52
Volume 12, Number 2 Journal of Education The Journal of the Council of Administrators of Special Education A Division of the Council for Exceptional Children Standards, Assessment, and Instructional Accountability District and State Standards and Assessments: Building an Inclusive Accountability System CASE IN POINT: Including Students with Disabilities in Large-Scale Testing: One Case at a Time ................................................................................3 ..........11 .......................................................................21 ...................................................................................31 ............................................................37 ............................................................39 —Martha L. Thurlow, Ph.D. and Sandra J. Thompson —Gerald Tindal, Ph.D., Robert Helwig, Ph.D., and Keith Hollenbeck, Ph.D. —Preston C. Green, Ed.D., Esq., and Stephen G. Sireci, Ph.D. —Thomas F. Hehir, Ed.D. An Update on Test Accommodations: Perspectives of Practice to Policy Legal and Psychometric Policy Considerations in the Testing of Students With Disabilities Begin Early, End Well: Strategies to Improve Results for Students With Disabilities Inclusive Assessments: A Means to an End —Edward Lee Vargas, Ed.D. —David J. Chard, Ph.D. Articles Special Issue Special Leadership

Volume 12, Number 2 Journal of Special Education … 12, Number 2 Journal of Education ... materi-als, and services. ... Journal of Special Education Leadership 12(2)

Embed Size (px)

Citation preview

Volume 12, Number 2

Journal of

Education

The Journal of the Council of Administrators of Special EducationA Division of the Council for Exceptional Children

Standards, Assessment, and Instructional Accountability

District and State Standards and Assessments: Building an Inclusive Accountability System

CASE IN POINT: Including Students with Disabilities in Large-Scale Testing: One Case at a Time

................................................................................3

..........11

.......................................................................21

...................................................................................31

............................................................37

............................................................39

—Martha L. Thurlow, Ph.D. and Sandra J. Thompson

—Gerald Tindal, Ph.D., Robert Helwig, Ph.D., and Keith Hollenbeck, Ph.D.

—Preston C. Green, Ed.D., Esq., and Stephen G. Sireci, Ph.D.

—Thomas F. Hehir, Ed.D.

An Update on Test Accommodations: Perspectives of Practice to Policy

Legal and Psychometric Policy Considerations in the Testing of Students With Disabilities

Begin Early, End Well: Strategies to Improve Results for Students With Disabilities

Inclusive Assessments: A Means to an End

—Edward Lee Vargas, Ed.D.

—David J. Chard, Ph.D.

Articles

Special Issue

Special

Le ad e rsh ip

Editor

Dr. Mary Lynn BoscardinUniversity of Massachusetts at Amherst

Assistant to the Editor

Rachel ParkerUniversity of Massachusetts at Amherst

Board of Associate Editors

Dr. Patricia AnthonyCape Cod Lighthouse Charter School • Orleans, MA

Dr. Colleen A. CapperUniversity of Wisconsin Madison

Dr. Carl LashleyUniversity of North Carolina at Greensboro

Dr. Edward Lee VargasSanta Fe Public Schools Santa Fe, NM

Review Board

Dr. Kenneth M. BirdWestside Community Schools Omaha, NE

Dr. Leonard C. BurrelloIndiana UniversityBloomington, IN

Dr. Jean B. CrockettVirginia Tech • Blacksburg, VA

Dr. Pia DurkinBoston Public Schools Dorchester, MA

Dr. Margaret E. GoertzUniversity of Pennsyvania Philadelphia, PA

Ms. Charlene A. GreenClark County School District Las Vegas, NV

Dr. Susan Brody HasaziUniversity of Vermont Burlington, VT

Dr. Robert HendersonUniversity of IllinoisUrbana-Champaign, IL

Dr. Dawn L. HunterChapman University Orange, CA

Dr. Shirley R. McBrideCanadian GovernmentVictoria, BC

Dr. Harold McGradyDivision of Learning DisabilitiesArlington, VA

Dr. Judy MontgomeryChapman UniversityOrange, CA

Dr. Michael OpudaMaine Department of Education Augusta, ME

Dr. Tom ParrishAmerican Institutes For Research Palo Alto, CA

Dr. Ted RiggenBarry Town School • Barry, VT

Dr. Kenneth E. SchneiderOrange County Public Schools Orlando, FL

Dr. Thomas M. SkrticUniversity of Kansas Lawrence, KS

Dr. Martha ThurlowNational Center on Educational Outcomes, University of Minnesota • Minneapolis, MN

Dr. Deborah A. VerstegenUniversity of Virginia Charlottesville, VA

Dr. David WoodAurora Public SchoolsAurora, CO

Dr. Jim YatesUniversity of Texas at Austin Austin, TX

Editorial Board

The Editorial MissionThe primary goal of the Journal of Special Education Leadership is to provide both practicing administrators and researchers of special education administration and policy with relevant tools and sources of informationbased on recent advances in administrative theory, research, and practice. The Journal of Special EducationLeadership is a journal dedicated to issues in special education administration, leadership, and policy issues. It is a refereed journal that directly supports CASE’s main objectives, which are to foster research, learning,teaching, and practice in the field of special education administration and to encourage the extension of special education administration knowledge to other fields. Articles for the Journal of Special EducationLeadership should enhance knowledge about the process of managing special education service delivery systems, as well as reflect on techniques, trends, and issues growing out of research on special education that is significant. Preference will be given to articles that have a broad appeal, wide applicability, and immediate usefulness to administrators, other practitioners, and researchers.

SubscriptionsThe Journal of Special Education Leadership is published by the Council of Administrators of Special Educationin conjunction with Sopris West. Copy requests should be made to Anne Miller, Sopris West, 4093 SpecialtyPlace, Longmont, CO 80504. Single copies may be purchased. Orders in multiples of 10 per issue can be pur-chased at a reduced rate. Members receive a copy of the Journal of Special Education Leadership as part of theirmembership fee. See back cover for subscription form.

AdvertisingThe Journal of Special Education Leadership will offer advertising for employment opportunities, conferenceannouncements, and those wishing to market educational and administrative publications, products, materi-als, and services. Please contact the editor for advertising rates.

PermissionsThe Journal of Special Education Leadership allows copies to be reproduced for nonprofit purposes without permission or charge by the publisher. For information on permission to quote, reprint, or translate material,please write or call the editor.

Dr. Mary Lynn Boscardin, EditorJournal of Special Education Leadership175 Hills-SouthSchool of EducationUniversity of MassachusettsAmherst, MA 01003

CopyrightThe Journal of Special Education Leadership, a journal for professionals in the field of special education adminis-tration, is published by the Council of Administrators of Special Education in conjunction with Sopris West tofoster the general advancement of research, learning, teaching, and practice in the field of special educationadministration. The Council of Administrators of Special Education retains literary property rights on copy-righted articles. Any signed article is the personal expression of the author; likewise, any advertisement is theresponsibility of the advertiser. Neither necessarily carries CASE endorsement unless specifically set forth byadopted resolution. Copies of the articles in this journal may be reproduced for nonprofit distribution with-out permission from the publisher.

Published in partnership with

Sopris West4093 Specialty PlaceLongmont, CO 80504

Phone: (303) 651-2829Fax: (303) 776-5934 www.sopriswest.com

Journal of Special Education LeadershipVolume 12, Number 2

2Journal of Special Education Leadership 12(2) • November 1999

As I am writing this letter, a new year is commencing for elementary and secondary schools punctuated by afocus on assessment and accountability like never before. Over the past ten years, assessment has come toplay a very large role in the lives of both teachers and students. The results of large-scale assessments, in par-ticular, are quickly becoming measures of teacher competency and student performance. This issue of theJournal of Special Education Leadership is dedicated to the topic of assessment and accountability as it relates tostudents with disabilities, extending beyond large-scale assessments and including curriculum-based mea-sures for determining instructional effectiveness.

Determining who should be administered various assessments and who should receive accommodationsand modifications, which might include the use of alternate measures, appears to be one of the greatest chal-lenges facing school districts today. Many have asked why it is important to include students with disabilitiesin the process at all. The key word is include. For too long students with disabilities have been excluded fromthe opportunity to participate in various forms of assessment, sometimes out of concern for the student andsometimes out of fear of what lower test score averages might do to districts or individual schools. Raisingthe bar was very necessary as special education programs slowly evolved through three phases (remediation,maintenance, and deterioration) when expectations for students with disabilities began to decline. One wayto raise that bar and let parents, teachers, and members of the community know about the seriousness of thisinitiative and to communicate that lower expectations for students with disabilities was unacceptable was toinclude all students in large-scale assessments. All students need to start each new school year with the samehopes, advantages, and promises.

The articles in this issue of JSEL consider a wide range of issues related to assessment and instructionalaccountability. Dr. Martha Thurlow and Sandra Thompson investigate the impact that district and state stan-dards have on building inclusive accountability systems. Drs. Gerald Tindal, Robert Helwig, and KeithHollenbeck provide a review of the investigations they have performed over the years on the use of curricu-lum-based measures for determining the effectiveness of instructional practices. Drs. Preston Green andStephen Sireci focus their discussion on the legal and psychometric issues affecting the assessment of stu-dents with disabilities. Dr. Thomas Hehir presents five strategies that research indicates are important toimprove the educational results of students with disabilities. Dr. David Chard offers a case example for CASEin Point to illustrate the impact of large-scale assessment on students with disabilities, and he argues for IEPteams assuming a greater role in the process of deciding the appropriate assessment procedures for studentswith disabilities. Dr. Edward Lee Vargas addresses the need for superintendents to surround themselves withskilled administrators of special education to more effectively meet expectations for higher levels of achieve-ment and accountability for all students.

On behalf of the members of the CASE Executive Committee and myself, we hope you enjoy this issue.Please do not hesitate to contact us with your comments and feedback.

Mary Lynn Boscardin, Ph.D., [email protected]

A Letter from the Editor

District and State Standards and AssessmentsBuilding an Inclusive Accountability System

Martha L. Thurlow, Ph.D. and Sandra J. ThompsonNational Center on Educational Outcomes, University of Minnesota

• Administrators need to promote the inclusion of students with disabilities in assessment and in accountability systems for these reasons: for an accurate picture of education; to make accurate comparisons; for students with disabilities to benefit from reforms; to avoid unintended consequences of exclusion; to meet legal requirements; and to promote high expectations.

• Reasons given to exclude students from district and state assessment include: they need accommodations that are not allowed; the tests are too hard and students will become frustrated;and guidelines about participation are inadequate.

• To increase participation, administrators need to provide: all students with opportunities to learn the curriculum; accommodations; information about assessment to parents and guardians;students with test preparation; professional development for educators.

3Journal of Special Education Leadership 12(2) • November 1999

“Educational accountability” is a phrase thatappears regularly in today’s media. The public

wants to know who is to be held accountable for theextent to which students are profiting from theireducational experiences. Educators and parentswant to know who is accountable, especially for thestudents who receive low scores on district andstatewide assessments. Although many elements cango into accountability determinations (e.g., drop-outrate, absenteeism), one of the more commonaccountability elements today is student perfor-mance. And, student performance is primarily basedon test results.

