College of William and Mary Center for Gifted Education
This article by Joyce VanTassel-Baska covers the problems schools face in trying to identify gifted students. Some of the issues are discussed and best practices, according to research, are explained. The author advocates above-level assessments to avoid the ceiling effect and provide better differentiation among those identified as gifted.
The identification of gifted children has long been a topic of great debate in the field of gifted education. More citations in the literature exist on this topic than on any other in the field. Moreover, it remains one of the most common problems of program development cited by school district personnel and state department coordinators in administering programs and services to gifted children.
There are many reasons for the intractable problems associated with identification of the gifted. One of them is related to the concept of absolute versus relative notions of giftedness. Newer definitional structures are attuned to the idea of relativity as we consider the context of the school, the nature of the student's background, and the demands of the program in order to make decisions about individual learners. A second issue that continues to be problematic is recognition of the range of individual differences within the group of learners who might be designated "gifted." We tend to spend a great deal of time deciding who is the last student in the program versus the first student not recommended. Cutting on a continuum of human ability is a risky venture and one many times difficult to justify. At the same time that such debates on identification rage, highly gifted students frequently idle without extensive and intensive enough services because programs are far more likely to focus resources on the mildly gifted group which may be larger and demand more attention. Finally, there is the nagging concern that underrepresented groups are not adequately being assessed to be included in gifted programs. Thus we make the test the proverbial messenger to be attacked and continue to search for a better instrument that may reveal greater parity in performance.
Any one of these issues would be sufficient to keep identification at the top of concerns for local school districts in planning and implementing programs. The three taken together guarantee that identification will always be a controversial topic.
Until our beliefs about identification change, little progress can be made in developing a better system that resolves all of the issues noted. Our task is not to identify only the truly gifted but also to locate students who demonstrate undeveloped potential intellectually and in specific areas including academic, artistic, and leadership domains. Our task is not to select students for all time but to select them for enhanced instructional opportunities that may benefit them at a given stage of development. Whether the intervention works or not, students should be regularly reassessed for new opportunities and dropped from those that are not meeting their needs. Our task is not to be gatekeepers to exclude students but rather custodians of promoting student growth by recognizing discernible strengths and working with the school community to enhance them whether through the gifted program or another medium. Establishing numerical cutoffs on relevant criteria may be less useful than gaining a holistic assessment of students being considered and matching program to strengths of a particular population.
What do we currently understand about the act of identification that may help us deal with the difficulties inherent in the process? First of all, many studies and authors favoring newer conceptual definitions of giftedness acknowledge the multidimensionality of the phenomenon (Gardner, 1991; Sternberg, 1985). Some students are omnibus gifted, capable across many domains and areas. Yet the majority of gifted students are not. They have distinct profiles of strengths and relative weaknesses. Their abilities may be discerned by performance and not paper and pencil tests. Their giftedness may not be evoked by the school environment but shine in the context of community. Some may experience developmental spurts at key stages of development which could not be discerned earlier. Interest may be piqued at some stage that motivates a student to develop abilities in relevant areas. In all of these examples, there is a clear sense that giftedness may be elusive in its manner and context of manifestation.
We also know that there are both genetic and environmental factors at work in the manifestation of giftedness. Individuals vary considerably in their ability to function effectively in various domains. Attention must be paid to the "rubber band" effect of human potential -- our genetic markers allow for expansive growth and development but not to an unlimited extent. We can stretch ourselves within a range based on the genetic potential which we possess. It is the role of education in the larger environment to provide the experiences which may stretch the individual potential in the areas of greatest flexibility for learning. This recognition of pre-existing individual differences would help educators realize the folly of trying to find a "one size fits all" program of study or curriculum. As long as differentiated practices are reserved for labeled special populations, the spirit of individualized learning will always be in jeopardy. Giftedness does not guarantee entitlement to educational privilege, but it does call for a flexible response by schools and other agencies to higher levels of functioning, based on the individual level of functioning not age.
