%), college degree and beyond (19%) (Nicholson &
Hibpshman, 1998).
- Reliability:
Internal consistency was calculated using the Kuder-Richardson - Formula 20
(KR-20). Reliability coefficients determined
by age level ranged from .88 to .97, with a median of .945, indicating
a high level of internal consistency (Nicholson & Hipbshman,
1996). Based on a sample size of 41 children and using a one-week
administration interval, the test-retest reliability was reported
to be .96 (Nicholson & Hipbshman, 1998).
- Validity:
Concurrent criterion-related validity has been reported through
correlations between the SIT-R Total Standard Score (TSS) and the
Wechsler Intelligence Scale for Children-Revised (WISC-R; Wechsler,
1974) IQ's. In four samples of children aged 6 to 16 (total n =
234). The SIT-R TSS correlated .829 - .914 with the WISC-R VIQ (verbal
intelligence quotient; median = .861). The SIT-R TSS correlated
.376 - .837 with the WISC-R PIQ (performance intelligence quotient;
median = .521), while the SIT-R TSS and WISC-R full-scale correlated
.612 - .920 (median r = .794; Nicholson & Hibpshman, 1998).
Also, “data do not seem sufficient to have a significant bearing
on concurrent validity for testing of individuals over the age of
16 years” (Campbell & Ashmore, 1995, p. 116). In a study
by Kunen, Overstreet and Salles (1996) the SIT-R was found to correlate
r = .92 with the Stanford-Binet (SBIS-4; Thorndike, Hagan &
Sattler, 1986), but the SIT-R did not adequately match the SBIS-4
IQ category assignments at levels other than mental retardation.
The SIT-R was intended as an intellectual screening test, specifically
measuring the verbal intelligence factor. The six domains from which
the SIT-R items were developed are similar to items on verbal subtests
by Wechsler (1991) and the Stanford Binet Intelligence Scale - Fourth
Edition (Thorndike et al., 1986). The domains are meant to measure
primarily global intellect, defined by Wechsler as "The capacity
of the individual to act purposefully, to think rationally, and
to deal effectively with his environment" (Nicholson &
Hibpshman, 1998, p. 1) and also include memory and crystallized
ability. Memory was defined as the ability to recall information
about what one has learned, including facts, visual images and taught
experiences. Crystallized ability was defined as intelligence related
to culture, life experiences and one's environment and can be broken
down into two further components: verbal ability and quantitative
reasoning (Nicholson & Hibpshman). No exploratory or confirmatory
factor analytic results were provided to support the unidimensionality
of the SIT-R. Correlations between the WISC –R verbal IQ and
SIT-R were very high (see concurrent validity) and the SIT-R does
show high internal consistency coefficients, a source of evidence
which is necessary but not sufficient to verify the construct validity
of the SIT-R.
- Evaluation
- Comments
of reviewers and general evaluation: The SIT-R is a quick,
easy to use screening test for verbal intelligence. The test instructions
are clear and easy to administer in a short period of time, which
makes it a useful tool for any professional administering the test.
The manual contains an easy to follow summary of administration
and scoring procedures, which is located inside the front cover
for quick reference. There is also a checklist for examiners with
ideas that help with proficiency in administering the test. The
cost is considerably less than comparable screening measures and
far less than diagnostic IQ tests. The manual is more comprehensive
than its predecessor providing adequate reliability data and some
basic validity data related to the WISC-R. When used as a verbal
intelligence screener, the SIT-R is an effective tool for identifying
individuals needing further evaluation. However, evaluators must
be careful to only use the SIT-R for screening, not diagnostic decisions.
Weaknesses of the original SIT included “… the lack of
a clearly stated theoretical rationale, the use of an inadequate
standardization sample, too little information on reliability and
validity, the difficulty of making comparisons between SIT IQs and
the scores of other tests… and the encouragement by the authors
for use of the instrument by untrained examiners…[and] have
been addressed in this revision, with varying degrees of success”
(Campbell & Ashmore, 1995, p. 116).
However, numerous concerns still remain. The collection of the normative
data is vague. There is no information regarding the percentages
of participants with disabilities or identified as gifted. This
is somewhat disturbing as the SIT-R is often used in determining
IQ for individuals with mental retardation and superior intellectual
ability. Although it is an effective tool in screening for mental
retardation, it does not provide appropriate information to determine
the level of mental retardation. The "Slosson Classification
Chart" for the Total Standard Scores may encourage misuse of
scores. This interpretive chart seems at odds with the cautions
against overinterpretation given elsewhere in the manual (Kamphaus,
1994). The table equates low test scores with various levels of
mental retardation, and may suggest the TSS score as a means to
diagnose or confirm mental retardation.
There is no mention of age differentiation upon which the norms
were constructed. The age range is from 4 years to 18+ years with
no differentiation past 18. Kemphaus (1994) noted that there is
no indication of the smoothing method used or whether or not one
was used to produce reasonable distributions of scores from age
to age. In addition, a rationale for changing the basal and ceiling
levels from seven in the original SIT to ten consecutive correct
or incorrect responses on the SIT-R was not provided and appears
to contribute to a lengthening of administration time.