Tests are administered for a variety of reasons.While special educators are very familiar with theuse of tests to determine whether a student is eligi-ble for special education services or to measure astudent’s progress, they are less familiar with large-scale assessments—those used for school, district, orstate accountability or to make decisions aboutwhether a student will graduate or be promoted tothe next grade in school. With recent changes in thegeneral education context, including a renewedemphasis on higher standards and the use of assess-ments to determine consequences assigned to students, educators, and educational units, it isessential that special education leaders understandstandards and assessments. It is also critical that

they understand what makes for an inclusiveaccountability system. Special education leadershave an important role to play in ensuring that gen-eral education and special education personnel worktoward building an inclusive accountability system.

In this article, we highlight several of the concepts that undergird an inclusive accountabilitysystem, examining both their current status and theways in which they need to change if we are tomove toward truly inclusive accountability.

Current Status of Standards andAssessment in the United StatesUsing assessments for accountability purposes ispart of a national movement referred to as “standards-based reform.” This movement promotessetting high educational standards in several contentareas, identifying indicators of successfully meetingthose standards, and measuring student progresstoward the indicators. District and statewide assess-ments provide information for the accountabilitysystem. Policymakers and administrators use theresults of assessments to develop educational policyand design improved instructional programs. Manydistricts and most states now have requirements toproduce data to use as evidence that schools aredoing what they are expected to be doing, that is,

4Journal of Special Education Leadership 12(2) • November 1999

meeting educational standards. In fact, more thantwo thirds of the states are emphasizing the establishment or revision of content standards orcurricular frameworks, with a similar number ofstates emphasizing the alignment of their statewideassessments with standards and curriculum frame-works (Erickson & Thurlow, 1998).

There are two types of assessments that districtsand states typically use in their accountability sys-tems. Norm-referenced tests (NRTs) compare studentperformance to that of a local or national (standard-ized) sample. Criterion-referenced tests (CRTs) eval-uate student performance relative to state or districtstandards. NRTs are used by 31 states; 33 states useCRTs and, of these, 19 states use both (Bond, Roeber,& Connealy, 1998). Although both NRTs and CRTsare used in accountability systems, they provide different kinds of information. Furthermore, they aredifferentially conducive to building inclusiveaccountability systems.

Students who are tested count in the informationused by policymakers to make decisions about cur-riculum; allocation of resources; and development ofschool, district, and state policies about the instruc-tional process. Unfortunately, not all students areassessed. Typically, students who are excluded arethose who are expected to receive low scores.Students who are not assessed may not count inimportant educational decisions, even though thedecisions affect their instruction. Researchers havefound that students with disabilities and studentswith limited English proficiency often are excludedfrom both assessment and accountability systems,especially when reports are released to the public(Thurlow, Langenfeld, Nelson, Shin, & Coleman,1998; Vanderwood, McGrew, & Ysseldyke, 1998;Ysseldyke, Thurlow, Langenfeld, Nelson,Teelucksingh, & Seyfarth, 1998).

There is great variability among states in the rateat which some groups of students, especially thosewith disabilities and limited English proficiency, par-ticipate in testing, and an overall lack of reliableinformation on the extent of their participation. Inthe past, it was common for many states to excludestudents with disabilities from testing (e.g., studentswere sent on a field trip on the testing day), or tohave their tests flagged and the results destroyedwithout being reported. It has been estimated that

approximately eighty-five percent of students withdisabilities are able to participate in district and stateassessments with or without accommodations(McGrew, Thurlow, & Spiegel, 1993; Ysseldyke,Thurlow, Vanderwood, & Shriner, 1994). Therefore,the majority of students with disabilities should par-ticipate in general assessment systems. A major chal-lenge for school administrators is to demonstrateaccountability for the performance of all students,and this must include students with disabilities.

Establishing High Standards and Aligned Assessment SystemsThere are several federal laws that have promptednational activity in this arena and that carry it to thestate and local level. These include Goals 2000:Educate America Act, Improving America’s Schools Act,School-to-Work Opportunities Act, and the 1997 reauthorization of the Individuals with DisabilitiesEducation Act. Provisions within each of these lawspromote and support a variety of state, district andschoolwide activities intended to promote qualityeducational opportunities for all students. The 1997reauthorization of IDEA is very specific in itsrequirements about the participation of studentswith disabilities in state and district assessments. Inaddition, the assessment provisions in IDEA requirethat information on participation in assessments andperformance be reported publicly in the same wayand with the same frequency as reports are given onstudents without disabilities.

Reasons to Include Students with Disabilities in District and State AssessmentsThere are many important reasons for administratorsto promote the inclusion of students with disabilitiesin assessments and in accountability systems. A listis provided here of reasons that administrators canprovide to teachers and others who wonder why.

For an accurate picture of education. It is notpossible to get an accurate picture of a district orstate’s educational system if ten percent or more ofthe students are removed from the picture.

To make accurate comparisons. Once assessmentresults are reported, the reports are comparedamong schools, districts and states. In order for

Inclusive Accountability System

5Journal of Special Education Leadership 12(2) • November 1999

accurate comparisons to be made, all students needto be included by all schools, districts, and states. Ifnot, a district that tested less than fifty percent of itsstudents would probably show scores that are notcomparable to a district that tested nearly one hun-dred percent of its students.

For students with disabilities to benefit fromreforms. Often, the results of statewide assessmentsare used to make district and state educational policydecisions about the curriculum or resource allocation.When students are not included in assessments, theirperformance does not influence policy decisions thatmay affect their education. For example, includingthe scores of students with disabilities might reveallow reading performance scores, indicating a needfor continued reading instruction that might other-wise be dropped from the curriculum. Only byincluding all students can the educational systemaddress the needs of all students.

To avoid unintended consequences of exclusion.If there are significant consequences associated withassessments (e.g., school accreditation, student grad-uation), one of the unintended consequences mightbe increased referrals to special education. Anotherconsequence has been an increase in the number ofstudents retained in the grade preceding a test grade,presumably to keep them from being included intesting until they are more mature or ready for thetest. For example, if a statewide test is given to thirdgraders, some students who are expected to performpoorly might be retained in second grade.

To meet legal requirements. There are now sev-eral federal mandates that require the participation ofall students in statewide assessments. Primaryamong these is the Individuals with Disabilities Edu-cation Act (IDEA), the federal law governing specialeducation services, and the Improving America’sSchools Act (IASA), which supports Title I.

To promote high expectations. The most frequentreason students with disabilities are excluded fromtesting is that they are not expected to do well. Theselow expectations can pervade a student’s entire educa-tional career, resulting in lowered standards for what astudent is taught and expected to achieve. Low expec-tations are often related to low achievement.

Why Some Students are Excluded fromDistrict and State AssessmentsMany reasons are given for excluding students fromdistrict and state assessments, especially in stateswhere accountability systems have significant consequences for students, schools, or school sys-tems (Erickson & Thurlow, 1998). Three reasons arefrequently given.

Students need accommodations that are notallowed. An assessment accommodation is an alter-ation in the way a test is given. The purpose of anassessment accommodation is to allow students toshow what they know without the impediment of adisability. Generally, criterion-referenced tests andclassroom tests allow for more accommodations thannorm-referenced tests. Providing assessment accom-modations has been controversial for many reasons.Some of these reasons reflect attitudes about what isfair, while others seem to rest on technical concerns.Thus, people have argued that providing assessmentaccommodations invalidates what the test is tryingto measure, and that the scores produced are invalid.Furthermore, because there is not a universal list ofacceptable assessment accommodations, one statemay allow accommodations that another state pro-hibits (Thurlow, Seyfarth, Scott, & Ysseldyke, 1997).For example, the use of a calculator during testing isprohibited by five states and allowed by four states,and reading a test aloud is prohibited by nine statesand allowed by two states (Thurlow et al., 1997). Thefact that IEP teams make decisions about whichaccommodations a student will use and that thesedecisions are inconsistent from one team to the nextfurther complicates the provision of accommoda-tions. Still, at the bottom of much exclusion of stu-dents with disabilities from assessments is thecomplaint that students need accommodations thatsomeone has decided are not allowed.

The tests are too hard and students will becomefrustrated. Some people argue that assessments foraccountability are too difficult for some students,especially for those with disabilities and limitedEnglish proficiency. Test frustration or anxiety occursfor many reasons. Two common reasons are: (1) thestudent has not been taught or has not had the oppor-tunity to learn the content of the test; and (2) the stu-dent does not have adequate test-taking skills or is

Inclusive Accountability System

6Journal of Special Education Leadership 12(2) • November 1999

unfamiliar with the test format. Those who argue stu-dents will become too frustrated must consider whattest-taking skills the student has been taught and pro-vide training on those needed, and then ask what testaccommodations a student may need. Lack of instruc-tion should not be a reason for excluding a studentfrom a test. It is a reason for making sure the studentreceives the needed instruction.

Guidelines about participation in assessmentsystems are exclusionary or poorly written, and frequently coupled with high stakes that promoteexclusion. Without clear guidelines, poor decisionsare made about participation for individual students.Often, exclusion is taken as the easier and less com-plicated avenue to follow. The difficulties created byunclear guidelines are reflected in differential partic-ipation rates from one state to the next. When stateor district accountability systems place high stakesconsequences on schools or districts for the academicachievement of students, those students expected toperform poorly may be excluded, further complicat-ing the problems created by exclusionary or poorlywritten guidelines.

How to Include All Students in District and State AssessmentsIn the spirit of education reform, and to meet therequirements of federal laws and initiatives, districtsand state administrators are implementing ways toincrease the participation of all students in districtand state assessment systems. These include:

Provide all students opportunities to learn thecurriculum. Content-related concerns are connectedto what students are learning. It is important to askwhy a student is not learning the content that is pre-sented on the test. Has the student been exposed tothe content in the curriculum? What are the stu-dent’s educational goals? Only a small number ofstudents are unable to participate in regular assess-ments because of significant disabilities. These stu-dents can be assessed using an alternate assessment.

Alternate assessments are designed for studentswith disabilities who are unable to participate mean-ingfully in general large-scale assessments used by adistrict or state for accountability purposes. Alternateassessments provide a mechanism for students witheven the most significant disabilities to be includedin the accountability system. The percentage of these

students is quite small, estimated to range from lessthan one-half of one percent to no more than twopercent of the total student population. Thus, instates that have about ten percent of their studentswith disabilities receiving special education services,approximately five percent to twenty percent of stu-dents with disabilities would be expected to take analternate assessment.

The purpose of alternate assessments in stateassessment systems is not to address individualaccountability measures, but rather, to increase thecapacity of large-scale accountability systems thatcreate information on how a school, district, or stateis doing in terms of student performance. Amongassumptions that might be considered the founda-tion of alternate assessments (Ysseldyke & Olsen,1997) are that the alternate assessments focus onauthentic skills and on assessing experiences in com-munity and other real-life environments, and thatalternate assessments should measure integratedskills across domains. Despite these assumptionsabout best practice alternate assessments, it is clearthat performance on alternate assessments must beaggregated and reported with other scores.

Gathering data on the performance of studentswith disabilities through alternate assessmentsrequires some rethinking of traditional assessmentmethods. An alternate assessment system is neither atraditional large-scale assessment system nor an indi-vidualized assessment. Alternate assessments are ahybrid that can be administered to students whohave a unique array of educational goals and experi-ences and who differ greatly in their ability to sensestimuli, solve problems, and provide responses.

