Printer Friendly

Special education screening services: group achievement test.

Special Education Screening System: Group Achievement Test

The purpose of a special education screening system is to identify students who deviate sufficiently from their peers as to require special attention (Salvia & Ysseldyke, 1985). The most common method of identifying students in need of special education services has been teacher-submitted nominations of students who are then individually administered traditional psychological and educational measures by a school psychologist. This method of teacher referral, followed by an individual psychological evaluation, is typical of the Utah procedure (Utah State Office of Education, 1981) and is consistent with federal guidelines (P.L. 94-142). This identification system has been demonstrated to qualify a large majority of those tested for special education, approximating 90% (Algozzine, Christenson, & Ysseldyke, 1982; Galagan, 1985; Utah State Office of Education, 1975). Because school psychologists currently spend half their time in performing assessment activities (Benson & Hughes, 1985), educators have proposed and used alternate systems. (Peterson, Heistad, Peterson, & Reynolds, 1985; Welton & Wedell, 1982). These systems decrease the frequency of standardized individual assessments for special education qualification.

Under the present sytem, a referred student most likely will qualify for special education. Thus the referral decision often determines a student's subsequent educational program (Christenson, Ysseldyke, & Algozzine, 1982). Current individual testing has two functions: (a) diagnosing (or confirming) a suspected handicap and (b) identifying a student's special needs in designing an individual education program (Tucker, 1985; Utah State Office of Education, 1981). Unfortunately, a growing body of literature suggests substantial flaws in both functions. First, regarding diagnostic individual testing, the problem exists in distinguishing between various categories of mildly handicapped children, as well as in determining who is and who is not mildly handicapped (GAO, 1981; Hagerty & Abramson, 1987; Tucker, 1985). Second, many educators have criticized the efficacy of the standardized tests typically used to assess students' individual needs (Galagan, 1985; Gresham, 1986; Ysseldyke et al., 1982). Researchers have argued that these test results do not effectively link assessment with educational intervention, thereby providing teachers with information that is not useful in planning effective instructional programs (Howell, 1986; Tucker, 1985).

An alternate identification system, the Montevideo, Minnesota Individualized Prescriptive Instructional Management System (Peterson et al., 1985), uses a 20th percentile cutoff point on a curriculum-based measure of academic progress over time, below which 100% of the special education students were identified. With this measure, however, approximately 8% of regular education students also fell below the cutoff in reading and math.

In the present study, we used existing standardized achievement test results to determine if group scores on an achievement test would be useful as a preliminary screening device to identify mildly handicapped special education students currently in the system.

The value of individual psychological evaluations at some point in the process of individualizing student programs is not disputed or indeed addressed in the current study. The problem of providing for individual differences is complex, and efforts to help individual children require multifaceted procedures. The possibility that group testing could provide an effective initial screening and identification of the mildly handicapped is an intriguing notion. If effective, it would provide more human resource use at other points in a continuum of educational and personal experiences.



In the first sample, four of eight elementary schools in the Uintah School District in Vernal, Utah, were chosen at random from within the district. All schools were located in or near a rural and industrial area surounding a small town (1984 population: 8,100). One of the schools has a minority American Indian population of approximately 40%. These four schools closely represent the diversity in ethnicity and socioeconomics status of the school district as a whole. A total sample of 1,434 students in Grades 1 through 6 was analyzed. This sample consisted of all the students who took the entire Stanford Achievement Test (SAT) battery for the spring of 1985 during the district's group testing program. The group included 1,249 regular education students and 126 special education students.

The special education students included the learning disabled (LD) (N = 81), behavior disordered (BD) (N = 35), and mildly intellectually handicapped (IH) (N = 10) whose are presently in the regular school system receiving from 30 minutes to one-half day resource. The majority of these students are receiving only one period a day of resource instruction. Children with speech or orthopedic handicaps were not included. The high incidence of learning disabled (LD) students in this study cold affect the results, since this classification implies an academic deficit. The rate of LD students (comprising 56% of total students classified as handicapped) in this sample, however, compares with the median value of 45% found by Gerber (1984) for all states in 1983. The national incidence rates varied from a low of 28% to a high of 86%, with 18 states registering at 50% or above for LD students (Gerber, 1984). Further, Gerber indicates that the rate for this classification had rapidly increased for the 7 years prior to 1983, and this trend is anticipated to continue. The proportions of LD students in this study are therefore in line with national averages.

