The work reported here addresses many of the questions raised in the previous paragraphs: some directly, some indirectly. As in other arts fields, teacher's beliefs have made it difficult to resolve questions about definition of high ability in the arts and about who may or may not be considered talented. Some art teachers believe, for instance, that all judgments about art, including assessment of students' art abilities, must be subjective and idiosyncratic while others believe that all forms of testing or assessment in the arts must be unreliable and subject to question (Sloan & Sullivan 1982). These teachers believe that art expression is inherently idiosyncratic and, therefore, cannot be judged in ways that will be agreed upon by other art teachers or students.
Another condition in visual arts education that has mitigated against clear definition of talented students is an almost total lack of standardized measurement instruments. There are available art tests that have been assessed and, in some instances, recommended in the gifted or talented literature, including the Meier (1929), Knauber (1932), and Horn (1935) tests (Khatena 1882). It has been shown, however, that these are not used in any identification programs for artistically talented students anywhere in the country and were not considered indicators of superior art abilities (Clark & Zimmerman 1987). These conclusions are based upon the dated nature of these older tests and upon the fact that they were never intended as diagnostic of superior abilities in the arts.
The five most commonly used identification procedures in programs for talented students in the visual arts, and the frequency of their use (shown in percentages), have been (1) self nomination 51%; (2) portfolio review 47%; (3) interview 27%; (4) informal art tests 25%; and (5) classroom teacher nomination 25%. These frequencies add to more than 100% because most of these practices are used in combination with one another or others. All other