Gathering data on the performance of students with

disabilities through alternate assessments requires

some rethinking of traditional assessment methods.

The phrase “alternate assessment” appears in the assessment provisions of the reauthorizedIndividuals with Disabilities Education Act. IDEArequires all states to have alternate assessments inplace by July 1, 2000.

Provide accommodations. Accommodations forstudent learning and assessment need to begin in theclassroom. Teachers need to provide accommodations

Inclusive Accountability System

7Journal of Special Education Leadership 12(2) • November 1999

during daily instruction, including classroom assess-ments. Once instructional accommodations are estab-lished, educators must decide which ones make themost sense to provide on a district or statewideassessment. States are required by federal law to pro-vide needed accommodations to students with IEPsor 504 accommodation plans. Several states nowhave policies in place that provide accommodationsto any student needing them.

Educators need to make decisions carefully;forms to guide accommodation decisions can behelpful (see Thurlow, Elliott, & Ysseldyke, 1998). It isimportant to identify accommodations starting froma student’s needs rather than from what the state ordistrict allows. (See Figure 1 at the end of this arti-cle.) Whenever possible, students should be includedin discussions about accommodations they will use,so that they can provide input to the process of mak-ing accommodation decisions.

It is important to identify accommodations starting

from a student’s needs rather than from what the

state or district allows.

It is important also for district administratorsand/or test coordinators to stay up-to-date on theaccommodations allowed for each district orstatewide assessment. Many states post currentassessment information on their education agency’swebsite. Links to state websites are availablethrough the Council of Chief State School Officers athttp://www.ccsso.org/seamenu.html

School administrators and local assessment per-sonnel should also keep track of the consequences ofusing specific accommodations. For example, using acalculator may be listed as an accommodation, butthe scores of students using calculators may be unex-plainably low, perhaps suggesting the need for addi-tional use of the accommodation in the classroom.Or, a student may not be using an accommodationeven though the IEP team has determined that it isneeded (e.g., extended time), perhaps suggestingeither that the accommodation is not needed, or thatit must be provided with another accommodation(e.g., individualized testing).

District level administrators need to make surethat all district and state guidelines on assessment

accommodations are available and clearly under-stood. Furthermore, it is important to know thenumber and nature of accommodations needed inspecific locations so that the logistics of who doeswhat, when, and where can be delineated. Differentaccommodations require different levels of prepara-tion. For example, much less preparation is neededfor a student requiring a large print test booklet thanis required for a student who needs to be tested in astudy carrel. A written timeline and other formsdetailing who needs what accommodations forwhich tests in what schools or classrooms, and whowill see that they are provided, helps keep everyoneinformed about responsibilities and expectations.

To the extent that administrators maintain con-sistent and clear documentation of how accommoda-tion decisions are made and when and where theyare provided, there will be decreased likelihood ofcomplaints. There will however, be lawsuits chal-lenging the declaration that certain accommodationsare not allowed on district and statewide assess-ments, especially when the stakes are high either forstudents or for systems. In facing these challenges itis important for school administrators to have asound, written decision-making process in place thatclearly defines accommodations that are allowedand not allowed on each test, with a rationale thateveryone involved can understand.

Provide information about assessments to parents and guardians. Make sure that parents/guardians understand the purpose of the tests. These individuals may not have access toinformation that helps them fully understand schooldistrict or state assessment systems, especially infor-mation on their purpose, accommodation policies,reporting procedures, and use of results. If parentsare to become partners in decisions about theirchild’s participation in assessment systems, theymust be provided with information and opportuni-ties to learn about inclusive assessment and account-ability. Parents can play a pivotal role in advocatingfor the inclusion of all students in assessments, butonly if they are provided information on how partic-ipation in testing can lead to higher expectations,broader curricular offerings, and ultimately, betterresults for students.

Prepare students for testing. Talk to studentsabout tests, their feelings, your expectations of them,

Inclusive Accountability System

8Journal of Special Education Leadership 12(2) • November 1999

and their expectations of themselves. Be positive andempathetic. Ensure that students know the purpose ofthe assessments and why it is important for them totake the tests. Thurlow, Elliott, and Ysseldyke (1998)recommend keeping a list of reasons why studentsshould participate in testing (rather than why not),keep adding to it, and have students add to it. Thismay help dispel myths about possible negative effectsof testing. Help students understand their personalstrengths, weaknesses, and preferred learning styles.Students who understand their challenges and areable to focus on their strengths have higher self-esteemand are able to deal with day-to-day pressures betterthan students who have resigned themselves to con-stant failure in all parts of their lives. Keep scores private. Don’t announce the names of students who“failed,” or put all of the low-scoring studentstogether, or make them stand out in any other way(e.g., one school posted all of the names of studentswho would have to take a test over again). Each per-son is working toward personal goals, even when tak-ing large-scale assessments—keep expectations high,but individualized, respectful, and private.

Provide professional development. Administra-tors and educators need professional development onassessment and accountability. Information needsrange from understanding the purpose and types of assessments, and how to include all students inassessments, to making the instruction-assessment-accommodation link, to developing alternate assess-ments. Preservice training programs are a vital link inthe process of change. Schools must link with highereducation institutions. Preservice teacher courses needto include not only best practices in curriculum,instruction, and changes in the law, but how teachersimplement these instructional practices within theday-to-day reality and pressures of assessment andaccountability in schools.

Respectful ReportingReporting the scores of students with disabilities isan important element of inclusive accountability systems. IDEA requires public reporting of scores of students with disabilities with the same frequencyas other students. Administrators need to think abouthow best to present and address the scores that arereported. The average scores of student with disabili-

ties, as a group, are lower than those of other stu-dents, even though the range of scores is similar(Trimble, 1998; Ysseldyke et al., 1998). It is importantto have a plan in place for how to present thisinformation. For example, if the scores are reportedwithout accompanying contextual information, theadministrator may want to think about providingsome kind of information, or make clear that the dataare baseline data for looking at change over time.

Just as it is important to plan how to best reportinformation, it is important to plan what will be doneas a result of the information reported. For example,specific plans for providing remedial programs, strat-egy instruction, or test-taking skills might be made.Further, these plans can be reported alongside scoresto show that test results are driving instruction.

SummaryToday there is more activity surrounding educa-tional accountability than ever before. Nationally, thefocus has been on test scores. If a school, district, orstate is to know how all students are performing, wemust account for all of them. Accountability doesnot mean that all students have to take the sameassessments. It does mean that all students must becounted and included in the accountability reportsof educational programs.

Leaders in special education need to make surethat all students are included in assessment systemsthat are used to measure educational accountability.This can be done by: (1) critiquing assessment policies;(2) making sure that appropriate accommodations areallowed for use by students who need them; (3)including the test results for all students when report-ing on school or district performance; and (4) usingthe results for planning and improvement efforts.

Questions and issues surrounding assessment andaccountability abound. There are few easy answers,but we do know that the learning and progress of allstudents has risen to the top of the nation’s educationagenda. Benefits to students with disabilities in aninclusive accountability system include higher expec-tations resulting from a broader, more inclusive sys-tem of general education. An inclusive systemincorporates the unique learning styles of every stu-dent along with schools that celebrate the diversityand individuality of each child. Inclusive assessment

Inclusive Accountability System

9Journal of Special Education Leadership 12(2) • November 1999

Inclusive Accountability System

From: Elliott, J., Ysseldyke, J., Thurlow, M., & Erickson, R. (1997). Providing assessment accommodations for students with disabilities in state and district assessments. (NCEO Policy Directions 7). University of Minnesota: National Center on Educational Outcomes.

Figure 1: Examples of Accommodations and Decision-Making Questions

Examples: • Provide test on audio tape. • Increase spacing between items or reduce items per

page or line.• Highlight key words or phrases in directions. • Provide cues (e.g., arrows and stop signs) on

answer form.

Questions:• Can the student listen and follow oral directions?• Can the student see and hear? • Can the student read?

Examples: • Administer the test in a small group or individually

in a separate location with minimal distractions.• Provide special lighting.• Provide special furniture or acoustics.

Questions:• Can the student focus on his or her own work in a

room with other students? • Does the student display behaviors that are

distracting to other students? • Can the student take the test in the same way as it is

administered to other students?

Setting Presentation

Examples: • Allow marking of answers in booklet. • Tape record responses for later translation.• Allow use of scribe. • Provide copying assistance between drafts.

Questions:• Can the student track from a test booklet to a test

response form?• Is the student able to manipulate a pencil or other

writing instrument?

Examples:• Allow a flexible schedule. • Extend the time allotted to take the test. • Tape record responses for later translation.• Allow frequent breaks during testing.

Questions:• Can the student work continuously for the entire length of

a typically administered portion of the test (e.g., 20 to 30minutes)?

• Does the student use accommodations that require more time to complete test items?

Timing Response

Examples: • Special test preparation • Any accommodation that a student needs that does not

fit under the existing categories

Questions:• Is this the first time that the student will be taking a

district or state assessment?• Does the student have the necessary test-taking skills?

Examples: • Administer the test in several sessions, possibly over

several days, specifying the duration of each session.• Allow subtests to be taken in a different order.• Administer the test at different times of day.

Questions:• Does the student take medication that dissipates over

time, with optimal performance at a certain time of day?• Does the student’s anxiety level increase dramatically

when working in certain content areas, so that theseshould be administered after other content areas?

Scheduling Other

10Journal of Special Education Leadership 12(2) • November 1999

Inclusive Accountability System

and accountability can become a reality only wheneducators and parents understand the purpose ofaccountability systems, the use of accommodations,and the need to provide all students with opportuni-ties to achieve the same high standards.

ReferencesBond, L., Roeber, E., & Connealy, S. (1998). Trends in state

student assessment programs. Washington, D.C.: Councilof Chief State School Officers.

Erickson, R. & Thurlow, M.L., (1998). 1997 State specialeducation outcomes: a report on state activities during edu-cational reform. Minneapolis: University of Minnesota,National Center on Educational Outcomes.

McGrew, K., Thurlow, M., & Spiegel, A. (1993). An investi-gation of the exclusion of students with disabilities innational data collection programs. EducationalEvaluation and Policy Analysis, 15(3), 339-352.

Thurlow, M.L., Elliott, J.L., & Ysseldyke, J.E. (1998). Testingstudents with disabilities: Practical strategies for comply-ing with district and state requirements. Thousand Oaks,CA: Corwin Press.

Thurlow, M.L., Langenfeld, K.L., Nelson, J.R., Shin, H., &Coleman, J.E. (1998). State accountability reports: Whatare states saying about students with disabilities?(Technical Report 20). Minneapolis: University ofMinnesota, National Center on Educational Outcomes.

Thurlow, M.L., Seyfarth, A., Scott, D., & Ysseldyke, J.E.(1997). State assessment policies on participation andaccommodations for students with disabilities: 1997 update.(Synthesis Report 29). Minneapolis: University ofMinnesota, National Center on Educational Outcomes.

Trimble, S. (1998). Performance trends and use of accommoda-tions on a statewide assessment: Students with disabilitiesin the Kiris On-Demand Assessments from 1992-93through 1995-96. (Maryland/Kentucky Report 3).Minneapolis: University of Minnesota, NationalCenter on Educational Outcomes.

Vanderwood, M., McGrew, K., & Ysseldyke, J.E. (1998).Why we can’t say much about the status of studentswith disabilities in education reform. ExceptionalChildren, 64, 359-370.

