|
Details
The DISCOVER assessment
was developed and refined over a 13-year period,
supported by the Office of Bilingual Education
and Minority Languages Affairs and the Javits
Gifted and Talented Education Program. It has
been used with varied multicultural populations
in the United States and abroad, and with students
from varied economic levels. One well-established
aspect of its validity is that when used to
assess abilities, the percentage of students
who receive the highest ratings is similar across
ethnic, racial, language, and economic groups
(Nielson, 1994; Maker, 1997; Sarouphim, 1999).
Shonebaum (1997) also found that deaf students
can be assessed effectively with DISCOVER, and
that the only modification needed is to videotape
rather than audiotape students' storytelling.
Reliability
Expert DISCOVER observers
agreed more frequently and consistently (r =.92
to 1.00) than novice DISCOVER observers (.47
to .92). However, across experience levels,
observers agreed 95% of the time on the highest
rating of problem solving, the category used
for placement in special programs (Griffiths,
1996). Recent studies showed that agreement
between DISCOVER personnel with varying levels
of experience was 81% overall and 100% on the
highest rating. Agreements between DISCOVER
personnel and school district teams was 85%,
and was 82% among the members of the district's
team (Kassymov, April, 2000).
Validity
In a concurrent validity
study, Sarouphim (1999a) found that DISCOVER
assessment results for spatial, logical-mathematical,
and linguistic abilities were consistent with
independent ratings made by both classroom teachers
and a researcher who observed students in the
classroom. Lori (1997) found a significant correlation
between storytelling ability and personal intelligences
in Bahraini children. Sarouphim (1999b) also
studied the relationship between scores on the
Raven Progressive Matrices in 257 kindergarten,
second, fourth, and fifth grade students. In
this group of predominately Navajo and Mexican
American students, she found the highest correlations
with spatial artistic (r = .58, p<.01), spatial
analytical (r = .39, p<.01) and math (r =
.35, p.<01) as one would expect, since the
Raven is a measure of non-verbal logical reasoning.
Lower correlations appeared between Raven scores
and oral (r = .20, ns) and written linguistic
(r = .093, ns) DISCOVER ratings.
In a study of 55 gifted
private school Caucasian students (without correcting
for restricted range of IQ) Stevens (2000) found
significant correlations between DISCOVER spatial
artistic (r = .373, p.<.01) and written linguistic
(r = .34, p<.05) activities and Full-Scale
IQ. He found significant correlations between
spatial artistic (r = .27, p<.05) and written
linguistic (r = .388, p<.01) activities and
verbal IQ. The only significant correlation
with performance IQ was the spatial artistic
activity (r = .369, p<.01). Griffiths (1997),
however, did not find significant relationships
between DISCOVER and WISC III or WPPSI verbal
or performance IQ or full-scale IQ in her study
of 34 Mexican American kindergarten students
identified as gifted using DISCOVER. In this
case, IQ scores ranged from 88 to 137 with a
mean of 115.
A recent analysis
showed that the spatial analytical activity
(solving tangram puzzles of increasing difficulty)
correlated with total IQ (r = .272, p<.05)
measured by the Otis-Lennon School Abilities
Test in second-grade students (N=72). For male
students (N=37) the relationship was stronger
(r = .441, p<.05), but for female students
(N=35), the only significant relationship was
between DISCOVER math and non-verbal IQ (r =.454,
p<.05). When scores were analyzed separately
by ethnicity, no significant relationships were
found for Hispanics (N=24), or African American
(N=18) students. However, for Caucasian students
(N=24), significant relationships were found
between the spatial analytical assessment and
total IQ (r =.477, p<.05), between spatial
artistic and non-verbal IQ (r =.719, p<.05).
In other internal analyses, we have found no
significant relationships between the DISCOVER
assessment activities and a cross-cultural non-verbal
test of creativity, the Test of Creative Thinking-Drawing
Production (TCT-DP) (Urban, 1996) in any cultural
or age group studied. A possible reason for
the lack of correlation is that the TCT-DP is
an assessment of divergent thinking in a decontextualized
situation.
|
|
References
Almegta. (1997). Correlation of
standardized math achievement, DISCOVER math
worksheet, ability, and gender of American Indian
students.
Unpublished Manuscript.
The University of Arizona, Tucson.
Griffiths,
S.E. (1996). The inter-observer reliability
of the DISCOVER problem-solving assessment.
Unpublished manuscript. The University
of Arizona, Tucson.
Griffiths, S. E. (1997). The comparitive
validity of assessments based on different theories
for the purpose of identifying gifted ethnic
minority students. Unpublished doctoral dissertation.
The University of Arizona, Tucson.
Kassamov, A. (2000). Researchers
and implementers: Comparison of inter-rater
reliability. Manuscript in preparation.
The University of Arizona DISCOVER Project.
Lori, A.A. (1997). Storytelling and personal
traits: Investigating the relationship between
children’s storytelling ability and their
interpersonal and intrapersonal traits.
Gifted
Education International.
13(1) 57- 66.
Maker, C.J. (1997). DISCOVER Problem-Solving
Assessment.
Quest, 8(1), 3, 5, 7, 9.
Nielson, A. (1994).
Traditional identification: Elitist,
racist, sexist?
New Evidence. CAG Communicator,
24(3), 18-19, 26-31.
Shonebaum, J.A. (1997). Assessing
the Multiple Intelligences of Children who are
Deaf with the DISCOVER process and the use of
American Sign Language.
Unpublished masters thesis.
The University of Arizona, Tucson.
Sarouphim, K. (2000).
Internal structure of
DISCOVER: A performance-based assessment.
Journal for the Education of the Gifted,
23(3), 314-327.
Sarouphim, K. (1999a). Discovering multiple
intelligences through a performance-based assessment:
Consistency with independent ratings.
Exceptional Children, 65(2), pp.151-161.
Sarouphim, K. (1999b).
DISCOVER:
A promising alternative assessment for
the identification of gifted minorities.
Gifted Child Quarterly,
43(4), 244-251.
Romanoff, B.S. (1999). Achievement
of African American students and White students:
A comparative study of placement in the program
for the gifted.
Unpublished doctoral dissertation.
The University of Arizona, Tucson.
Stevens, B. (2000). Relationships
between the DISCOVER assessment and Wechsler
Intelligence Scales as identifiers of gifted
children. Unpublished masters thesis. The
University of Arizona, Tucson.
Urban, K. (1996). Test of Creative
Thinking – Drawing Production, Swets &
Zeitlinger, Inc.
|