Regarding construct validity, the SIT-R fails to provide evidence
of factorial validity. There were 1,854 subjects who participated
in the sample, an adequate number to determine factor structure.
The authors should have determined whether the SIT-R was unidimensional
as hypothesized at the time of sampling, as it will be costly and
time consuming to sample another 1,800+ participants to reach reliable
factor conclusions. The standardization sample also over represents
small population centers and under represents large population centers
(Campbell & Ashmore, 1995). No data regarding the number of
individuals per age category was provided.
Although the authors claim to have selected a sample to match the
U.S. population, the SIT-R manual fails to provide the U.S. census
statistics, making comparisons difficult. Also, racial composition
did not closely reflect U.S. population demographics. This has been
attributed to the testing protocol requiring the subject’s
first language to be English. In addition, the norming methods are
poorly described and sample characteristics are not broken down
by age category. Given the verbal nature of the SIT-R items and
under
representation of minorities, the SIT-R may be less useful in an
increasingly multicultural society in which English may not be an
individual's primary language, as well as for the assessment of
less verbal and undereducated individuals (Kamphaus, 1994).
Watson (1994) pointed out that major difficulties continue to exist
in understanding the SIT-R’s reliability and validity because
of small sample sizes and restricted populations. Regarding test-retest
reliability, a sample size of 41 children is insufficient for determining
reliability, particularly given that there was no indication of
the demographic composition of the sample. Thus, additional data
must be gathered before the stability of the SIT-R can be evaluated.
Also, while it was helpful to provide concurrent validity coefficients
with the WISC-R (Wechsler, 1974), the lack of coefficients with
other commonly used diagnostic IQ tests, including the SBIS–4
(Thorndike et al., 1986) and K-ABC (Kaufman & Kaufman, 1983)
was noticeable.
REFERENCES
Campbell, C. A., & Ashmore, R. J. (1995). Test review: The Slosson
Intelligence Test-Revised (SIT-R). Measurement and Evaluation in Counseling
and Development, 28, 116-118.
Clark, P., McCallum, R. S., Edwards, R. P., & Hildman, L. K. (1987).
Use of the Slosson Intelligence Test in screening of gifted children.
Journal of School Psychology, 25,189-192.
Harris, K. R., & Reid, R. (1991). A critical review of the Slosson
Intelligence Test. Learning Disabilities Research and Practice, 6, 188-191
Kamphaus, R. W. (1994). Review of the Slosson Intelligence Test-Revised.
In J. C. Coneley (Ed.) The eleventh mental measurements yearbook (pp.
954-956). Lincoln, NE: University of Nebraska Press.
Karnes, F. A., Whorton, J. E., & Currie, B. B. (1986). Correlations
of scores on the WISC-R, the S-B, the Slosson Intelligence Test, and the
Developing Cognitive Abilities Test for intellectually gifted youth. Psychological
Reports, 58, 887-889.
Kaufman, A. S., & Kaufman, N. L. (1983). K-ABC: Kaufman Assessment
Battery for Children. Circle Pines, MN: American Guidance Service.
Kunen, S., Overstreet, S., & Salles, C. (1996). Concurrent validity
study of the Slosson Intelligence Test-Revised. Mental Retardation, 34,
380-386.
McCormick, P. K., Campbell, J. W., Pasnak, R., & Perry, P. (1990).
Instruction of Piagetian concepts for children with mental retardation.
Mental Retardation, 28, 359-366.
McCuller, G. L., Salzberg, C. L., & Linnuris Kraft, B. (1987). Producing
generalized job initiative in severely mentally retarded sheltered workshops.
Journal of Applied Behavioral Analysis, 20, 413-420.
Nicholson, C. L., & Hibpshman, T. H. (1990). Administration and scoring
manual for the Slosson Intelligence Test- Revised. East Aurora, NY: Slosson
Educational Publications.
Nicholson, C. L., & Hibpshman, T. H., (1998). Technical manual for
the Slosson Intelligence Test- Revised. East Aurora, NY: Slosson Educational
Publications.
Salvia, J., & Ysseldyke, J. E. (1995). Assessment (6th ed.). Boston:
Houghton Mifflin.
Slosson, R. (1963). The Slosson Intelligence Test. East Aurora, NY: Slosson
Educational Publications.
Thorndike, R. L., Hagan, E., & Sattler, J. (1986). Manual for the
Stanford-Binet Intelligence Scale - Fourth Edition. Chicago, IL: Riverside.
Watson, T. S. (1994). Review of the Slosson Intelligence Test-Revised.
In J. C. Coneley (Ed.) The eleventh mental measurements yearbook, (pp.956-958).
Lincoln, NE:
University of Nebraska Press.
Wechsler, D. (1974). The Wechsler Intelligence Scale for Children - Revised.
San Antonio, TX: The Psychological Corporation.
Wechsler, D. (1991). The Wechsler Intelligence Scale for Children - Third
Edition. San Antonio, TX: The Psychological Corporation.
|