Ysseldyke, J.E. & Olsen, K. (1997). Putting alternate assess-ments into practice: what to measure and possible sources ofdata. (Synthesis Report 28). Minneapolis: University ofMinnesota, National Center on Educational Outcomes.

Ysseldyke, J.E., Thurlow, M.L., Langenfeld, K.L., Nelson,J.R., Teelucksingh, E., & Seyfarth, A. (1998).Educational results for students with disabilities: What dothe data tell us? (Technical Report 23). Minneapolis:University of Minnesota, National Center onEducational Outcomes.

Ysseldyke, J.E., Thurlow, M.L., Vanderwood, M. &Shriner, J. (1994). Recommendations for making decisionsabout the participation of students with disabilities instatewide assessment programs. (Synthesis Report 15).Minneapolis: University of Minnesota, NationalCenter on Educational Outcomes.

Sidebar 1: 10th grader: “Man, I flunked the big testagain. I only get two more tries and if I don’t pass,I’m not going to graduate.” Teacher #1: “I don’tthink you’ll ever pass–why don’t you just stayhome the next time the test is given so you don’tpull the school’s average down?” Teacher #2: “No,you deserve the opportunity to graduate from highschool with your friends. Let’s sit down and makea plan to help make sure you have a chance to learnall of the material the test covers, and that you havethe accommodations you need so that we test whatyou know and not your learning disability.”

Sidebar 2: School district administrator: “I calledyou all together to talk about what we should doabout our low test scores. You know, if we don’tlook better this year, the state is going to cut ourfunding.” Respondent #1: “Well, it’s those darnspecial ed. and LEP kids that are pulling our aver-age down. Why don’t we just pull their scoresout?” Respondent #2: “No, those two groups makeup almost twenty percent of the students in ourdistrict. I think we should take a hard look at whatwe are teaching all of our students and make sureour curriculum is in line with state standards,which is what the tests are based on. Then we needto make sure all of our students have access to thecurriculum and can use accommodations to levelthe playing field.”

About the AuthorsMartha Thurlow, Ph.D., is the director of theNational Center on Educational Outcomes (NCEO),NCEO is supported through a cooperative agree-ment (#H159C50004) with the U.S. Department ofEducation, Office of Special Education Programs.Sandra Thompson is a research associate in theDepartment of Educational Psychology, at theUniversity of Minnesota. E-mail: Martha Thurlow [email protected] and Sandy Thompsonat [email protected]

11Journal of Special Education Leadership 12(2) • November 1999

Given inclusion of students with disabilities inlarge-scale assessments, a number of questions

have begun to emerge about how tests can be mademore accommodating for them and yet maintainconstruct validity for making decisions about out-comes and accountability. Tindal (1998) framed theissue around task comparability and proposed threestrategies for justifying an accommodation. Phillips(1994) posed five questions and maintained thatanswering any one of them as “yes” meant that thechange was more than an accommodation. Fuchs(1999) described an accommodation as the boost inperformance for students with disabilities thatexceeded a set amount attained by students in gen-eral education. All three perspectives have the gen-eral aim to avoid access problems, level the playingfield, and ensure integrity in decision making. Theyalso differ in the degree to which they are imple-mented and evaluated. In this article, a perspectiveis presented that uses empirical data to determinewhether an accommodation is both necessary andeffective. First, a framework is described that inte-grates these three approaches to accommodation val-idation. Second, a program of research is presentedin multiple choice mathematics testing. Third, a pro-gram of research in writing assessment is presented.In both mathematics and writing assessment, the

findings are extrapolated to help connect practicewith policy.

Tindal (1998) argues that accommodations can bejustified in terms of task comparability and articu-lates three strategies as having been used in makingthis determination. If two tasks are judged compara-ble (one of which represents a change from a stan-dard test administration), then the change wouldserve as an appropriate accommodation. If two tasksare judged noncomparable, the change would not bean accommodation but a modification. The threestrategies for making this determination, reflectingstate and district policies, can be compared to eachother in the kinds of evidence they utilize. The firstand most prominent strategy is descriptive with theonly evidence being prima facie documentation inpolicy. If the policy describes two tasks as compara-ble, then the accommodation is justified. Usually thisform of evidence is simply documentation of testchanges in a list of allowable changes or in a formalpolicy document. The major problem with this strat-egy is that “evidence” is authoritarian and tautologi-cal. The second type of evidence is an analysis of useand possibly outcomes, using post hoc programevaluation. For example, data are collected on howmany students receive an accommodation and howtheir performance is influenced comparing it to

An Update on Test Accommodations Perspectives of Practice to Policy

Gerald Tindal, Ph.D., Robert Helwig, Ph.D., and Keith Hollenbeck, Ph.D.University of Oregon

• With the current increase in the use of accommodations in testing, it is critical to evaluatetheir impact using experimental models rather than simply post hoc evaluations.

• Accommodations to large-scale testing are more than just leveling the playing field and need to be con-sidered as changes in the way tests are given and taken that are centered on the construct validity of thedecisions being made.

• When thinking about changes in testing, specific attention needs to be given to the influence of well-defined tasks on individual students.

• In both math and writing, the findings from our experimental research have validated specific changes inthe way tests are given and taken, and brought into question the construct validity of the measurementdecision-making system.

12Journal of Special Education Leadership 12(2) • November 1999

other (similar) students who did not receive theaccommodation. The problem with this strategy isthat incorrect interpretations can be made givennoncomparability of student populations and no infor-mation about implementation of the change. The thirdstrategy involves conducting an actual experiment inwhich students are randomly assigned to treatmentsand a control group used to document outcomes. Bothgroup design studies and single case studies are possi-ble in justifying whether two tasks are comparable.While difficult to implement, this strategy is the onlyone to advance the field of testing to rigorous conclu-sions with any kind of validity. Typically, an interactionis expected in that the treatment (method of test admin-istration) appears effective for the experimental groupbut not for the control group.

Phillips (1993, 1994) established five questionsthat need to be answered no for an accommodationto be considered appropriate.• Will format changes or alterations in testing con-

ditions change the skill being measured?• Will the scores of examinees tested under stan-

dard conditions have a different meaning from scores for examinees tested with the requested accommodation?

• Would examinees without disabilities benefit ifallowed the same accommodation?

• Does the examinee with a disability have anycapability for adapting to standard test adminis-tration conditions?

• Is the disability evidence or testing accommoda-tions policy based on procedures with doubtfulvalidity or reliability?Finally, Fuchs (1999) has described an accommo-

dation as justified when students with disabilitiesperform at least one standard deviation higher witha change in testing than students without disabilitiesreceiving the same test changes. She also anchorsoutcomes to a decision-making system for awardingan accommodation, comparing student performanceon curriculum-based measures (CBM) with teachers’recommendations. In her analysis, it is possible to

track the different decisions (by teachers with andwithout CBM data) to document false positives andfalse negatives.

All three systems for justifying accommodationsare similar in their use of differential outcomes:Accommodations should work for those who need itand should be neutral for students who do not needit. Thus, the main indicator that two tasks are com-parable is the degree to which a change in test scoreor performance is affected. Beyond that similarity,several differences are noteworthy. Tindal (1998)includes program evaluation information for judgingcomparability, though only with considerable cau-tion. Phillips (1994) poses several peripheral ques-tions, some of which require sophisticatedjudgments. Finally, Fuchs (1999) also uses CBM datato partially validate the need for an accommodation.

Accommodations should work for those who need it

and should be neutral for students who do not need it.

As the research on test accommodations becomesmore complete and studies begin to accumulate (seeTindal & Fuchs, 1999 for a review of research onaccommodations), patterns of findings can be assem-bled. It is this program of research that is neededbefore test changes truly can be validated. In theremainder of this paper, we use Messick’s (1989)two-by-two matrix for understanding the constructvalidity of score meaning and social values. Onefacet is the function or outcome from testing, whichaddresses how the test is to be interpreted or used.The other facet is the source of justification for test-ing, which uses either evidence to understand scoremeaning or values to understand consequences.Crossing these two facets creates a four-celled tablehighlighting a unitary view of validity integratingmeaning and values. The matrix is designed to buildunderstanding in a progressive manner so that anyconstruct being measured is neither underrepre-sented nor with irrelevant variance. (See Figure 1.)

Test Accommodation: Perspectives of Practice to Policy

Test Interpretation Test Use

Evidential Basis 1. Construct Validity (CV) 2. CV + Relevance/Utility (R/U)

Consequential Basis 3. CV + Value Implications (VI) 4. CV + Soc. Conseq. (SC)

Figure 1. Facets of Validity as a Progressive Matrix

13Journal of Special Education Leadership 12(2) • November 1999

The evidential basis of test interpretation reflectsconstruct validity (CV) in addressing convergent-discriminant evidence; the focus of interpretation isprimarily scientific and empirical. The evidentialbasis of test use focuses on the construct validity(CV) of performance in applied settings with thebenefits of testing considered in relation to costs andrelevance/utility (R/U).

The consequential basis of test interpretation iscomprised of construct validity (CV) with referenceto broad theories and philosophical views, all ofwhich address value implications (VI) and becomeembedded within score meaning. This block often“triggers” score-based actions.

The consequential basis of test use considers con-struct validity (CV), relevance/utility (R/U) andvalue implications (VI), and potential as well associal consequences (SC) in applied settings, focus-ing on equity and fairness along with many otherbroad social interpretations, in a sense the functionalworth of the test.

The validation process in this progressive matrixis never definitive but proceeds in an iterative mannerwith evidence interacting with values to create scoremeaning and interpretations. The validation processincludes both evidence and value in an evaluativerhetoric of argumentation. Ultimately, validation isinterpretative, addressing both the meaning of scoresand their use in making decisions (Messick, 1995).

In the following two examples of research pro-grams, the mark of distinction is the emphasis onconstruct validity. Improving performance is not thesole criterion for justifying an accommodation.Rather, the very construct of what is being measuredis under scrutiny. In this process, three componentsare considered: (1) the task demands; (2) the scalingof behavior; and (3) the student being tested. In theend, these programs of research are based on multi-ple studies that build and interleave both previousmethodologies and subsequent findings.

Research on Mathematics Multiple Choice TestingThe focus of several Behavioral Research andTeaching (BRT) research projects over the past fewyears has been to better understand testing practicesboth in validating accommodations and at the same

time verifying construct validity of achievement testsin math that employ multiple choice formats.Through empirical research it is possible to presentobjective evidence on the effectiveness of variousaccommodations as well as present convergent evi-dence for test interpretation. In this section, we detailfour BRT studies done in mathematics. Our researchhas evolved from a concern with documentingchanges in performance with accommodations to aspecific focus on the interaction between studentskills and test demands and finally to investigatingthe robustness of test “standardization.”

In our initial research, we examined the effec-tiveness of a reading accommodation (Tindal,Heath, Hollenbeck, Almond, & Harniss, 1998). Onehundred twenty-two low achieving general educa-tion students and forty-two special educationfourth-grade students with IEPs in reading and/ormath were randomly assigned to one of two treat-ments. One group took a 30-problem mathematicsmultiple choice test under standard conditions. Theother group took a comparable version of the sametest with all test questions and item choices readaloud (with each item read twice). An overhead ofevery test item was used to highlight each word/line as it was read. We found that the performanceof general education students was not affected bythe mode of item presentation, while the specialeducation students significantly improved in theirperformance when the test questions and choiceswere read aloud.