In a cross-validation sample, 2 of 64 elementary schools were assigned to the investigators at random from the Special Education Director of the Granite School District, Salt Lake City, Utah. This school district is located in a major metropolitan urban and suburban area. A total of 1,011 students in Grades 1-6 were included from this city school district. This group included 921 regular education students and 90 special education students. This sample was used to demonstrate whether equivalent results could be obtained from a different population.


The Stanford Achievement Test Series was standardized on a large national sample, and the reliability, validity, and norming procedures are well documented (Psychological Corporation, 1983). The spring standardization of this instrument involved 200,000 students from 300 school districts chosen to represent the national population in terms of school system enrollment, geographic region, and public versus nonpublic enrollment, and it was stratified by socioeconomic status. The total battery score consists of 13 subtest scores for first grade and 14 subtest scores for second through sixth grades. The total reading score consists of four subtests for first and second grades and two subtest scores for third through sixth grades.

Design and Procedure

The district scores were downloaded from a state computer (which accessed all Stanford Achievement testing in the State of Utah) into a dBase II data base on the Corvus hard drive of an Apple IIe computer system.

A rank ordering of SAT total battery percentile scores was done for each school and for all schools combined, first for the original group and 5 months later for the cross-validation group. Another set of rank orderings using the SAT total reading scores was completed in the same manner. The results were then listed in a tabular format. The incidence rates for special education and regular students appearing below the four arbitrarily predetermined cutoffs were then listed. Individuals were described as "hits" if the placement agreed with their current classification and "misses" if the placement disagreed with their current classification.


The hits, misses, and percentage correctly identified by using the total battery score of the SAT are given in Table 1. As one might expect, the lower the cutoff is set, the greater the likelihood that regular education students will be identified as above the cutoff, or currently in regular education. In the first group, at the 5th percentile 8 regular education students out of a total of 1,249 are "missed" and would be classified as special education. At the 20th percentile, 143 regular education students from the 1,249 total are "missed" and would be classified as special education. In the cross-validation group, similar results are obtained. At the 5th percentile cutoff, 14 out of 921 regular education students are "misclassified," and at the 10th percentile level, 39 of the 921 regular education students appear below the cutoff.

Again, consistent with expectations, the higher the cutoff score is set, the greater the likelihood that special education students will be classified as special education students, and the lower the cutoff, the greater the likelihood that they will be "missed." In the first group, at the 5th percentile, 65 special education students out of a total of 126 are "missed" and would be classified as regular education. At the 20th percentile, only 8 of the 126 total would be "missed" and classified as regular education students. In the cross-validation group, 14 out of 921 regular education students appear below the 5th percentile cutoff, and 36 of 921 appear below th e 10th percentile cutoff.

The hits, misses, and percentage correctly identified by using the total reading score of the SAT are given in Table 2.

The differences between the total battery scores and the total reading scores appear inconsequential. The total number of hits for the special education students are generally slightly higher for the total battery score than for the total reading score. It is considerably easier, however, to obtain total reading scores (which require fewer subtests) than to obtain total battery scores.

In identifying students previously identified as handicapped, it appears that this screening system is capable of the same basic 90% accuracy found in the present teacher referral system.

Data on the special education students above the 20th percentile revealed that five LD, three BD, and no IH students were above this cutoff on total battery scores, and nine LD and two BD students were above the same cutoff for total reading scores.


The group achievement test employed in this investigation discriminates remarkably well between the special and regular education populations of these two school districts. If a 20th percentile cutoff is used as the screening criterion, all but a handful of special education children are identified. At this level, however, many more children who are currently classified as regular education students would be included in special education. It is indeed difficult to determine where to draw the line that determines where mildly handicapped scores leave off and nonhandicapping scores begin.

