Handbook of psychology volume 7 educational psychology

Gifted Education Programs and Procedures

495

1986b). The supportive conditions include a domain or field

that is structured in a way and developed to the extent that it

is available and comprehensible to a young child. The child’s

capacity and ways of learning must fit well with knowledge

base of the domain and the forms in which learning and

instruction occur. The talented child must be living in a his-

torical time in which the domain is valued and high-level

mastery of it is prized. Furthermore, the prodigy must come

from a home with a family that recognizes and supports the

ability and can obtain resources to insure its development

(Feldman, 1986b).

According to Feldman (1986b) only some domains that

can be described as developmental produce prodigies. Devel-

opmental domains typically have a long history of knowledge

development and major changes over time in structure, tech-

nology, organization, and practice. They have several dif-

ferent levels of expertise to master in sequence—each level

marked by a different set of skills. Feldman notes that the

largest numbers of child prodigies have been found in chess

and music, whereas many fewer have been found in writing,

the visual arts, or mathematics. Child prodigies in fields such

as physics or the natural sciences are virtually unheard of

(Feldman, 1986b).

Although prodigies display exceptional capacity to master

the levels of a particular field, their tremendously fast learn-

ing rate appears limited to a single domain. “Perhaps the

purest essence of domain-specific talent is the ability to holis-

tically intuit the syntactic core of rules and regularities lying

at the heart of a domain of knowledge . . .” (Morelock &

Feldman, 1993, p. 179). That these children do develop skills

to such a high level in an area is a result of a delicate and rare

interaction between the capabilities and developmental tra-

jectory of the child and of the domain (Feldman, 1986b).

And, despite early and rapid advancement in a field,

Feldman’s prodigies did not necessarily stay with the same

field into adulthood. Most often, they gave up their commit-

ment to that one field and branched out to others.

The Relationship Between Creativity

and Intellectual Giftedness

The field of gifted education has had an ambivalent attitude

toward the concept of creativity. Some, like Renzulli (de-

scribed previously), believe that creativity is the essence of

giftedness—one should not even speak about giftedness

except to mean creative production. Alternatively, creativity is

included as one of several categories of giftedness within the

Marland definition (described previously) equal in status with

other categories. Others believe that creativity is a phenome-

non distinct from, yet larger than intellectual giftedness; cre-

ativity rests on the acquisition of skills and knowledge acquired

via intellectual giftedness but goes beyond it (see Feldman, de-

scribed previously). Others believe that creativity can only be

thought of within the context of a domain and in reference to

adult work. It is “impossible to define creativity independently

of a judgment based on criteria that change from domain to do-

main and across time” (Csikszentmihalyi, 1994, p. 143).

Definitions of creativity can be categorized according to

four different emphases—personality, process, press (situa-

tion), and product (Fishkin, 1999). Work on the creative per-

sonality aims to illuminate the personality traits of creative

producers, usually through retrospective research studies

(examples of such studies are listed earlier in this chapter).

Creative producers have been found to possess personality

characteristics such as risk taking, flexibility, a preference for

disorder, androgyny, the ability to tolerate ambiguity and

delay gratification, and an introspective and introverted

personality (Rogers, 1999).

Other researchers primarily focus their definition on the

process of creativity, attempting to delineate the intellectual

activities involved in creative thought such as use of imagery,

problem-finding, and metacognitive components (Fishkin,

1999). Studies that focus on press are interested in the con-

textual determinants of creative productivity, particularly en-

vironmental factors. Work in this area has looked at aspects

of the home environment such as encouragement of freedom

of thought and risk taking, features of the work place such

as leadership style of a supervisor, temperature, and physical

climate, and other psychological and social variables that

affect creativity (Keller-Mathers & Murdock, 1999).

Research on creative products has focused on determining

how decisions about the uniqueness of products are made

(e.g., criteria used, judgements by whom) within and across

different domains, the relevance of creative production to the

definition of creativity (e.g., can one be creative if one does

not produce?), and the development of rating scales to make

judgments about children’s creative products (Keller-Mathers

& Murdock, 1999).

Researchers who have tried to assess creativity have used

a variety of instruments and techniques depending upon their

theoretical perspective. Creativity measures include tests of

divergent thinking, attitude and interest inventories, person-

ality inventories, biographical analyses, ratings by teachers

or others, judgments of products, criteria of eminence, and

self-reports of creative activities or achievements (Johnson &

Fishkin, 1999). The most extensively used and researched

instruments to identify creative children are standardized

measures of divergent thinking such as the Torrance Tests of

Creative Thinking or several subtests of the Structure of the

Intellect Test. Many assessments of creativity suffer from low

reliability and questionable validity (see Johnson & Fishkin

for a review).