...[T]he performance of general education students

was not affected by the mode of item presenta-

tion, while the special education students signifi-

cantly improved in their performance when the

test questions and choices were read aloud.

This study is important in that it identifies read-ing as an access skill in multiple choice mathematicstests. If students with disabilities and reading IEPsare to have their mathematics achievement docu-mented without bias, then read-aloud assistancemay be needed. On its surface, this differential effect,based on student status, appears to satisfy one of therequirements for a justifiable accommodation as put

Test Accommodation: Perspectives of Practice to Policy

14Journal of Special Education Leadership 12(2) • November 1999

forth by Tindal (1998), Phillips (1993, 1994), andFuchs (1999), i. e., accommodations should producean effect only for those students in need of a particu-lar change in test administration conditions. Weknow little, however, about identifying who mayneed this accommodation or why it is effective.Other than placement in special education and beingserved in reading, there is no other indication thatthere existed a particular need for an accommoda-tion for any specific student. By itself, an IEP inreading or math is an indirect indicator that a stu-dent needs a reading accommodation.

A follow-up study focused on documenting theeffectiveness of an accommodation for a particulargroup of students with a specific need (Helwig,Rozek-Tedesco, Heath, Tindal, & Almond, in press).Six items from a math test were identified as havingparticularly difficult reading levels. This determina-tion was based on a number of criteria includingword count, difficult vocabulary, and sentence struc-ture. As part of a larger study, these items were pre-sented both in standard test booklet format, as well avideo format as described below. Approximately 250subjects were randomly assigned to one of these con-ditions. Prior to testing, all students were given bothreading and basic math skill tests. Based on theseresults, a subset of students was identified as beingmost in need of an accommodation. This groupincluded only students with average and abovemath skills who concurrently were reading signifi-cantly below grade level. It was reasoned that onlylow skilled readers would require items being readaloud. Within this group of low readers, however, itwas hypothesized that only those that possessed abasic level of math skill would be able to benefitfrom the accommodation. In other words, hearingthe problems read aloud would enable low readersto understand what was being asked, but those withadditional deficiencies in math skill would not havethe tools to solve the problems. Our hypothesis wascorrect. On four of the six test items, the targetedgroup (below grade level readers possessing averageand above math skills) performed better on thevideo version than on the standard version. For allother groups there was no difference. It appearedthat higher level readers, regardless of math skill,were not in need of the accommodation, while thelow reading and low math group lacked the requisite

math skill to translate reading access into successfulproblem solutions.

As a result of this research, we have established aclearer picture of student need. It is certainly benefi-cial to know that these accommodations do not affectthe performance of nontargeted groups. But, from apractical perspective, in terms of time, logistics, andcost, it is best to have tools that precisely identify apopulation of students for whom a specific accom-modation is needed. Also, by narrowing our definition of need (in this case a specific combinationof reading and math skills), and empirically compar-ing results from various testing conditions, we canbetter develop specific accommodation criteria andbe confident that a change in format is truly justified.

...[F]rom a practical perspective, in terms of time,

logistics, and cost, it is best to have tools that pre-

cisely identify a population of students for whom a

specific accommodation is needed

In another follow-up study we compared theeffectiveness of two different accommodations(Hollenbeck, Rozek-Tedesco, Tindal, & Almond,1999) on 180 middle school students. One version ofa multiple choice mathematics test was administeredusing a video of an actor reading all test questionsand item choices. The test was paced with each itembeing presented to a class of students; therefore, allstudents worked on the same problem at the sametime. Problem difficulty was used to prescribe theamount of time that students were given to solve theproblem before the next problem was presented viarelease of the remote control “pause” button.Another version was administered with a computerCD using nearly the same format as the videotape,with the same actor reading all test items. When tak-ing the computer version of the test, however, onlythe audio portion (not the video) was used and stu-dents were free to answer questions in any order,spend as much time as desired on any problem, andreturn to problems they had skipped. Resultsshowed no differences between the video and com-puter administration. Furthermore, no interactionwas found between type of administration and typeof student (with or without disability).

Test Accommodation: Perspectives of Practice to Policy

15Journal of Special Education Leadership 12(2) • November 1999

In our last study, we simplified the wording ofthe problems on a mathematics multiple choice test(Tindal, Anderson, Helwig, Miller, & Glasgow, 1999).Students from a middle school were given two ver-sions of a 25-item test. One version was adminis-tered in its standard form. On another version, thelanguage was simplified by (1) using only activevoice; (2) shortening or removing noun phrases; (3)separating sentences instead of using conditionalclauses; (4) reducing questions to simple “wh” for-mat; (5) personalizing tests with “you”; and (6) sub-stituting, where appropriate, shorter or more easilyunderstandable words. As in the earlier video/CDstudy, students participated in both administrationconditions with the order counterbalanced for halfthe group. Unexpectedly, no treatment effects werefound for either students with learning disabilities orthose in general education.

Although accommodations appeared to be inef-fective in these last two studies, we believe the find-ings establish the robustness of test administrationconditions. Specifically, this research suggests that cer-tain changes in standardized testing can occur with-out changing the score or invalidating interpretationsof the score. Furthermore, without having to docu-ment that a specific student is in particular need of anaccommodation, teachers can administer a math testby reading it, presenting the test via a video tape orcomputer, or using simplified language to focus onthe math skill. As a consequence, any student with anIEP in reading might be given the choice of having amathematics test read aloud without fear that if a stu-dent did not really need the accommodation, his/herscore would be unfairly inflated. Of course, not allaccommodations fall into this category and moreresearch is needed to document such reliance forother changes in test administration.

The work we have done in mathematics testaccommodations addresses both the evidential andconsequential aspects of validity, in how tests are tobe interpreted and used. Both the construct validityof math achievement is being addressed and theconditions for testing are being explicated. In bothinstances, policy is being informed.

We clearly must distinguish math from reading ifwe are to advance our theories of skill-specificachievement or to understand the cognitive interac-tions of two different symbol systems. From a practi-cal, classroom perspective, students may need to

have the test read to them if we want to documentmath proficiency using a multiple choice test. Thisaccommodation has the primary purpose of levelingthe playing field by providing students access to thetest. Otherwise, when students incorrectly respondedto standardized test items, “it is unknown whetherthey do so because of lack of knowledge or becauseof inability to successfully comprehend the test andits items” (Homan, Hewitt, & Linder, 1994, p. 349).As we develop supplemental classroom measures ofreading and computing that document student skillsand profiles, teachers may begin to better relate theirteaching to the testing and make better decisionsabout who needs what accommodation.

We clearly must distinguish math from reading if

we are to advance our theories of skill-specific

achievement or to understand the cognitive inter-

actions of two different symbol systems.

Yet, many teachers have little basis for recom-mending an accommodation, knowing little about the interaction between specific student characteristics and test demands. From our work onvideo presentations and computer read-aloud pre-sentations, we know, however, that artificially pacingstudents does not change their score. Similarly, fromour work on simplifying language, we see no changein performance over that attained under standardconditions. In both studies, the evidence indicatesthat these particular accommodations do not signifi-cantly change the construct of math problem solving(via multiple choice tests) and may help us rethinkour notions of “standardization.” By confining thisconcept to very narrow topographically similaradministrations, we may be denying students multi-ple access to the skill being tested. Probably the mostsignificant social consequence would be a betterintegration of testing into the educational practicesof teaching, which is unlikely to ever be narrowly“standardized” in the same manner as tests.

Writing Research InformingPolicy DecisionsOur six research programs on writing performanceassessment utilized the criterion of Fuchs (1999),

Test Accommodation: Perspectives of Practice to Policy

16Journal of Special Education Leadership 12(2) • November 1999

Test Accommodation: Perspectives of Practice to Policy

Messick (1989, 1995), Phillips (1993, 1994), andTindal (1998) to document and justify viable accom-modations. Principally utilizing Messick’s (1989,1995) validity arguments, the six research outcomesserved to inform policy decisions. The first threestudies (Helwig, Stieber, Tindal, Hollenbeck, Heath,& Almond, 1999; Hollenbeck, Tindal, Harniss, &Almond, 1999; Hollenbeck, Tindal, Stieber, &Harniss, 1999) analyzed the construct validity ofhandwritten versus word-processed essays. Thefourth study (Hollenbeck, Tindal, & Almond, inpress), analyzed the consequential basis of test useas it pertains to judges’ ratings of statewide essays.The last two studies (Hollenbeck & Tindal, 1999;Hollenbeck, Tindal, & Harniss, 1999) examined theevidential basis of test use as it relates to screen editing and writing time.

Helwig, Stieber, Tindal, Hollenbeck, Heath, andAlmond (1999) examined the comparative factorstructures (state identified writing traits) of eighthgrade handwritten and word-processed essays.Helwig et al.’s research supported the unidimension-ality of handwritten and word-processed essays.When handwritten or word-processed writing samples were analyzed by themselves a single factorwas found for each, but two factors were foundwhen handwritten and word-processed essays wereevaluated simultaneously. Taken together, the two-factor solution accounted for sixty-four percent of thetotal variance. Factor I consisted entirely of the word-processed traits, while Factor II consisted entirely ofthe handwritten traits. Helwig et al. noted that “theloadings of the six writing traits for each administra-tion exhibited a similar ranking in terms of magni-tude within each writing typology. Organization hadthe highest loading on both factors, while Voice andConventions had the two lowest loadings on eachfactor” (p. 12). Similar patterns held when subsets ofthe study population based on (1) teacher-rated profi-ciency; (2) type of computer; (3) type of writing; or(4) gender were factor analyzed.

Helwig et al.’s results raise validity issues withthe scoring systems in state-mandated writing assess-ments. These results show that judging individualtraits lacks empirical support and that validity gener-alizations would be strengthened by employingholistic rating scales. The between-group results also

suggest that handwritten essays and word-processedessays are noncomparable and, thus, not interchange-able, so they should be thought of as a modification,not an accommodation. (Tindal, 1998).

Besides factor structures, Hollenbeck, Tindal,Stieber, and Harniss (1999) analyzed the validity ofeighth grade word-processed essays as an accommo-dation for handwritten essays. They hypothesizedthat handwritten or word-processed responses to thesame prompt must create comparable score distribu-tions around equivalent means and standard devia-tions with the same degree of reliability to have thescores represent the same construct-of-interest.When the original handwritten compositions weretranscribed into a word-processed format and thenrated by state judges, the original handwritten com-positions were rated significantly higher than theword-processed composition on four of the six traitsfor the total group. When the ratings of the composi-tions that were handwritten were compared againstthe ratings assigned to the same essay that wasword-processed, significant differences for the totalsample were found on three writing traits: Ideas andContent (p = .0005), Organization (p = .0097), andConventions (p = .0002). While no significant differences were found for the traits of Voice (p =.3041) and Word Choice (p = .3690), the trait ofSentence Fluency approached significance (p =.0992). Further, collateral evidence was gathered bycomparing mean scores for each trait for practicalsignificance. The mean handwritten trait score washigher for five of the six traits for the total essays.