The concept of degree or extent of giftedness is an important aspect to consider in developing identification processes. When I directed the talent search program at Northwestern University, I would have teachers tell me that seventh grade students who were scoring at the 600 level in mathematics on the Scholastic Aptitude Test (SAT) weren't truly precocious in mathematics, even though their scores placed them in the top 2 % of the population. Only the 700's met that criterion. What these teachers were noting is the wide band of difference that exists within a gifted population such that students at the bottom of a particular group may function very differently from those at the top of the group. In psychometric language, this means that gifted students may vary among themselves by as much as three standard deviations in respect to mental functioning in one or more areas. Reading level, for example, in a fifth grade gifted program could range from seventh to college level. The implications of this phenomenon for identification is to decide how broad a group might benefit from a particular intervention and then ensure differentiation of instruction in the delivery of that intervention to ensure adequate challenge for those at the top of the group and yet not cause anxiety to set in for those at the bottom. Wide ranges of abilities within a gifted population have to be tolerated in most gifted programs since the context of delivery frequently requires sufficient numbers of students to justify the special intervention.
We also know that the recognition of advanced behavior is the most critical variable in determining who can best profit from advanced work and instruction. To deny services to students clearly advanced in reading, mathematics, the arts, or other domains because they have not been formally assessed calls into question a school system's capacity to respond to individual differences. This principle of responding to advanced student behaviors is central to including teacher, parent, and community input into the identification process. Use of domain-specific checklists is one way to assess such behavior in context. Such checklists also contribute important insights into effective programming for individual children.
Work in talent development (e.g. Csikszentmihalyi, 1996; Simonton, 1999) has convinced most people in our field that ability alone may be insufficient to predict success in gifted programs, let alone life endeavors. Non-intellectual factors like motivation, personality, persistence, and concentration impact greatly on creative productivity at particular stages of development but also over the lifespan. Thus our identification processes may need to be sensitive to students whose ability threshold may be slightly lower but whose capacity and zeal to do work in a given domain may be very high. Tapping into these non-intellectual strengths can best be accomplished through performance and portfolio-based assessment protocols coupled with careful observation of performance over time.
What are best practices for identification based on research? Currently there is a call for a new paradigm for identification, in line with the new constructs of giftedness that have been conceptualized (Passow & Frasier, 1996). This new paradigm of identification would recognize the different ways in which students display giftedness and would call for more varied and authentic assessment. Instead of relying on intelligence and achievement test scores solely for identification, multiple criteria would be used, including more non-traditional measures such as observing students interacting with a variety of learning opportunities (Passow & Frasier, 1996) it is a belief of many in the field of gifted education that new conceptions of giftedness and a new paradigm for identifying and selecting students will help minority and disadvantaged students become more represented in gifted programs (VanTassel-Baska, Patton, & Prillaman, 1991; Ford, 1996).
Part of the process of non-traditional assessment involves trying to tap into fluid rather than crystallized abilities. Dynamic assessment is one such non-traditional approach used to assess cognitive abilities that are often not apparent when most forms of standardized tests are used. This type of assessment usually consists of a test-intervention-retest format, with the focus being on the improvement students make after an intervention, based on learning cognitive strategies related to mastery of the testing task (Kirschenbaum, 1998).
Research evidence also suggests that disadvantaged learners perform better on tasks that emphasize fluid over crystallized intelligence (Mills & Tissot, 1995), and spatial reasoning over verbal and mathematical (Naglieri, 1999). By employing an assessment approach that contains a strong spatial component, disparities between scores by socio-economic status (SES) levels or ethnic group may be reduced (B. Bracken, presentation at College of William and Mary, April, 1999). Thus using instruments like the Matrix Analysis Test and the Ravens Matrices may yield somewhat different populations of students than the use of traditional intelligence tests that emphasize verbal tasks. The new UNIT test also offers promise in this regard as a full scale measure.
