Test uses

Screening and identification

Clark (1989) described two different ways of using the CDAT. The original scoring method resulted in quantitative results expected of gifted and talented identification procedures. For initial screening, however, he did not score the responses, but made holistic judgments responses into ability levels. This sorting system, evolving after the original development of the test, broadly distinguished between those responses evincing an aptitude for advanced art instruction and those lacking evidence of such aptitude.


Clark (1989) describes the sorting process as follows:

Completed test booklets are scanned and simply sorted into poor, average, and excellent categories. The central or average group is then resorted into weak, good, and very good categories. Results of this two-step system are five stacks of test booklets in graded levels of competence (poor, weak, good, very good, and excellent). Screening decisions might include only excellent subjects or may include both very good and excellent subjects. The critical condition for determining a cut-off point would be how many people might best be served by a program and how many are needed for a pool of possible participants. (p. 103)

Clark later used a Grading Scale, provided to me by Enid Zimmerman (personal communication, April 10, 2019). Judges used a scale of scores 0 to 10 with descriptors for each of four screening levels. This scale would be applied in holistic judgments made about each of the four CDAT drawing tasks, resembling Hurwitz’s (1983) general recommendations for judging children’s responses as part of a talent identification process.

9-10 Clearly high ability, high potential; Well above grade levelThese students are demonstrating clearly superior representation abilities, perceptual abilities, and spatial abilities for their age. Their work clearly is well-above grade level expectations and demonstrated a great deal of maturity in terms of visual presentation of ideas. Their instructional needs would best be met with individualized instruction, specifically designed to improve their skills.
6-7-8 High ability, high potential; Above grade levelThese students are drawing above grade level. They work well with representation, perceptual, and spatial tasks. Their work is above grade level and demonstrates strong abilities in terms of drawing skills. Their instructional needs would best be met with individualized instruction, specifically designed to improve their skills.
3-4-5 Medium ability, medium potential; At grade levelThese students are drawing with median grade level abilities. Their delineation, perceptual, and spatial abilities are at grade-level expectations, but can be further developed with directed art instruction. Group lessons and practice are required, specifically to improve their representation, perception, and delineation abilities.
0-1-2 Immature ability, immature potential; Below grade levelThese students demonstrate a lack of previous instruction and/or practice with drawing tasks. Their abilities remain relatively below grade-level, due to lack of exposure to–or practice with–visual tasks and concepts. They probably are capable of great improvement, but need a skillful, directive teacher and a great deal of practice and study in visual representation tasks.


Clark (1989) found very high inter-rater reliability in his analysis of scores that he and two of his art education students produced. Even so, he did not recommend relying on the results of the CDAT in isolation to identify and label students in need of advanced art instruction. He agreed with the standard rationale and recommendation to consider a variety of qualitative and quantitative measures. Clark’s list of other measures included assessments of motivation, perception, and aesthetic sensibility, a portfolio of work, and nominations by self, parent, teacher, peer, or art professional.

In their Project ARTS three-year project with upper elementary students, Clark & Zimmerman’s (2001) methodology allowed organizers in each of three different rural, multicultural locations to develop their own local identification plan, considering diverse forms of relevant data. This fit with their recommendation that “definitions used in such programs should be broad and open-ended because it is important to be expansive when seeking or identifying high-ability performance levels in rural schools without previous programs for artistically talented students” (p. 104).

Clark (1989, 1993) denied requests to specify cutoff scores for identification of art giftedness. He recommended instead that a predetermined percentage of high performers on the CDAT be admitted to a specific program based on the needs of the program. Clark and Zimmerman (2001) elaborated further in their report on Project ARTS: “Such cut-off scores would not be appropriate to use in most rural schools, particularly those serving economically disadvantaged or ethnically diverse students; few students would be accepted into art talent development programs on the basis of these types of test scores.” (p. 104).

To put the requests in context, what is now known as The Marland Report (1972) had somewhat recently established a new national definition of giftedness and gifted education in schools that included specific giftedness or talent in the arts. The report expanded upon an earlier definition involving only general intellectual giftedness and gifted education. The request for a cut-off score in measures of art ability may come out of the tradition of relying on cut-off scores of standardized measure of IQ for identification of intellectual giftedness. Clark showed no intention of developing and publishing a norm set larger than the sample provided by each local program in order to accommodate such a request.

Clark and Zimmerman (2001) recommended the following principles for identification of students gifted in art, including:

  • Consider a child’s potential and work in progress as well as final products;
  • Consider process portfolios, work samples, and biographical inventories as alternatives to standardized test scores;
  • Consider factors of background, personality, cultural values, and age;
  • Use “diverse multiple-criteria systems in all identification programs” (p. 105).

Research uses

Clark (1989) expected the CDAT would meet the need for an instrument to screen for high potential and identify high-ability visual artists in specific art talent programs across the country. Clark (1993) also recommended the CDAT as a diagnostic tool for identifying the instructional needs for learners of all abilities. As such, he recommended a pretest to identify remedial or advanced instructional needs and a post-test to show evidence of achievement of drawing competencies.

Clark (1989) suggested also that further research could be conducted using the CDAT in order to more precisely explore the distribution of drawing ability across the population, the relationship of drawing ability to previous opportunities to learn, and the types of learning opportunities that contribute to high achievement on the drawing tasks. Studies have used the instrument for collecting quantitative data (Chan & Chan, 2007, Stewart, 1999).