If the 5th percentile is used, about half the current special education children are identified (which is statistically all that could be expected, given the state's 12.18% allowable percentage of handicapped students within a population) and only a small minority of children who are currently in regular education would be included. If the cutoff is set a a very low level, it would be reasonable to assume that any children with such low achievement levels are in need of special instruction.

One might argue that the purpose of all screening and classification is to identify children who are having difficulty in the educational system and to give them focused instruction. Special education programs, with their attendant resource rooms, are only a vehicle to efficiently and humanely deal with individual differences. The logic of allowing some children with profound educational deficits to have a special educational experience, while denying others with equal deficitls, could be construed as discriminatory. If one assumes that most special education programs stress: (a) starting at the individual's actual achievement level, (b) giving the opportunity for greater drill and repetition, (c) affording more personal attention, and (d) using rewards and incentives, it would be difficult to believe that special instruction would not be justifiable for any child not achieving well in his or her current educational setting.

Handicapping Incidence Rates

Above the 20th Percentile


Above the higher cutoffs it would seem that the most likely handicapping condition to occur would be BD students, as handicapping definition an extreme academic deficit is not required of the BD student. However, this was not the case. The total ratio of LD to BD students in this study is approximately 2.3:1. The ratio of LD to BD students above the 20th percentile cutoff for both scales combined is 2.8:1 (no IH students scored above the 20th percentile on either measure).

This appears to be a function of the day-to-day practicalities operating within the rubric of the referral system. Students who are referred for special education are most often those that simply are not "keeping up" academically. These low achievers are fit into the system with whatever classification is appropriate or, possibly, convenient. therefore, it appears that students with behavior problems and low academic achievement scores are served, or at least referred, in lieu of students with behavior problems who do not demonstrate low academic achievement.

Implications for


The present study supports the rationale that using group achievement test data and setting the cutoff at the 5% level would identify almost exclusively children who are currently identified as special education students. Those few who are not so identified should probably have been. If resources are available to service a larger number of children, higher cutoffs could be used. If this procedure were used for this particular group of students, between 400 and 600 hours of school psychologist and classroom teacher time could have been spared.

To use the results of this study and still maintain flexibility within the system, various procedures could be instituted. First, we propose using standardized test data to form the initial roster for possible inclusion into special programs. Second, the district would then set a cutoff that would include all children scoring below it (we advocate the 5% cutoff). The present study furnishes data on these first two steps only. Third, the parents and teachers of children so identified would meet and reach consensus on a placement in or out of the special programs. If they believed that further information was necessary before a determination would be made, a school psychologist could do an assessment taking into account the areas of concern for both the parent and the teacher. Fourth, those scoring above the proposed cutoff could still be referred, tested, and placed, as in the present identification system.


Different instruments may be capable of identifying different students as handicapped. The present system is, no doubt, identifying students that it finds useful to identify as handicapped, although just who constitutes the mildly handicapped population is open to debate (Hagerty & Abramson, 1987; Sapon-Shevin, 1987). This study attempts to nominate by group test a set of low-achieving students, without regard to IQ score. These students are generally being qualified presently, but through individual testing rather than a group screening system.

In a time when funding is becoming more difficult in every tax-dependent sector of the economy, and nondiscrimination and freedom of choice are significant issues, educational practices should be under continuous review. The present individual identification and classification system requires a considerable expenditure of time, particularly for school psychologists, and serves to limit their role (Carroll, Harris, & Bretzing, 1979; Cook & Patterson, 1977; Lacayo, Sherwood, & Morris, 1981). Changes in the initial identification of children with learning problems appear warranted; standardized group achievement tests could be used in such a procedure.


Algozzine, B., Christenson, S., & Ysseldyke, J. (1982). Probabilities associated with the referral to placement process. Teacher Education and Special Education, 5, 19-23.

Benson, J. A., & Hughes, J. (1985). Perceptions as role definition processes in school psychology: A national survey. School Psychology Review, 14, 64-74.