Hollenbeck, Tindal, Stieber, and Harniss’ (1999)results indicate that state raters graded word-processed essays more harshly. Results showed thatthe statewide scoring rubric lacked scale stabilitywhen applied to identical essays presented in differ-ent response modes. Instead of providing betteraccess that reveals student competence in the writingprocess, the word processing response mode appearsto punish students for using it. These rater con-founds negatively influence the essay scores so thatstudents using word processing are not accuratelyand fairly evaluated. If scores vary depending on thevehicle (response mode) used to create them, then itis not likely that scores can be interpreted meaning-fully (Dietel, Herman, & Knuth, 1991). Most importantly,

17Journal of Special Education Leadership 12(2) • November 1999

these findings suggest that word-processed essaysare not an appropriate accommodation for statewidewriting tasks (usually handwritten) because theresults violate Tindal’s (1998) task comparabilityrequirement and Phillips’ (1993, 1994) requirementthat standard administration scores have the samemeaning as the accommodated condition. Given theexisting evidence, the two response modes (writingcompositions by hand and writing them with com-puters) are not comparable and should not be usedin the same evaluation system.

Hollenbeck, Tindal, Harniss, and Almond (1999)investigated the construct validity of three computerwriting accommodations for eighth graders. Manyeducators assume that a word processing accommo-dation would enable students to pay attention to thewriting process because the computer executes thesimplistic, laborious secretarial tasks. The threegroups compared were (1) Computer Group—Students completed the entire test using a computer-based word processor (without the use of thespellchecker or grammar checker), from planning tofinal essay; (2) Computer, Last-Day Group—Studentscompleted the planning and draft stages using paper-pencil and then completed their final draft by trans-ferring their handwritten compositions to a computerusing a word processor (without the use of thespellchecker or grammar checker); and (3) Computer,Last-Day with Spellcheck Group—Students com-pleted the planning and draft stages using paper-pen-cil and then completed their final draft by transferringtheir handwritten compositions to a computer using aword processor (with the use of the spellchecker).

The students in the Computer Group had lowermean scores than either the Computer, Last-DayGroup or the Computer, Last-Day with SpellcheckGroup. For all six traits, the mean trait score for theComputer, Last-Day with Spellcheck Group washigher than the Computer, Last-Day Group, whichwas higher than the Computer Group. However, nosignificant differences were found for four traits:Ideas and Content (F(2,83) = .80, p = .4549),Organization (F(2,83) = 2.34, p = .1031), Voice(F(2,83) = .57, p = .5702), and Word Choice (F(2,83) =2.62, p = .0791). Significant differences were foundfor two traits: Sentence Fluency (F(2,83) = 3.53, p =.0337) and Conventions (F(2,83) = 3.54, p = .0336).

For both traits, the mean Computer Group score wassignificantly lower than the Computer, Last-Daywith Spellcheck Group; p = .0489 and p = .0361,respectively. The use of the computer spellcheckernot only significantly influenced the Conventions, italso influenced the writing traits of Organization,Word Choice, and Sentence Fluency. While spellingmay be thought of as a singular skill, its control ontext flow and cohesion cannot be understated, andone could deduce that correct spelling is a significantcomponent of Organization, Word Choice, andSentence Fluency. From the state policy perspective,allowing some students to use word processors withspellcheckers could significantly skew traditionalscore distributions and give those students an unfairadvantage. Furthermore, the spellchecker should bethought of as a modification because it creates non-comparable tasks (Tindal, 1998) and would allowexaminees without disabilities to benefit from theaccommodation (Phillips, 1993, 1994).

Hollenbeck, Tindal, and Almond (1999) investi-gated the validity consequences of rater decisions asit related to the high-stakes decision reproducibilityor consistency. They found that correlations betweenraters on the individual traits were moderate, with p-values for the handwritten essay trait scores significantly different on four of the six traits. Whilethe mean handwritten trait scores for Organization (p = .002), Voice (p = .0085), Sentence Fluency (p =.0351), and Conventions (p = .0009) were significantlydifferent, the traits of Ideas and Content (p = .4944)and Word Choice (p = .897) were not. Further, exactmatch statistics for individual traits never exceeded58.4%, and most correlations were in the .6 range.

Hollenbeck, Tindal, and Almond’s (1999) findingsindicated that the state raters have not achieved theconsistency necessary for high-stakes decisions,because “judgment is the crucial link in the chain ofreasoning from performance to inference about stu-dents” (Mislevy, 1994, p. 34). High-stakes assessmentis of little value if the stability of results from onejudge to another cannot be relied upon. Because thestatewide writing test has high-stakes consequences,valid generalizations from it are suspect. Exact matchstatistics were low enough to undermine the validityof the scores for comparing groups of students,schools, or districts across the state. Any standard is

Test Accommodation: Perspectives of Practice to Policy

18Journal of Special Education Leadership 12(2) • November 1999

only as meaningful as the reproducibility, reliability,and validity of the score interpretation allows.

Hollenbeck, Tindal, and Harniss (1999) examinedthe relevance and utility of computer accommodationsfor eighth graders—specifically whether the size ofthe computer screens predicted the students’ writingscores on a statewide writing test. They hypothe-sized that text editing is based upon the computertext viewing area. Logically, then, the size of thecomputer screen should greatly affect editing accu-racy and, thus, the scores on the final product.Larger computer screens should produce higheressay scores, whereas the smaller screen (a 32-character by 4-line text window) should hinder paragraph, page, and total text editing.

Overall, Hollenbeck, Tindal, and Harniss’ (1999)results of this study showed that a larger computerscreen did not significantly increase student writingscores. In fact, just the opposite was found. Of six traitdifferences only the traits of Ideas and Content [F(1,74)= 2.31, p = .1331], Voice [F(1,74) = .10, p = .7478],Organization [F(1,74) = 3.00, p = .0875] were nonsignifi-cant. Word Choice, Sentence Fluency, and Conventionssignificantly favored the Small Screen Group evenwhen controlling for student writing differences thatexisted prior to the study. For the group differencesthat were not significant (the traits of Ideas andContent, Organization, and Voice), the small screen andthe large screen were equally effective for students.

From a policy perspective, these results showthat before computer technology becomes useful as astatewide writing accommodation, students must betrained adequately to utilize the text editing featuresof the word processor. However, before the validityof word processors as a statewide writing testaccommodation can be established, concurrent valid-ity issues (like keyboarding skills, utilizing comput-ers in the writing process for classroom assignments,and instruction in the word processors’ editing func-tions) must be resolved. Furthermore, the resultsfrom Helwig et al. (1999) and Hollenbeck, Tindal,Stieber, and Harniss (1999) confound the notion thata computer could be an accommodation and not amodification (Phillips, 1993, 1994; Tindal, 1998).

...[B]efore the validity of word processors as a

statewide writing test accommodation can be estab-

lished, concurrent validity issues (like keyboarding

skills, utilizing computers in the writing process for

classroom assignments, and instruction in the word

processors’ editing functions) must be resolved

The last research project (Hollenbeck & Tindal,1999) evaluated the relevance and utility of complet-ing the statewide assessment for eighth graders intwo days versus three days. While the OregonStatewide Assessment Program Writing andMathematics Problem-Solving administration man-ual stipulates that “the writing assessment must beadministered on three consecutive days” (p. 1), themotivation for imposing the three day time limitmay be “to ensure that all examinees face the same,standardized task, and thus . . . the test is fair forevery test taker” (Powers and Fowles, 1996, p. 433).The question to be asked is whether this edict is sup-ported by empirical evidence.

Hollenbeck and Tindal (1999) found that somestudent groups write better with no stipulated timeconstraints. A significant difference was found forstudent status with general education students per-forming at significantly higher levels than studentsin special education (F=30.596[1,54], p<.0001). Nomain effect was found for two versus three days oftest administration duration. However, the interac-tion between student status and test duration wassignificant with special education students perform-ing significantly lower when writing for only twodays (F=7.276[1,54], p=.0093). However, no signifi-cant differences were found for either test durationor the interaction between student status and testduration. Overall, they discovered that whether stu-dents were given two or three days, no significantdifferences were found on most traits, except forIdeas and Content. For this particular trait, studentsbeing served in special education wrote composi-tions that were rated significantly lower when theyhad only two days to write while their general edu-cation counterparts were not so affected by one dayless. Otherwise, for all other traits, students’ quality

Test Accommodation: Perspectives of Practice to Policy

19Journal of Special Education Leadership 12(2) • November 1999

of composition was the same whether they wrote fortwo or three days.

The validity of score interpretation is paramountand is related to the amount of error present in themeasurement system. For Oregon’s writing assess-ment policy, the mandated three-day period may bemore valid for students with disabilities than for itsgeneral education students. Phillips’ (1993, 1994) andFuchs’ (1999) criteria suggest that Oregon already hasa valid accommodation in place because studentswith disabilities bettered their score by writing forthe third day while general education students didnot. Tindal (1998), however, would counsel that theform of the task would need to be considered beforejudging extra time as an appropriate accommodation.

SummaryIn this article, we have provided considerable dataon the effects of test changes, anchoring the out-comes to a definition of accommodation. We havefocused on several aspects of validity à la Messick(1989), relating test interpretations with test use andconnecting an evidential with a consequential basisfor making decisions. In some studies, we have oper-ationalized a critical attribute of an accommodationthat is based on individual need, not general benefit.In other studies, we have begun to re-examine thevery construct that is being measured. To the degreethat students are provided or denied opportunitiesbased on the results of standardized tests, we believesuch a program of research is necessary to informpolicy. Otherwise, with policy based on traditionalpractices, best guesses, or program evaluations, neg-ative outcomes are likely.

ReferencesDietel, R.J., Herman, J.L., & Knuth, R.A. (1991). What does

research say about assessment? [On-line]. Available:http://www.ncre.org/sdrs/areas/stw_egys/4assess.htm

Fuchs, L.S. (April 1999). Curriculum-based measurement:Updates on its application in standards-based assessment systems. Charlotte, NC: Council ofExceptional Children.

Helwig, R., Rozek-Tedesco, M., Heath, B., Tindal, G., &Almond, P. (in press). Reading as an access to mathproblem solving on multiple choice tests. Journal ofEducational Research.

Helwig R., Stieber, S., Tindal, G., Hollenbeck, K., Heath,B., & Almond, P. (1999). A comparison of factor analysesof handwritten and word-processed writing of middleschool students. Manuscript submitted for publication.

Hollenbeck, K., Rozek-Tedesco, M., Tindal, G., & Almond,P. (1999). A student-controlled versus teacher-controlledaccommodation for large-scale math tests. Manuscriptsubmitted for publication.

Hollenbeck, K. & Tindal, G. (1999). Accommodatingstudents on large scale writing tests: consideration of time in the writing process. Manuscript submitted for publication.

Hollenbeck, K., Tindal, G., & Almond, P. (in press).Reliability and decision consistency: An analysis ofwriting mode at two times on a statewide test.Educational Assessment.

Hollenbeck, K., Tindal, G., & Harniss, M. (1999). The influ-ence of computer screen size on statewide writing testscores. Manuscript submitted for publication.

Hollenbeck, K., Tindal, G., Harniss, M., & Almond, P.(1999). The effect of using computers as an accommo-dation in a statewide writing test. Manuscript submittedfor publication.

Hollenbeck, K., Tindal, G., Stieber, S., & Harniss, M.(1999). Handwritten versus word-processed statewide com-positions: do judges rate them differently? Manuscriptsubmitted for publication.

Homan, S., Hewitt, M., & Linder, J. (1994). The develop-ment and validation of a formula for measuring single-sentence test item readability. Journal ofEducational Measurement, 31, 349-358.

Mislevy, R.J. (1994). Test theory reconceived (CSE TechnicalReport 376). Los Angeles: UCLA, CRESST.

Messick, S. (1989). Validity. In R. Linn (Ed.), Educa-tional measurement (3rd ed.), pp. 13-105. New York: MacMillan.

Messick, S. (1995). Validity of psychological assessment:Validations of inferences from persons’ responses andperformances as scientific inquiry into score meaning.American Psychologist, 50, 741-749.

Oregon Department of Education (undated). Oregonstatewide assessment program writing and mathemat-ics problem-solving administration manual. Salem, OR:Author.

Phillips, S.E. (1993). Testing accommodations for disabledstudents. Education Law Reporter, 80, 9-32.

Phillips, S.E. (1994). High stakes testing accommodations:validity versus disabled rights. Applied Measurement inEducation, 7, 93-120.

Powers, D.E. & Fowles, M.E. (1996). Effects of ApplyingDifferent Time Limits to a Proposed GRE Writing Test.Journal of Educational Measurement, 33, 433-452.

Test Accommodation: Perspectives of Practice to Policy

20Journal of Special Education Leadership 12(2) • November 1999

Tindal, G. (1998). Models for understanding task compara-bility in accommodated testing. A publication for the Council of Chief State School Officers.Washington, D.C.

Tindal, G., Anderson, L., Helwig, R., Miller, S., & Glasgow,A. (1999). Accommodating students with learning disabili-ties on math tests using language simplification.Unpublished manuscript.

Tindal, G. & Fuchs, L. S. (1999). A summary of research ontest changes: an empirical basis for defining accommoda-tions: Volume I. Unpublished manuscript, Lexington,KY: Mid-South Regional Resource Center.

Tindal, G., Heath, B., Hollenbeck, K., Almond, P., &Harniss, M. (1998). Accommodating students withdisabilities on large-scale tests: an empirical study ofstudent response and test administration demands.Exceptional Children, 64 (4), 439-450.

About the AuthorsGerald Tindal, Ph.D., is an associate professor andRobert Helwig, Ph.D., and Keith Hollenbeck, Ph.D.,are research associates in the Department ofBehavioral Research and Teaching, 232 College ofEducation, University of Oregon, Eugene, OR 97403-5262. E-mail: Gerald Tindal at [email protected], Robert Helwig at [email protected], and Keith Hollenbeck [email protected]

Test Accommodation: Perspectives of Practice to Policy

21Journal of Special Education Leadership 12(2) • November 1999

An accountability movement pervades our educa-tional system. Educational policymakers, politi-

cians, and parents want to be sure our children arereceiving an education that will make them compet-itive in the global marketplace. Such assurancesrequire outcome measures, which are usually in theform of standardized tests. Standardized tests areused to evaluate teachers, schools, school districts,and states. They are also used to make decisionsregarding graduation and promotion. Twenty-fivestates currently require, or will soon require, stu-dents to pass a standardized test as a prerequisitefor graduation (Murphy, 1999), and several schooldistricts are considering using standardized tests asa prerequisite for grade promotion (Heubert &Hauser, 1999).

Standardized tests are designed to promote fair-ness in decision making by ensuring the contenttested and the test administration conditions areuniform for all test-takers. This uniformity providesa level playing field for comparing students to oneanother (norm-referenced testing) and for determin-ing how well students have learned specific subjectmatter (criterion-referenced testing). Although this“one size fits all” approach to assessment is suitablefor most of the tested population, it has been criti-cized when applied to students with disabilities.Many students with disabilities are unable to

display their true knowledge and skills when takingtests under standardized conditions. For this reason,standardized tests are usually accommodated whentesting students with disabilities. However, an“accommodated standardized test” is an oxymoron.Thus, it is not surprising that valid assessment ofstudents with disabilities is a complex and confus-ing process.

Although this “one size fits all” approach to assess-

ment is suitable for most of the tested population,

it has been criticized when applied to students with

disabilities.

Professionals involved in educating studentswith disabilities are struggling to keep up withfederal and state laws regarding the inclusion ofstudents with disabilities in statewide and nationaltesting. The testing of students with disabilitiesraises legal questions, such as the extent to whichschools must provide testing accommodations.Additionally, psychometricians have long pointedout that test scores from accommodated tests maynot be comparable to scores from the original, standardized test (American Educational ResearchAssociation (AERA), American Psychological Association(APA), National Council on Measurement in

Legal and Psychometric Policy Considerations inthe Testing of Students With Disabilities

Preston C. Green, Ed.D., Esq., and Stephen G. Sireci, Ph.D.University of Massachusetts at Amherst

• Complex legal and psychometric issues surround the testing of students with disabilities.

• Special education administrators must be familiar with the laws regarding provision of test accommodations to students with disabilities.

• Special education administrators must be familiar with psychometric issues regarding the validity andcomparability of accommodated test scores.

• By drawing on professional testing standards and recent court cases in this area, special educationadministrators can develop sound and legally defensible policies regarding testing of students with disabilities.

22Journal of Special Education Leadership 12(2) • November 1999

Education (NCME), 1994; Geisinger, 1994; Phillips,1994; Sireci & Geisinger, 1998). In some cases, anaccommodated test may be measuring differentconstructs than the original test. In this article, wediscuss some of the major legal and psychometricissues surrounding the testing of students with dis-abilities. The first section discusses legal questionsraised by statutory and constitutional law, the sec-ond section addresses psychometric issues, and thethird section provides important guidelines foreducators to follow to ensure that their testing procedures are both legally and psychometricallyacceptable.

Legal Issues Regarding Testing Students With DisabilitiesSeveral federal civil rights statutes clearly requirestudents with disabilities to be included in statewideand districtwide testing programs. For example,states participating in the Goals 2000: EducateAmerica Act (Goals 2000, 1999) must develop stateimprovement plans that include performance assess-ments for all students, including students with dis-abilities. Such performance assessments mustprovide for the participation of students withdiverse needs, provide necessary adaptations andaccommodations, and include multiple measures ofstudent performance. Title I of the ImprovingAmerica’s Schools Act of 1994 (Title I, 1999) alsorequires states to provide for the participation of allstudents in assessment programs, including studentswith disabilities. Each school must provide reason-able accommodations for students with diverselearning needs. Additionally, states must disaggre-gate the results each year to compare students withdisabilities to nondisabled students.

As states attempt to meet these federal man-dates, they should be aware of the legal issuesraised by both federal statutes and the UnitedStates Constitution. Important provisions includethe Individuals with Disabilities Education Act(IDEA, 1999), Section 504 of the Rehabilitation Actof 1973 (Section 504, 1999), the Americans withDisabilities Act (ADA, 1999), the Family EducationRights and Privacy Act (FERPA, 1999), and the DueProcess Clause of the Fourteenth Amendment(United States Constitution Amendment 14, Section1). A discussion of the legal issues involving theseprovisions follows.

IDEA

The recent reauthorization of the IDEA compels thestate or local education agency to develop guidelinesto enable children with disabilities to participate instate or district assessments. States must also pro-vide alternate assessments for those students whocannot participate in state or district testing pro-grams. The student’s Individual Education Program(IEP) must include a description of the modificationsmade to enable the student to participate in theassessment program. If the IEP team concludes thatthe student should not participate in the assessmentprogram, it should state why the assessment is notappropriate, and how the child will be assessed.

Does the Denial of Diplomas to Students WithDisabilities Violate IDEA’s “Free AppropriatePublic Education Requirement”?

Courts have held that the denial of diplomas to stu-dents with disabilities who received servicesrequired by the IDEA, but were unable to pass astandardized test, does not violate the statute’s freeappropriate public education requirement. InBrookhart v. Illinois State Board of Education (1983),for example, the Seventh Circuit held that thePeoria (Illinois) school district’s requirement that allstudents pass a minimum competency test toreceive a diploma was permissible under IDEA.The court reached this conclusion because thestatute mandated only that schools provide handi-capped students with access to specialized andindividualized educational services. The statute didnot require a particular result, such as the award ofa diploma. Other details regarding the Brookhartcase are described in the section on constitutionaldue process.

Must School Districts Develop Dispute Resolution Procedures for Tests for Graduation and Promotion?

It is unclear whether standardized tests for gradua-tion or promotion fall into one of the categories forwhich dispute resolution procedures under IDEAmust be developed. IDEA provides parents theopportunity to present complaints with respect toany matter related to the identification, evaluation,or educational placement of a child; or the provi-sion of a free appropriate public education. A court

Legal and Psychometric Policy Considerations

23Journal of Special Education Leadership 12(2) • November 1999

would probably find that standardized tests are acomponent of a free appropriate public education.Several IDEA provisions and implementation regu-lations clearly demonstrate that students with disabilities are expected to participate in statewideand districtwide testing programs. Therefore,school districts must provide parents with media-tion, or an impartial due process hearing to resolvedisputes involving standardized tests for gradua-tion and promotion.

Section 504 and ADA

Section 504 forbids programs receiving federal fund-ing from discriminating against “otherwise quali-fied” disabled persons. The ADA prohibits publicentities from discriminating against “qualified” indi-viduals with disabilities. The ADA also requiresexaminations related to secondary education to beaccessible to disabled persons, or that alternativeaccessible arrangements be made for such individu-als. Section 504 and the ADA are similar in substanceand, with the exception of Section 504’s federalfunding requirement, interchangeable (Allison v.Department of Corrections, 1996).

What Are Appropriate Testing Modifications for Students With Disabilities?

Future litigation surrounding Section 504 and ADAwill involve the appropriateness of testing modifica-tions for “otherwise qualified” students with disabil-ities. Wynne v. Tufts University School of Medicine(1991/1992) applied this principle to the academicsetting. In this case, the Tufts University School ofMedicine dismissed a learning-disabled student afterhe had repeatedly failed biochemistry. The studentclaimed the school had refused to test his proficiencyin biochemistry by some means other than multiplechoice tests. The district court granted summaryjudgment for the school because the student was notable to meet the school’s requirements; therefore, hewas not otherwise qualified under Section 504.

On appeal, the First Circuit reversed the lowercourt’s finding of summary judgment and remandedthe case for further proceedings. In granting sum-mary judgment, the lower court had failed to con-sider (1) the extent to which reasonableaccommodations that would satisfy the legitimateinterests of both the school and the student were

available; and (2) if such accommodations did notexist, the extent to which the institution exploredpotential alternatives. The First Circuit vacated sum-mary judgment and sent the case back to the lowercourt to determine whether the university had satis-fied its duty. On remand, the district court againgranted summary judgment to the medical school.The First Circuit upheld the summary judgmentbecause the school demonstrated that it consideredother options, but rationally concluded that suchoptions would either lower academic standards orrequire substantial program alteration. The Wynnecase suggests a school district’s modifications willsurvive legal challenge if the district shows that ithas considered the requests of disabled students, butdecided after careful deliberation that the accommo-dation request would require substantial alterationor lower the district’s academic standards.

