The Reynolds Intellectual Assessment Scales (RIAS) and Assessment of Intellectual Giftedness

Home , Intellectual giftedness, Verbal intelligence

Amber E. Brueggemann University of Georgia Cecil R. Reynolds Texas A & M University R.W. Kamphaus University of Georgia the reynolds intellectual assessment scales (RIAS) and assessment of intellectual giftedness

Abstract

The Reynolds Intellectual Assessment Scales (RIAS) (Reynolds & Kamphaus, 2003) is a psychometrically sound, individually administered test of intelligence developed and standardized for ages 3 through 94 years. This article describes the goals for development of the RIAS and its underlying theory, emphasizing its applicability to the identification of intellectually gifted individuals. In addition, an overview of the test’s administration and psychometric properties is provided.

The Reynolds Intellectual Assessment Scales that is not as strongly associated with (RIAS; Reynolds & Kamphaus, 2003) is a general intelligence (defined more strongly recently published, rapidly administered as representing problem solving skill on the (25-35 minutes for most school-aged RIAS). The memory scores do not contribute individuals), individually administered test to the calculation of the various intelligence of intelligence. Developed and standardized indexes on the RIAS.1 for ages 3 years to 94 years, the RIAS yields three intelligence scores, a Verbal Intelligence Index (VIX), a Nonverbal Theory and Goals Intelligence Index (NIX), and a global Composite Intelligence Index (CIX), derived The RIAS is based on the consensus findings from the VIX and NIX. Co-normed verbal of Cattell and Horn’s (Horn & Cattell, 1966) and nonverbal memory subtests are theory of fluid and crystallized abilities and included as well but are considered to Carroll’s (1993) three-stratum theory of represent a related but separate construct intelligence (Reynolds & Kamphaus, 2005).