496

Gifted Education Programs and Procedures

An important issue of construct validity for creativity mea-

sures are their distinctiveness or independence from IQ. The

relationship between IQ and divergent thinking measures has

been described as nonlinear, with evidence of a threshold

effect (Johnson & Fishkin, 1999). Across a wide range of IQ

scores, the correlation between intelligence and creativity is

about .40 (Johnson & Fishkin); however, beyond an IQ of

120, the correlation is negligible. The threshold relationship is

interpreted to mean that greater creativity is associated with

higher intelligence up to a certain point only. Beyond an IQ of

120, higher intelligence does not guarantee greater creativity,

and other variables such as motivation and personality dispo-

sition are more influential. The relationship between creativity

and academic achievement is complicated; some studies show

a positive correlation, some show no correlation, and others

indicate that the relationship is mediated by gender, type of

creativity, and type of creativity assessment (Ai, 1999).

There have been many programs developed to advance the

creative thinking skills of children. The assumption behind

these programs is that creative thinking skills are not innate

but instead are acquired through practice and deliberate

teaching. Creative thinking programs include synectics, a

form of analogical reasoning in which thinkers examine con-

nections between seemingly unrelated objects; lateral think-

ing (e.g., the Cognitive Research Trust program or CoRt);

Odyssey of the Mind (OM), a creativity training program in

which teams of children practice skills such as brainstorming,

suspending judgment, and listening to others in order to solve

complex, open-ended problems in a competition; and the

Future Problem Solving program (FPS), which teaches stu-

dents to use creative problem-solving techniques applied to

ill-defined, complex problems about futuristic issues in com-

petition with other teams of students (Meador, Fishkin, &

Hoover, 1999).

Pyryt (1999) reports on several meta-analyses that exam-

ined the effects of various types of creativity training on as-

pects of children’s thinking. The outcome measure in these

studies is typically performance on divergent thinking tests,

mainly the Torrance Test of Creative Thinking. Pyryt found

effects sizes averaging close to one standard deviation across

a variety of groups of students, not just the gifted ones. Effect

size correlated with the amount of training, and different

types of training differentially affected performance on ver-

bal versus figural divergent thinking measures.

An important issue for researchers and educators is the

predictive validity of creativity measures for adult accom-

plishments. Cramond (1994), in a review of research on the

Torrance tests, concludes that there are moderate correlations

between childhood divergent thinking scores and adult

accomplishments. However, Tannenbaum (1983) notes that

Torrance’s studies are the only ones that consistently support

the predictive validity of divergent thinking measures for

adult creative activities.

There is an accumulated body of research about highly

creative, eminent individuals. These studies have sought

to understand the factors that contribute to high levels of

creative production. A major finding is that many creative

producers experienced stressful and traumatic childhoods—

particularly, the early loss of parents (Olszewski-Kubilius,

2000). Stressful childhoods are thought to contribute to the

development of creativity by engendering strong motivations

to express and thereby ameliorate childhood wounds through

some creative outlet and to gain acceptance, love, or admira-

tion from others. Early family environments that stress indi-

vidual thought, expression, and independence, and those in

which there is an emotional distance between parent and

child and less vigilant parent monitoring and socialization of

children create contexts for the development of personality

traits (e.g., risk-taking) that are thought to be essential to the

development of creativity (Subotnik, Olszewski-Kubilius, &

Arnold, in press).

THE EDUCATION OF GIFTED CHILDREN

Identification

The identification of gifted children is a major issue for edu-

cators. It is inextricably tied to one’s definition of and beliefs

about the nature of giftedness and also to the programs and

services that are put into place for children who have been

identified as gifted. These two components serve as bookends

to determine or bound the identification procedures.

For example, if you believe strongly that the aim of iden-

tification is to find children who have the potential to become

creative producers in adulthood (see discussion of Renzulli

earlier in this chapter), identification procedures might in-

clude a focus on demonstration of exceptionally creative

work in school coupled with task persistence and motivation

to produce unusual products at a high level. On the other

hand, if your beliefs about giftedness are that it is exceptional

intellectual ability regardless of actual achievement or per-

formance, measures such as IQ scores could be used for iden-

tification, and you would aim to include children with high

ability yet low school achievement. If you subscribe to an

educational definition of giftedness and believe that gifted

children are those for whom the typical school curriculum is

inadequate, you would use measures of achievement to find

students who are able to perform beyond their current school

placement in the subjects typically taught in schools. If you

The Education of Gifted Children

497

subscribe to a multiple intelligences view of ability, you

would want to establish identification procedures to find chil-

dren with talent in the various domains and provide programs

to help them develop that talent.

Programming or services affect identification procedures.

If you believe that your goal is to prepare children to become

creative producers in adulthood, you would build a program

that gives them many opportunities to practice making or

generating creative products. If you believe that your goal

is to ensure that every child in school experiences academic

challenges, you would design classes that through pacing or

rigor of the content provide challenge to gifted learners.

In reality, identification procedures for gifted children are

often put into place without a clear rationale or understanding

of the beliefs about giftedness and its development upon

which they rest. Moreover, there is often a mismatch between

identification procedures and programming. A frequent and

often valid criticism of gifted programs is that they often

have very selective identification procedures that include

high IQ and achievement scores, but the enrichment pro-

grams provide general enrichment appropriate for which

most students could be successful.

Identification of Academic Talent

A national survey of school gifted programs conducted by the

Richardson Foundation in the mid-1980s showed that teacher

nomination was the most frequently used means to identify

academically gifted students (91%), followed by achieve-

ment tests (90%), and IQ tests (82%; Cox et al., 1985). All

other types of criteria—grades, peer or self-nomination, and

creativity measures—were used significantly less often by

schools (less than 10% in most cases).

Although it is frequently used, teacher nomination is not

highly regarded as a method of identifying gifted students

(Tannenbaum, 1983). Teachers tend to nominate children who

are high achievers, polite, and well-behaved; sometimes use

placement in a gifted program as a reward for high achieve-

ment; and often fail to identify underachieving gifted children

(Borland, 1989). One of the problems in assessing the accuracy

of teacher nomination is the lack of precise standards against

which their judgments can be compared. Teachers improve in

identifying children who turn out to be gifted as defined by IQ

scores when they receive training about the behavioral charac-

teristics of gifted children (Tannenbaum, 1983).

IQ tests, which are probably best viewed as predictors of ap-

titude for academic achievement, are relevant to educationally

oriented definitions of giftedness. One of the problems with IQ

tests is that the information is not very useful for programming

planning because it does not relate well to the content areas

typically included in the school curriculum. Furthermore, IQ

tests are limited in their prediction of adult success as well.

However, IQ tests may reveal intellectual ability among under-

achieving students (Subotnik & Arnold, 1994).

Achievement tests and domain-specific aptitude tests do

map onto the school curriculum better than do IQ tests and

can indicate students’ mastery of the in-grade curriculum.

Due to severe ceiling problems, however, these tests do not

reveal what students know beyond the information assessed

on the test, which may be considerable. Sometimes, achieve-

ment tests are used off-level, so to speak, or tests (or test

forms) designed for older students are given to younger ones,

thereby eliminating the ceiling problems. However, it is

also not clear to what extent achievement or special aptitude

test performance predicts adult success in different fields

(Tannenbaum, 1983). A substantial amount of evidence exists

about this issue for mathematics; high mathematics scores on

the Scholastic Achievement Test (SAT) in the seventh or

eighth grade are associated with high academic achievement

in high school and college and with the choice of mathemat-

ics as a career (Lubinski & Benbow, 1993).

Checklists for teachers and parents are also frequently used

as part of an identification system. Very often, these are in-

cluded primarily for political reasons—that is, to include the

opinions of a particular group in the process. There are some

published instruments available, including the Scales for

Rating the Behavioral Characteristics of Superior Students

(Renzulli, Smith, White, Callahan, & Hartman, 1976), which

assesses abilities in 10 areas: learning, motivation, leader-

ship, artistic, musical, dramatics, communication-precision,

communication-expressive, planning, and creativity. Also

available are the Purdue Academic Rating Scales, designed to

identify secondary students for honors, advanced placement

(AP), or accelerated classes: It focuses on signs of superior

academic performance in mathematics, science, English, so-

cial studies, and foreign languages (Feldhusen, Hoover, &

Saylor, 1990). See Johnson and Fishkin (1990) for a com-

pendium of rating scales for creativity. Most often, schools

devise their own questionnaires; hence, there is typically little

or no validity or reliability information available on them.

Schools generally use multiple pieces of information

about a student to make a decision about placement into

a gifted program and employ some means or system of

organizing and evaluating the data. Borland (1989) and others

(Feldhusen & Jarwan, 1993) recommend that identification

be viewed as a process consisting of several distinct phases.

These phases include screening, which involves using avail-

able information about students to nominate them as potential

candidates for the gifted program—candidates who need fur-

ther scrutiny or examination to determine their match to the

498

Gifted Education Programs and Procedures

program. Sources of information for screening include exist-

ing student records, group tests, referrals, and nominations.

At the screening phase, the goal is to cast a wide net, so to

speak, using generous cutoffs so in order to include as many

students as possible.

The second phase is selection or placement. This phase

involves the gathering of more specific and reliable infor-

mation, usually through additional testing (Borland, 1989).

Tests to be used for selection or placement should be selected

on the basis of their relevance to the educational program,

their reliability and validity, the availability of normative

data, and the extent to which they are free from ceiling effects

and test bias (Feldhusen & Jarwan, 1993).

Placement decisions are usually based on the results of

synthesizing the data in some fashion and typically involve

one of the following mechanisms: constructing a matrix that

involves assigning points to scores on different tests and

summing them to yield a single giftedness index; computing

standard scores for each measure in order to ensure compara-

bility across measures; a comprehensive case study on each

nominated student; or a multiple-cutoff method, which

requires students to meet minimum score cutoffs for each mea-

sure included in the identification system (Feldhusen &

Jarwan). More sophisticated methods such as multiple regres-

sion models that determine a formula based on the predictive

validity of each identification instrument for program perfor-

mance are rarely used (Feldhusen & Jarwan). Domain-specific

measures may be used in areas such as the visual or performing

arts and include portfolio assessment, product assessment or

critique, auditions, and so on.

The third phase of the identification process that is rarely

implemented is the evaluation of the viability of the identifica-

tion protocol (Borland, 1989). This involves an analysis of

students’ performance in relation to the identification criteria

and can involve a possible reformulation of the identification

protocol. The questions being addressed at this phase of the

identification process are Are the students who were selected

for the program succeeding in it?, Are the selection procedures

excluding children who also could succeed in the program?,

and Are the children in the program achieving at expected high

levels over the long term? Methods to assess the appropriate-

ness of the selection protocol include the use of multiple re-

gression to determine the extent to which identification criteria

predict achievement in the program or the comparison of the

achievement of children who were selected for the program to

those who were nominated for the program but not selected.

Schools infrequently evaluate their identification protocols.

In summary, typical problems with identification systems

include the following: a lack of consistency between the

philosophy of the program and the identification protocol; the

overreliance on a single measure or instrument for identifica-

tion; a mismatch between the identification protocol and the

program being provided—most typically requiring high

scores on measures of verbal and quantitative ability and pro-

viding a program in only one area or using very high cutoff

scores on IQ or achievement tests and providing a general en-

richment program suitable for mildly above-average learners;

lack of evaluation of identification criteria to assess their pre-

dictive validity for program performance or outcomes; prob-

lems with instruments used, such as low reliability or validity

and ceiling effects; use of some kind of summative score

based on different selection criteria that has dubious validity;

and the lack of procedures for periodic reassessment of

students, both those in and those not in the program.

Recommended identification procedures include the use

of multiple screening measures in an effort to be very inclu-

sive and capture nontraditional students and underachievers,

identification protocols tailored to the particular school do-

main (e.g. using math tests for an accelerated math program),

placement decisions made by a committee using all available

information, and off-level testing to deal with ceiling effects

on on-level tests. See Feldhusen and Jarwan (1993) and

Borland (1989) for a fuller discussion of recommended iden-

tification procedures for gifted students.

Identification of Underrepresented Students

A major issue within the field of gifted education is the

typical underrepresentation of children of color within

gifted programs or advanced classes (Olszewski-Kubilius &

Laubscher, 1996) regardless of socioeconomic level (Ford,

1996). Reasons for this situation include lower performance

on typically used identification instruments by minority stu-

dents due to environmental factors such as fewer educational

opportunities and qualitatively poorer educational environ-

ments, instrumentation problems such as cultural bias of tests

and their lack of predictive validity for academic achieve-

ment of minority students, and the prejudices of teachers and

other educational personnel making decisions about students.

Ford (1996) recommends using alternative identification pro-

cedures that include nonverbal problemsolving tests (e.g., the

Ravens Progressive Matrices tests or the Naglieri Nonverbal

Ability Test), surveys and instruments specifically designed

to identify giftedness among minority students, and train-

ing for educators regarding cultural sensitivity. Although the

research evidence suggests that students of color are more

likely to be identified as gifted when alternatives are used,

they are still much less likely to be referred or nominated for