There is also a need to employ a two-stage process of screening and identification to ensure that appropriate measures are used in the selection of students for a program. It is not highly defensible to use group achievement and intelligence test score data as the final arbiters for selection by merely raising the cutoff, let's say to 98%. Many times school districts will have large numbers of students who would qualify at 95%. To use a norm-referenced test that is grade-level calibrated to make judgments about students at the top end is not justifiable, given the problems of ceiling effect. A better and more defensible strategy is to use off level aptitude and achievement measures to ascertain a true dispersion of the student scores in order to select the most able. Off level instrumentation like the PLUS test, the SCAT test and the SAT all provide such information so that identification can be more precise. Use of these instruments over the past 25 years has continued to demonstrate effectiveness and efficiency in discerning able students' range of functioning in critical domains (Benbow & Stanley, 1996).
The use of measures that are relevant to program emphasis is also a crucial consideration. Using verbal measures to decide who should be in a math program makes no sense. Ensuring that an identification system is geared to the nature of the program intervention is crucial, especially at the second stage of the process. Thus, if the program emphasis is writing, a writing sample would be included at the identification stage, or if the program emphasis is science, a performance-based science assessment or science project portfolio would be included to make final selections. Such authentic assessment data strengthen the case for selecting the most apt individual students for participation in carefully defined program areas (VanTassel-Baska, 1998)
The use of identification protocols that are appropriate at different stages in the development of students is also a best practice in the field. Early childhood identification procedures, because of age and lack of contact with the school, have to consider parental feedback more carefully, use testing data more judiciously, and consider advanced performance tasks as an important part of the process. At secondary level, based on different organizational contexts, identification procedures need to be distinctive in respect to protocols for finding students in a broader range of talent areas and for considering domain-specific approaches based on departmental courses of study.
Finally, the identification process must be equitable in respect to selection, validation, and placement of students. Making placement decisions based on individual profile data is also considered best practice as it allows professional judgment to be exercised rather than just allowing a numerical cut-off score on a matrix model to determine placement (Borland & Wright, 1994).
Identification will continue to present a challenge to educators of the gifted. Yet thoughtful consideration of and reflection on various problems, issues, and current best practices can make the process more feasible and credible in school contexts.
Benbow, C. P. & Stanley, J. C. (1996). Inequity in equity: How "equity" can lead to inequity for high-potential students. Psychology, Public Policy, and Law, 2,
Borland, J. H., & Wright, L. (1994). Identifying young, potentially gifted, economically disadvantaged students. Gifted Child Quarterly, 38, 164-171.
Csikszentmihaly, M. (1996). Creativity: Flow and the psychology of discovery and invention. NY: Harper Collins
Ford, D. Y. (1996). Reversing underachievement among gifted black students: Promising programs and practices. New York: Teachers College Press.
Gardner, H. (1991). Creating minds. New York: Basic Books
Kirschenbaum, R.J. (1998). Dynamic assessment and its use with underserved gifted and talented. Gifted Child Quarterly, 42, 140-147.
Mills, C. & Tissot, S. (1995). Identifying academic potential in students from underrepresented populations: Is using the Ravens Progressive Matrices a good idea?
Gifted Child Quarterly, 39 (4), 209-217.
Naglieri, J. A. (1991). The essentials of CAS assessment. NY: J. Wiley & Sons.
Passow, A. H., & Frasier, M. M. (1996). Toward improving identification of talent potential among minority and disadvantaged students. Roeper Review, 18,
Simonton, D. K. (1999). Origins of Genius, New York: Oxford University Press.
Sternberg,R. J. (1985). Beyond I.Q. NY: Basic Books
VanTassel-Baska, J. (1998). Excellence in educating the gifted (3rd ed.). Denver, CO: Love
VanTassel-Baska, J., Patton, J., & Prillamon, D. (1991). Gifted youth at risk: A report of a national study. Reston, Virginia: The Council for Exceptional Children.