Email: [email protected]. Volume 21 No 2/3, 2006, 127 While the indexes are not explicitly divided measurement of g and its two primary determining access to services despite the focuses in order to gain a better measure of g. into fluid and crystallized domains, components, verbal and nonverbal differences in their definitions (McCoach et In the case of gifted identification, gauging combinations of the six RIAS subtests were intelligence, with close correspondence al, 2001). verbal abilities not only explains a greater designed to measure four aspects of to crystallized and fluid intelligence. proportion of variation in intelligence, or g, intelligence, which are related to these Goal 2: Provide a practical measurement According to Reis and Small (2005), but also assesses a central characteristic of concepts. The first, general intelligence, or g, device in terms of efficacies of time, intellectually gifted students display several giftedness, and the component that is most is assessed by the Composite Intelligence direct costs, and information needed cognitive characteristics that make them highly predictive of success in academic and Index (CIX) and corresponds to Carroll’s from a measure of intelligence. unique from their peers. These traits include related learning environments. (1993) stratum three, which is best Goal 3: Allow continuity of measurement advanced verbal aptitude, the ability to represented by fluid abilities. Verbal across all developmental levels from ages understand complex and abstract concepts, Not only does the RIAS mainly assess the intelligence is closely related to the concept 3 years through 94 years for both clinical and advanced thinking, processing, and primary traits of intellectual giftedness, of of crystallized ability, although not identical. and research purposes. problem-solving skills. Since these are the equal importance is the fact that the RIAS Nonverbal intelligence, also referred to as Goal 4: Substantially reduce or eliminate traits that characterize gifted individuals, it does not measure domains that are not visualization or spatial abilities by some dependence on motor coordination and reasonably follows that the intelligence associated with typical conceptions of the theories, is similar to fluid intelligence. In visual-motor speed in the measurement measure used for gifted identification should intellectually gifted. As stated in the fourth addition, memory, or learning, may be of intelligence. assess problem-solving and higher-order goal, the RIAS was specifically developed so assessed by the two supplemental subtests. Goal 5: Eliminate dependence on reading in skills commonly thought of as general that its measurement of intelligence would the measurement of intelligence. intelligence. The RIAS emphasizes problem- not be confounded with motor coordination Most of the variance measured by Goal 6: Provide for accurate prediction of solving and higher order processing skills in and visual-motor speed. Recall that the intelligence tests is accounted for by basic academic achievement at levels that assessing intelligence and does not consider fourth stratum two trait is broad visual Carroll’s (1993) stratum three, which is solely are at least comparable to that of visual-motor speed, rote memory, or retrieval perception, referred to by Carroll as composed of g (general intelligence). The intelligence tests twice its length. of facts (e. g., names of US Presidents, etc.) in visualization (1993). Tests of this factor greater the amount of g measured by an Goal 7: Apply familiar, common concepts determining an individual’s general require the ability to perceive and transform intelligence test, the higher the test’s that are clear and easy to interpret, intellectual level. stimuli accurately, possibly including visual correlation with important outcomes such as coupled with simple administration and manipulation (Kamphaus, 2001), and also academic achievement. Stratum three can be scoring. As discussed previously, the primary requires motor coordination. assessed through measures of stratum two Goal 8: Eliminate items that show differential goal for development of the RIAS was to traits, which are composed of combinations item functioning, or DIF, associated with assess g and its main components, fluid and Another trait excluded by the RIAS, of subtests, or stratum one measures. These gender or ethnicity. crystallized intelligence. Any test of g must visual-motor speed, is best captured by the second stratum traits include, in order of measure higher-order cognitive abilities, tail-end of Carroll’s (1993) second-stratum – relation to g, fluid intelligence, crystallized namely fluid intelligence, such as general broad cognitive speediness and processing intelligence, general memory and learning, The RIAS for Gifted Assessment sequential reasoning, induction, deduction, speed. Broad cognitive speediness requires broad visual perception, broad auditory analogies, etc (Carroll, 1993). These are the quick responding with little decision- perception, broad retrieval ability, broad There are various definitions of intellectual characteristics most important in identifying making, while processing speed involves cognitive speed, and processing speed. In giftedness and the determination of criteria gifted students, making fluid intelligence both a quick reaction time and decision- other words, fluid and crystallized abilities for eligibility to receive gifted services is key to assessing for giftedness. All of the making. The former is well-represented on are the best candidates for quick and typically left to officials at the State, RIAS subtests have a substantial g tests of intelligence in the form of design accurate assessment of general intelligence Provincial, or school district level. This saturation, and therefore are suited for gifted copying, while the latter is represented by (Kamphaus, 2001). situation leads to varying conceptions of identification (Reynolds & Kamphaus, 2003). items that require rapid choices (Kamphaus, what it is to be gifted and how best to 2001). These measures typically require These theories combined with practical identify this population. Many school Crystallized ability is measured by first- rapid motor coordination and are poorly considerations led to the primary goals for districts define intellectual giftedness as order factors that mainly involve language correlated with g. development as provided in the RIAS general intellectual ability or the ability to abilities, which are more commonly referred professional manual (Reynolds & achieve high academic achievement. Thus, to as verbal abilities by clinicians (Kamphaus, Because intellectual giftedness is so Kamphaus, 2003) and listed here. many schools include an intelligence test as 2001). Again, this second-order factor is one closely tied to the concept of general Goal 1: Provide a reliable and valid one, or sometimes the only, criteria for area of assessment on which the RIAS intelligence, it would be deleterious to

128, Gifted Education International Volume 21 No 2/3, 2006, 129 include measures of traits that are not at least Working memory is a lower level skill aged individuals should make it particularly good correlates of g. The assessment of Crystallized ability, as aforementioned, is that does form a basis for intelligence to useful in the search for intellectually gifted intellectual giftedness would be confounded the second stratum two trait and therefore function well. However, working memory and talented individuals. with domains that are not of interest, one of the best measures of g. This well- has relatively low correlations with possibly decreasing the likelihood of replicated finding suggests that based on the reasoning test scores and while it has some In summary, the RIAS was developed identification. The RIAS avoids this VCI scores of the motor impaired and overlap with intelligence (as do all cognitive with the goal of measuring factors that are confounding by excluding speed and motor control group we should expect the groups tasks), it is a different, lower level construct closely associated with g. Given the nature of coordination components. to also have similar Full Scale IQs (FSIQ). (e. g., see Ackerman, Beier, & Boyle, 2005; intellectual giftedness, these g-saturated Yet, this study resulted in a significant Deary, 2000; Kline, 2000). Working memory traits are well suited for the assessment of Even mild motor impairments can have difference between the mean FSIQ for the requires very little manipulation of this population. The RIAS is beneficial in significant, deleterious effects on IQ two groups (Mean = 85.7 vs Mean = 96.7, p = information and manipulation of evaluating the intellectually gifted because it estimation for some of the most popular of .08). This difference demonstrates that by information is what is crucial to reasoning. is not confounded with poor correlates of g, intelligence tests. An individual who is measuring constructs poorly related to g, Working memory is a basic building block i.e. visual-motor speed, motor coordination, highly skilled at problem-solving, deductive such as motor coordination and visual- for cognitive activity but operates with a and short-term memory. The score range of and inductive reasoning, and other higher motor speed, the overall intelligence threshold effect. Clearly a base level of the RIAS is also useful across the entire order processing skills, but who has motor measure is not an accurate representation of working memory is required to perform school-age range as well. problems, will have their overall intelligence higher level intellectual ability. reasoning tasks but having a really superb level underestimated on tests that include working memory does not make you any visual-motor skills and processing speed (as Similarly, the RIAS avoids confounding more superior at manipulation of RIAS Overview typically measured) in the estimation of IQs the Composite Intelligence Index (CIX) with information--so a basic level of working or related scores. a memory measure, but instead provides two memory is a necessary but insufficient skill Before employing a test for the assessment of supplementary subtests as a separate or ability for good reasoning, i.e., you can any population it is important to understand An illustration of this effect can be found memory composite (CMI). Memory loads on not reason very well in the absence of a base its administration, development, and in the Wechsler Intelligence Scale for the third stratum two factor, general memory level of working memory, but really high psychometric properties. A brief description Children-Fourth Edition (WISC-IV; Wechsler, and learning. As such, it is not as highly levels of working memory do make you of these aspects of the RIAS follows. The 2003) manual. In a clinical study, 21 children correlated with g as crystallized and fluid reason better past this threshold. The RIAS goals and theory on which the creation of the with significant motor impairments were intelligence and hence not of great interest in model notes the importance of having RIAS was based have been discussed administered the WISC-IV and compared to determining whether an individual meets the measures of memory conormed with previously (for a more in depth discussion of a demographically matched control group criteria for intellectual giftedness. The RIAS intellectual reasoning tasks and provides these issues see Reynolds & Kamphaus, administered the same test. Compared to the is rare in that it provides a measure of them, but relegates them to a separate scale 2003, and Reynolds & Kamphaus, 2005, control group, the motor impaired group memory if needed for a specific population, that does not contribute to the intelligence which also includes a case study performed significantly worse on the but it is a separate composite and not indexes in the proffered model of Reynolds application). Processing Speed Index (PSI), which confounded with the overall measure of and Kamphaus (2003).. includes tests such as copying symbols based general intelligence. The RIAS includes six subtests that are on a given code, scanning for matching In evaluating potentially high scoring based on formats common in intellectual symbols, and then marking an answer box Rote memory, such as represented in individuals, the range of possible scores is of assessment providing an instant familiarity (Mean = 78.2 vs Mean = 97.7, p < .01). There tasks like digit or letter recall, remembering importance as well. The RIAS is scaled such for the user. There are four intelligence was also a significant difference between symbols associated with single numbers, and that the indexes provide an eight standard subtests requiring integration of several groups’ performance on the Perceptual the like, require little in the way of mental deviation score range, from 40 to 160. While intellectual functions and two memory Reasoning Index (PRI), which includes a manipulation and their inclusion in estimates floor and ceiling effects occur at some ages, subtests. The intelligence subtests are further block design subtest requiring motor of intellectual ability seems questionable. the age range (3 years to 94 years) and subdivided into a Verbal Intelligence Index coordination (Mean = 83.8 vs Mean = 94.8, p Although knowledge of memory skills of associated range of difficulty of the various (VIX) and a Nonverbal Intelligence Index = .05). However, the groups performed various sorts is very useful from a diagnostic tasks on the RIAS, provides it with an upper (NIX). equally well on stratum two measures of perspective, - it just should not be confused range of IQs (CIX) fully to 160 for ages 3 crystallized ability represented by the Verbal with higher order information processing years to 14 years, 159 at age 15, and 150 or The VIX includes the Guess What (GWH) Comprehension Index (VCI; Mean = 95.5 vs when referring to working memory and higher at remaining ages. The scaling of the and Verbal Reasoning (VRZ) subtests. Guess Mean = 97.9, p = .62). other forms of rote recall. RIAS and its range of difficulty for school- What requires verbal reasoning, vocabulary,

130, Gifted Education International Volume 21 No 2/3, 2006, 131 language development, and verbal must repeat the words verbatim. Nonverbal parameter estimation and minimizes the stability, was examined using the test-retest knowledge base. It consists of questions Memory requires the ability to encode, store, effects of sampling irregularities across the method for 86 individuals with a median beginning with “What is…” followed and recall pictorial stimuli that are concrete 52 age groups in the normative sample. For time interval of 21 days. Test-retest reliability typically by three clues (occasionally more or abstract. The examinee is presented with the subtests, T scores with a mean of 50 and correlations and corrected coefficients for the are given) describing the target object, the target picture for 5 seconds and then a standard deviation of 10 were chosen over subtests and indexes are provided in the person, place, or concept and requires only a asked to identify the target within a set of the traditional scaled score due to the higher RIAS Manual (Reynolds & Kamphaus, 2003). one or two word response from the pictures presented on the following card. reliability coefficients for RIAS subtest scores The corrected coefficients range from .76 to examinee. The Verbal Reasoning subtest thus allowing finer discriminations. For the .91 for the 3- to 4-year-olds, .76 to .92 for the measures verbal-analytical reasoning ability The RIAS is easy to administer and intelligence indexes (VIX, NIX, CIX) as well 5- to 8-year-olds, .77 to .99 for the 9- to 12- but with fewer vocabulary and general objectively scored. Judgments are seldom as the memory index (CMX), the traditional year-olds, .73 to .99 for the 13- to 82-year- knowledge demands than GWH. The required in scoring, with the exception of the IQ-metric is used (Mean=100, SD=15). For olds, and .76 to .91 for the entire test-retest examinee is required to supply one or two Verbal Memory subtest, and interrater those who are interested, percentiles, T sample. Based on these reliabilities, words that complete a complex analogy. reliability exceeds the recommended value scores, z scores, normal curve equivalents examiners may expect small practice effects of .90 for all 6 subtests with a median value (NCE’s), age equivalents, and stanines are when using the RIAS. In addition, as The NIX scale includes the Odd-Item Out of 1.00 (Reynolds & Kamphaus, 2003). After provided in the RIAS Manual (Reynolds & previously mentioned, all subtests have (OIO) and What’s Missing (WHM) subtests. the examiner has become familiar with the Kamphaus, 2003). more than sufficient interrater reliabilities, The examinee has two opportunities to administration, the entire RIAS requires an with the lowest being .95 for Verbal Memory. generate the correct response on both average of 25 to 35 minutes, including the Internal consistency reliability of the Given that the RIAS is a relatively new subtests, receiving two points if correct the memory tasks. The intelligence subtests RIAS was assessed using Cronbach’s intelligence scale, the RIAS Manual first time and one point for the second. Odd- alone can be administered in about 20 to 25 coefficient alpha, as is commonly done for (Reynolds & Kamphaus, 2003) is the main Item Out measures nonverbal reasoning minutes, although it may take longer for such tests. The lowest alpha reliability for the source of validity evidence. Theory-based skills and also requires the subject to use bright individuals such as those who are RIAS subtests scores is .84 across all age evidence, content evidence, response spatial ability and visualization. It is a assessed for being potentially gifted. Basal groups, with the medians across age for the processes evidence, internal structure reverse matrix analogies task. The examinee and ceiling rules as well as age designated subtests equaling .90 or greater. Similarly evidence, external relations evidence, and is presented with a card containing several starting points were determined based on high alpha coefficients were obtained for the consequential evidence are provided as figures that complete a form of matrix responses of the standardization sample in indexes, with the lowest median coefficient recommended by the Standards (AERA, analogy but with one extra item. The order to control administration time. In equal to .94. These reliability estimates are APA, & NCME, 1999). Theory will not be examinee must identify the one that is addition, administration is made more quite strong and exceed values of composite discussed here as it was discussed different from the others, the item that does efficient by the presence of all necessary indexes of some tests considerably longer previously, and an explanation of not belong. What’s Missing requires instructions on the RIAS record form. than the RIAS and are as high as or higher consequential validity will be omitted due to nonverbal reasoning where the individual than most popular IQ tests which are its controversial nature. must conceptualize a picture, analyze its The RIAS was normed on a significantly longer to administer and score. gestalt, and deduce the essential missing representative sample of 2,438 participants This is partly due to the focus of the RIAS on Content evidence of validity is provided element, indicating the response verbally or based on the demographic characteristics of g and excluding extraneous variables, by direct examination of the RIAS items and by simply pointing to the location of the the U.S. population as defined by the 2001 making the items and subtests more format. Items were reviewed by experts for missing elements. It is not necessary to give U.S. Bureau of the Census Report (Reynolds homogenous and therefore more reliable. cultural bias and appropriateness, with a verbal response or to know the “name” of & Kamphaus, 2003). Stratification variables questionable items being eliminated or the missing essential element. included age, gender, ethnicity, educational The RIAS manual also reports the modified. Response process evidence is level, and region. Oversampling of the South reliability of the test’s scores across gender based on fit between what the subtests The Composite Memory Index (CMI) region and minority ethnic groups were and by ethnicity. No significant differences propose to measure and the tasks employed. includes the Verbal Memory (VRM) and corrected for using a weighting procedure. were found for any age level as a function of In this respect, all subtests are devoid of Nonverbal Memory (NVM) subtests. Verbal The resulting norms represent a nearly gender. Additionally, no significant contamination from nonintellectual Memory assesses the ability to encode, store, perfect match to the population statistics. differences in reliability were found for processes, such as motor coordination, and recall verbal material in a meaningful whites versus African Americans, although speed, and acuity. The verbal subtests context. Sentences or stories with lengths All standard scores for the RIAS were at some age levels the absolute value of the measure vocabulary and language determined by the examinee’s age are read obtained using continuous norming, a differences was greater than for gender. comprehension, not simply acquired skills aloud by the examiner and the examinee regression procedure that stabilizes Error due to time sampling, or test score and little expressive language is required.

132, Gifted Education International Volume 21 No 2/3, 2006, 133 The nonverbal subtests are designed to consisting of verbal, nonverbal, and memory indexes correlated highly with the WISC-III populations. Further, its exclusion of require primarily spatial encoding, and components. Model 1 fit better when only Full Scale IQ (FSIQ), ranging from .60 for the confounding factors such as motor ability avoid verbal strategies. the four intelligence subtests were included NIX to .78 for the VIX. This pattern is and speed provides an even more accurate and the results suggest that the RIAS has a expected because the VIX and FSIQ are most picture of intellectual functions related to Evidence based on internal structure was large first factor, yet inspection of Model 2 associated with g based on their factor intellectual giftedness than other currently supplied by confirmatory and exploratory and Model 3 was still necessary. Model 2 analyses. Additionally, differences are to be available measures. The range of intelligence factor analyses, in addition to internal provided a good fit for the four subtests as expected between the NIX and WISC-III IQs index scores available (up to 160 through age consistency as noted above. To enhance well with chi-square values between .22 and due to the WISC-III’s Performance IQ, which 14 years and above 150 thereafter) coupled stability and generalizability of the results 1.49 and small RMSEA values. Model 3 for incorporates the motor and language skills, with their reliability should make the RIAS for the factor analysis of subtest six subtests did not demonstrate as good a and speed components that are purposely maximally useful with high IQ populations. intercorrelations without masking fit, but there were indications that Model 2 absent from the RIAS nonverbal scale. A brief version of the RIAS, the Reynolds developmental shifts, the sample was for the six subtests was a possibility for Additionally, the RIAS was shown to be Intellectual Screening Test (RIST; Reynolds & divided into five age groups: early individuals 19 years of age and older. While predictive of educational achievement based Kamphaus, 2003) is also available although childhood (3 to 5 years), childhood (6 to 11 the use of all six subtests for assessing on correlations in the .60s and .70s with the not specifically reviewed here. years), adolescence (12 to 18 years), nonverbal and verbal intelligence in adults Wechsler Individual Achievement Test adulthood (19 to 54 years), and senior may be plausible, Reynolds and Kamphaus (WIAT; The Psychological Corporation, adulthood (55 to 94 years). (2003) recommend the use of the four-subtest 1992). References two-factor model based on the factor The principal factors solution with two- analysis results and the theories discussed The RIAS scores for clinical samples Ackerman, P. L., Beier, M., & Boyle, M. (2005). factors makes the most sense for both the previously. identified during standardization were Working memory and intelligence: The same or four intelligence subtests and all six subtests obtained to provide evidence of validity different constructs. Psychological Bulletin, 131 (1), as a set. The g factor for the RIAS is the The RIAS shows evidence of external based on the match between observed 30-60. strongest with all but one of the loadings on validity based on developmental trends, performance and hypothesized performance g being .60 or higher. The two verbal subtests relations to other measures, and for a given disorder. The clinical groups American Educational Research Association, American Psychological Association, and have the highest loadings on g, but the performance of clinical groups. To obtain a included five organic syndromes, such as National Council on Measurement in Education nonverbal subtests are also good measures measure of relationship between deafness and mental retardation, five (1999). Standards for educational and psychological of the unrotated factor. Two components of performance on the RIAS and age, psychiatric groups, such as depression and testing. Washington, DC: American Psychological g, verbal and nonverbal, were clearly correlations were calculated for ages 3 polysubstance abuse, and learning Association. delineated for every age group. The six- through 18 years. The correlations were then disabilities and attention deficit subtest solution also followed these two examined to determine whether they hyperactivity disorder groups. The scores for Carroll, J. B. (1993). Human cognitive abilities: A dimensions, with Verbal Memory on the first followed the hypothesized early and rapid these samples displayed the expected survey of factor analytic studies. New York: factor with the verbal subtests and increase, plateau in the teen years, and patterns of mild cognitive impairment with Cambridge University Press. Nonverbal memory on the second factor decline in the older years of the intelligence greater impairment for more severe Deary, I. J. (2000). Looking down on human with nonverbal subtests. Despite these composite. The correlations were large disorders, providing further proof of validity intelligence: From psychometrics to the brain. New loadings, the authors contend that the (greater than .80) for the sample as well as for the RIAS. York: Oxford University Press. memory subtests are best presented as a for gender and ethnicity showing that raw Conclusion separate memory index. The four-subtest scores increase with age across subtests with Horn, J. L., & Cattell, R. B. (1966). Refinement and and six-subtest solutions demonstrate the common stimuli. In addition, polynomial The RIAS is a quick and easy to test of the theory of fluid and crystallized general same factor structure across gender and curves were constructed to illustrate that the administer (25 –35 minutes) measure of intelligences. Journal of Educational Psychology, ethnic groups based on factorial similarity expected developmental trend was evident intelligence that provides a highly reliable 57(5), 253-270. requirements and coefficients of congruence. for the entire age range from 3 to 94 years. and valid measurement of general Clinical assessment of child Confirmatory factor analyses were intelligence and its two main components, Kamphaus, R. W. (2001). and adolescent intelligence. (Second edition). conducted to compare the chosen two factor To demonstrate external validity with verbal and nonverbal abilities. Its focus on Boston: Allyn and Bacon. model (Model 2) of the RIAS to a single other measures, each RIAS index was tasks most related to higher-order factor model consisting simply of g (Model correlated with the three IQs provided by the conceptions of intelligence makes it a prime Kline, P. (2000). Handbook of psychological testing 1) and a three factor model (Model 3) WISC-III (Wechsler, 1991). All of the RIAS candidate for use with intellectually gifted (2nd ed.). New York: Routledge.