Carroll, J. L., Harris, J. D., & Bretzing, B. H. (1979). A survey of psychologists serving secondary schools. Professional Psychology, 10, 766-770.

Christenson, S., Ysseldyke, J., & Algozzine, B. (1982). Institutional constraints and external pressures influencing referral decisions. Psychology in the Schools, 19, 341-345.

Cook, V. J., & Patterson, J. G. (1977). Psychologists in the schools of Nebraska: Professional function. Psychology in the Schools, 14, 371-376.

Galagan, J. (1985). Psychoeducational testing: Turn out the lights, the party's over. Exceptional Children, 52, 288-299.

General Accounting Office (GAO). (1981). Disparities still exist in who gets special education. Washington, DC: U.S. Government Printing Office.

Gerber, M. M. (1984). The Department of Education's Sixth Annual Report to Congress on P.L. 94-142: Is Congress getting the full story? Exceptional Children, 51, 209-224.

Gresham, F. (1986). On the malleability of intelligence: Unnecessary assumptions, reifications, and occlusion. School Psychology Review, 15, 261-263.

Hagerty, G. J., & Abramson, M. (1987). Impediments to implementing national policy change for mildly handicapped students. Exceptional Children, 53, 315-323.

Howell, K. W. (1986). Direct assessment of academic performance. School Psychology Review, 15, 324-335.

Lacayo, N., Sherwood, G., & Morris, J. (1981). Daily activities of school psychologists: A national survey. Psychology in the Schools, 18, 18-190.

Peterson, J., Heistad, D., Peterson, D., & Reynolds, M. (1985). Montevideo individualized prescriptive instructional management system. Exceptional Children, 52, 239-243.

Public Law 94-142, The Education for All Handicapped Children Act of 1975. Washington, DC: Federal Register.

Psychological Corporation. (1983). Stanford Achievement Test technical manual. San Antonio: Harcourt Brace Jovanovich.

Salvia, J., & Ysseldyke, J. (1985). Assessment in special and remedial education. Boston: Houghton Mifflin.

Sapon-Shevin, M. (1987). The National Education Reports and special education: Implications for students. Exceptional Children, 53, 300-306.

Tucker, J. A. (1985). Curriculum-based assessment: An introduction. Exceptional Children, 52, 199-204.

Utah State Office of Education (1975). Project identification. Salt Lake City: Author.

Utah State Office of Education. (1981). Rules and regulations for education programs for the handicapped. Salt Lake City: Author.

Welton, J., & Wedell, K. (1982). Special education in England and Wales. Report printed at the Univerity of London, Institute of Education.

Ysseldyke, J., Thurlow, M., Graden, J., Wesson, C., Deno, S., & Algozzine, B. (1983). Generalizations from five years of research on assessment and decision making. Exceptional Education Quarterly, 4, 75-93.

BRIAN STONE is School Psychologist, Uintah School District, Vernal, Utah. BERT P. CUNDICK is Professor of Psychology, Psychology Department, Comprehensive Clinic, Brigham Young University, Provo, Utah. DAVE SWANSON is Special Education Director, Uintah School District, Vernal, Utah.
COPYRIGHT 1988 Council for Exceptional Children
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1988 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Stone, Brian; Cundick, Bert P.; Swanson, Dave
Publication:Exceptional Children
Date:Sep 1, 1988
Previous Article:Service coordination between correctional and public school systems for handicapped juvenile offenders.
Next Article:The special education professional.

Related Articles
Perspectives on eligibility for placement in special education programs.
Categorical programs for remedial and handicapped students: issues of validity.
Curriculum-based measurement and two models of follow-up consultation.
The pitfalls and promises of special education practice.
Who becomes an "at-risk" student? The predictive value of a kindergarten screening battery.
The RISK screening test: using kindergarten teachers' ratings to predict future placement in resource classrooms.
Categorical and programming changes in special education services.
The need for technological advances in assessment related to national educational reform.
20/20 analysis: estimating school effectiveness in serving students at the margins.
Effect of different methods of reporting and reclassification on trends in test scores for students with disabilities.

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters