Cautions in interpreting WRAT standard scores as criterion measures of achievement in young children

Abstract

The present article calls into question the use of the WRAT as a criterion measure for determining ability/achievement discrepancies in young school-aged children. It is demonstrated that the use of regression equations to predict the expected achievement levels of five-and six-year-old children on the WRAT still does not rectify the difficulties in identifying underachievement for children in the lower primary grades. Implications for the early identification of specific learning disabilities are discussed.The predictive validity of the Wechsler Intelligence Scale for Children-Revised (WISC-R) in relation to measures of achievement has received considerable attention during the past few years (Hale, 1978; Hartlage & Steele, 1977; Schwarting & Schwarting, 1977; Wikoff, 1979). Many of these inquiries have studied the relationship between the WISC-R and Wide Range Achievement Test (WRAT). Hale, for example, investigated relationships between WISC-R Verbal and Performance IQ scores and WRAT standard scores in Reading, Spelling, and Arithmetic, and discovered that WISC-R Verbal IQ significantly predicted WRAT Reading and Arithmetic standard scores. Hartlage and Steele reported significant positive correlations between all WISC-R IQ scales and all WRAT subtests, as did Schwarting and Schwarting. Such findings have lent further support to the already widespread use of the WRAT as a valid criterion measure of achievement levels.Hale addressed the need to construct prediction rules to determine whether an individual is achieving within expected ranges based upon his/her ability. In effect, his analysis generated regression equations that can be used to predict an individual's WRAT Reading and Arithmetic standard scores from his/her WISC-R Verbal IQ. Hale's decision to employ WRAT standard scores rather than grade level differences to define underachievement was based upon the contention that the use of grade level scores tends to result in disproportionate numbers of older children being identified as underachievers to the relative exclusion of younger children from this categorization.The purpose of the present paper is to raise the issue that while the use of WRAT standard scores as criteria in the identification of underachievement is a preferable method compared to the use of grade level designations, limitations with regard to sole reliance upon such scores as diagnostic indicators of underachievement are evident. Specifically, an examination of the normative tables of the WRAT, particularly at the lower chronological age levels (e.g., 5-0 to 6-1 l), reveals that even when standard scores are used, it remains difficult to identify five-and six-year-old youngsters as underachievers. DOES THE WRAT REFLECT CLASSROOM ACHIEVEMENT?It is of interest to note the skills that are required of five-and six-year-old children to achieve specific grade ratings and standard scores on the WRAT. As an example, let us assume that a psychologist administers the WRAT to a child of age six years, four months. An examination of the Spelli...

On the adequacy of the Wide Range Achievement Test (WRAT): A reply to Snart, Dennis, and Brailsford.

“…Snart et al claim that Silverstein (1978) found that WRAT scores were overestimated but in fact there seems to be, in most cases, little difference (0-8 points) between the standard scores that were calculated on the basis of the WRAT norms and those calculated norms from another sample. Contrary to the assertions of Snart et al, Grossman (1981) does not actually present any evidence that WRAT scores overestimate the performance of young children.…”

Concerns regarding the Wide Range Achievement Test (WRAT).

“…Obviously we are not the first to have concerns about the accuracy of the WRAT in reflecting the true academic or achievement levels of children and adults. However, despite an abundance of recent criticism by American psychologists and educators (Grossman, 1981;Salvia & Ysseldyke, 1978;Silverstein, 1978), the WRAT continues to be used in both the U.S. and Canada, perhaps because it can be administered and scored relatively easily and quickly, and perhaps because users lack an awareness of better alternatives and/or the serious problems which exist with the instrument. In Canadian populations, for whom the WRAT was never normed, the latter problems are further compounded, and we would therefore like to make some cautionary comments to WRAT users.…”

“…Given the present data and an abundance of anecdotal evidence gleaned from conversations with elementary school teachers, we would suggest that both the 1976 and 1978 norms, for reading at least, consistently overestimate grade-level performance on word recognition by approximately 1 to 2 years, thus negating their utility either for pinpointing a disability or for suggesting appropriate instructional levels. Especially at younger ages the overestimation and variability of scores are even greater (Grossman, 1981). If the WRAT reading test is given on an individual basis, one could infer from the published norms that a child who is actually 1 or 2 years behind age-mates is functioning appropriately in reading, or conversely that a child who in fact is functioning in line with age-mates should be in an advanced reading program.…”

“…Pikulski and Shanahan, commenting on the inflated WRAT scores, noted that 37% of the students were placed at least one grade higher than on the informal reading inventories and 30% were placed three grades higher. Grossman (1981) demonstrated in his study with young school-aged children that even the use of standard scores and regression equations to predict expected achievement did not rectify the difficulties in assessing achievement with the WRAT. Grossman commented:…”

“…Therefore we will utilize these scores to underline the points which we wish to make in the present discussion. In fact, since we are dealing with average classes of children at each grade, the inflated grade scores would also translate into inflated standard scores (Grossman, 1981). In all cases in Table 1, both the 1976 and 1978 norms serve to overestimate reading performance in terms of the child's actual grade placement.…”