134, Gifted Education International Volume 21 No 2/3, 2006, 135 McCoach, D. B., Kehle, T. J., Bray, M. A., & Siegle, Reynolds, C. R., & Kamphaus, R. W. (2003). D. (2001). Best practices in the identification of Reynolds Intellectual Assessment Scales and Reynolds gifted students with learning disabilities. Intellectual Screening Test: Professional Manual. Psychology in the Schools, 38, 403-411. Odessa, FL: PAR, Inc.

Reis, S. M., & Small, M. A. (2005). The varied and The Psychological Corporation. (1992). Wechsler unique characteristics exhibited by diverse gifted individual achievement test. San Antonio, TX: and talented learners. In F. A. Karnes, & S. M. Psychological Corporation. Bean (Eds.), Methods and materials for teaching the gifted (Second edition; pp. 3-35). Waco, TX: Wechsler, D. (2003). Wechsler intelligence scale for Prufrock. children-fourth edition. San Antonio: The Psychological Corporation. Reynolds, C. R., & Kamphaus, R. W. (2005). Introduction to the Reynolds Intellectual Wechsler, D. (1991). Wechsler intelligence scale for Assessment Scales (RIAS) and the Reynolds children-third edition. San Antonio, TX: The Intellectual Screening Test (RIST). In D. P. Psychological Corporation. Flanagan & P. L. Harrison (Eds.), Contemporary intellectual assessment, Second edition (pp. xxx- xxx). New York: Guilford.

136, Gifted Education International