Does the Flagging of Test Scores Violate Section 504 and the ADA?

Courts may also have to determine whether“flagged” test scores violate Section 504 and theADA (Pullin & Heaney, 1997). A flagged test scorerefers to a signal, usually in the form of an asterisk,that alerts a decisionmaker (e.g., college admissionsofficer or employer) that the test was given in a non-standard administration. Education Testing Service(ETS) and other testing agencies have flagged theirtest scores because: (1) they cannot guarantee thatthe accommodated tests can be compared to teststaken under normal conditions; and (2) it is difficultto conduct true validity studies because of the smallnumber of test-takers in each disability group(Mayer, 1998). The practice of test flagging is prob-lematic because the reviewer of the test score mayinfer that the flag signifies that the test-takerreceived an accommodation due to a disability(Mayer, 1998; Pullin & Heaney, 1997). Consequently,a college may use this information to deny admis-sion to students with disabilities, even though thestudent may be “otherwise qualified” to attend thatinstitution. Such an action would violate bothSection 504 and the ADA.

There has been disagreement as to how schooldistricts, testing agencies, and universities mayavoid liability under these statutes. Mayer (1998)asserts that the practice of flagging test scores vio-lates Section 504 and the ADA and should be aban-

Legal and Psychometric Policy Considerations

24Journal of Special Education Leadership 12(2) • November 1999

doned. However, the courts’ general grant of defer-ence to the policy making decisions of educationalinstitutions suggests that they will withstand legalchallenges by providing safeguards to disabledapplicants. Phillips (1994) recommends that testingagencies should inform parents and students ifscores from an accommodated test will be flagged,and obtain the consent to the release of such infor-mation prior to taking the test. Pullin and Heaney(1997) further advise that the admissions process ofuniversities “should include decisionmakers well-versed in the assessment and educational capabili-ties” of students with disabilities (pp. 827-28).

FERPADoes the Practice of Test Flagging Violate the FERPA?

A school district that reports flagged test scores tocolleges or employers might also be in violation ofthe Family Education Rights and Privacy Act(FERPA). FERPA prohibits educational institutionsfrom receiving federal funds if they have a policyof releasing school records or personally identifi-able information to unauthorized persons, agen-cies, or organizations. FERPA was adopted toprevent systematic violations of students’ privacyby unauthorized releases of sensitive informationin their educational records (Jensen v. Reeves, 1999).Courts have held that no private cause of actionexists under FERPA; however, parents may bring acause of action under Title 42 United States CodeService § 1983 (1999) for a violation of FERPA, ifschool officials acted under the color of state law.Thus, the practice of providing test scores withoutwritten parental consent may cause a school dis-trict to lose federal funding, as well as expose themto legal challenges from parents. To avoid liability,school districts should inform students and parentsif scores from accommodated tests will be flagged,and if so, require parents and students to consentto the release of this information prior to the takingof the test.

Constitutional Due Process

If a state decides to require students with disabilitiesto pass a standardized test, it may have to providesuch students with “constitutional due process” protection. The Due Process Clause of theFourteenth Amendment forbids a governmental

entity from depriving a person of a property or liberty interest without providing procedural dueprocess or substantive due process. A liberty interestmay arise in situations where injury to a person’sreputation, honor, or integrity is combined with“governmental action [that] deprive the individualof a right previously held under state law” (Hall v.Davis, 1976, pp. 708-09). In general, procedural dueprocess requires public schools to provide studentswith adequate notice, and a hearing providing students with an opportunity to be heard (Dixon v.Alabama State Board of Education, 1961). Substantivedue process prevents governmental entities fromdepriving students of that right in an arbitrary orcapricious manner (Debra P. v. Turlington, 1981).

Are Students with Disabilities Entitled to Constitutional Due Process?

In Board of Education of Northport-East NorthportUnion v. Ambach (1983), the Court of Appeals of NewYork found that students with disabilities who arefunctionally unable to pass a standardized test donot have a liberty interest to a diploma. The schooldistrict awarded high school diplomas to two dis-abled students who failed competency tests in read-ing and mathematics, but successfully completedtheir IEPs. One of the students had a neurologicaldisorder while the other was mentally retarded. Thecommissioner sought to revoke the diplomasawarded to these students. The court of appeals heldthat the students did not have a liberty interest to adiploma because they were incapable of advancingto the point where they could pass the test. Thecourt also found that students with disabilities whowere capable of passing the test were not deprivedof procedural due process because they had threeyears notice of the testing requirement.

By contrast, in Brookhart v. Illinois State Board ofEducation (1983), the Seventh Circuit found that stu-dents with disabilities have a due process right toreceive a diploma upon completion of previouslyrequired work. Thus, a school district could not sub-sequently require students with disabilities to pass ahigh stakes test to receive a diploma without firstproviding procedural and substantive due processprotections. In Brookhart, the Peoria (Illinois) SchoolDistrict decided to require all students to pass a minimum competency test to receive a diploma.

Legal and Psychometric Policy Considerations

25Journal of Special Education Leadership 12(2) • November 1999

Prior to the adoption of the testing requirement, stu-dents could receive a diploma under state law ifthey completed 17 course credits and fulfilled thestate’s graduation requirements. Fourteen elemen-tary and high school students with disabilities whodid not pass the test alleged that the school districtviolated their due process rights by failing to pro-vide them with adequate notice to prepare for theminimum competency test. The Seventh Circuitfound that the students had a liberty interest for tworeasons. First, failure of the test attached a “stigma”on the students that could have had devastatingeffects on their educational and employment oppor-tunities. Second, by altering the graduation require-ments, the school district deprived students of aright previously held under state law. It appears thatthe Brookhart court reached a different decision fromthat of Ambach because the students could have con-ceivably passed the minimum competency test ifthey had received the opportunity.

The Brookhart court then found that the schooldistrict violated the due process rights of the highschool students. The Seventh Circuit distinguishedBrookhart from Ambach in that the students receivedonly 1½ years notice of the graduation changes,which was insufficient time to provide students withdisabilities with the necessary preparation to passthe test. To support this conclusion, the court notedstudents lacked exposure to as much as ninety per-cent of the material on the minimum competencytest. Moreover, the students’ programs of instructionwere not designed to meet the goals of passing thetest, but rather were developed to address individ-ual educational needs. Thus, Brookhart demonstratesthat the extent of notification of a testing require-ment is a critical issue in evaluating due process.

Psychometric Issues Regarding Testing Students With Disabilities

In addition to legal mandates, leaders in special edu-cation must also consider validity issues when test-ing students with disabilities. To psychometricians(i.e., professionals who build and evaluate tests), allissues of test fairness are issues of validity. Seminalconceptualizations of validity (e.g., AERA, APA,NCME, 1994; Messick, 1989; Shepard, 1993) assertthat validity is not a property of a test, but rather a

concept that refers to the fairness and utility of deci-sions made on the basis of test scores. For example,when test scores are used to place students intoinstructional programs, it is the placement decisionsthat must be validated, not the test itself. Well-developedtests may be used inappropriately. Therefore, issuesof test validity must always be considered withrespect to the purposes and consequences of the test-ing. When students with disabilities are tested, anobvious validity issue is whether their test scores areappropriate for making placement, promotion, grad-uation, admissions, and other specific decisions.

The Testing of Students with Disabilities Raises Several Measurement Problems.

There are two key psychometric terms to be under-stood when evaluating any educational test: con-struct validity and content validity. Construct validityrefers to the degree to which a test measures the con-struct it is designed to measure (Messick, 1989).Constructs are unobservable characteristics of peoplethat are hypothesized to account for individual dif-ferences and behaviors. For example, “mathematicsproficiency” is a construct. We cannot observe a stu-dent’s mathematics proficiency, so we develop a testto measure this construct. Evaluations of constructvalidity investigate how well test scores reflect theconstruct the test is designed to measure. Contentvalidity refers to the degree to which the content of atest is congruent with the purposes of the testing(Sireci, 1998). For example, if a test is designed tomeasure how well eighth grade students in a partic-ular school district have learned earth science, thecontent of the test should reflect the district’s eighthgrade earth science curriculum. Construct and con-tent validity are central to the evaluation of theappropriateness and utility of all students’ testscores. However, with respect to testing studentswith disabilities, several additional questions are raised: (1) Do standardized (i.e., non-accommodated) tests provide valid assessment ofthe knowledge, skills, and abilities possessed by stu-dents with disabilities?; (2) Do scores resulting fromaccommodated tests possess the same psychometricproperties as scores derived from standardized tests?;and (3) If a test or a test administration is accommo-dated, are students’ scores from the accommodatedtest comparable to students’ scores from the original

Legal and Psychometric Policy Considerations

26Journal of Special Education Leadership 12(2) • November 1999

test? The answer to these questions is typically, “itdepends.” This is the most precise answer becausestudents with disabilities are a heterogeneous group,and test accommodations vary widely from one situ-ation to the next. Thus, when considering thesevalidity issues, the nature of a student’s disabilitymust be considered alongside the nature of the testaccommodation.

When Should a Test Accommodation Be Made for Students with Disabilities?

A fundamental issue underlying the first question iswhether a test accommodation is warranted for aparticular student. If the constraints of a standard-ized test administration make it difficult or impossi-ble for a student to demonstrate her/his trueknowledge, skills, and abilities, a case can be madefor an accommodated test. However, if the accom-modation changes the nature of what the test is mea-suring, a case can be made against providing anaccommodated test. For example, if a reading com-prehension test is accommodated for a dyslexic stu-dent by allowing someone to read the test aloud tothe student, the construct measured may changefrom reading comprehension (the original test) to lis-tening comprehension (the accommodated test). Thisis an obvious problem with respect to test scoreinterpretation. On the other hand, an accommoda-tion such as allowing extended time on a readingcomprehension test may retain measurement of theoriginal construct. Therefore, when determiningwhether an accommodation is warranted, test usersmust consider the need for the accommodation aswell as the effect the accommodation will have onthe nature of the construct measured. In addressingthis issue, Phillips (1994) urged testing agencies toclearly document the objectives of the test so that theeffect of the accommodation on measurement of theintended objectives could be ascertained.

What Types of Accommodations Are Available for Students with Disabilities?

Requests for accommodations to testing and thetypes of accommodations provided by test develop-ers are numerous. Many of these accommodationsare described in detail elsewhere (e.g., Geisinger;1994; Phillips, 1994; Scarpati, 1991; Scarpati &

Sireci, 1998; Sireci & Geisinger, 1998; Thurlow,Elliot, & Ysseldyke, 1998; Thurlow, Hurley,Spicuzza, & Sawaf, 1996). Some accommodationsare designed for specific sensory or motor disabili-ties. Examples of such accommodations are audio-cassette, Braille, and large-print versions of a testfor students with visual impairments; use of ascribe or recorder to assist in recording of answersfor students with motor or verbal disabilities; andsign language recorders or videocassette recordersfor students with hearing impairments. A variety ofaccommodations are also provided for applicantswith learning disabilities (e.g., dyslexia, dys-graphia, attention deficit disorder, dyscalculia, etc.).These accommodations include providing frequentbreaks, oral presentation of instructions or testquestions, use of a calculator or computer, extendedtesting time, and taking the test in a private roomor in a small group setting. For both physical andlearning disabilities, some accommodations includeomitting specific test content.

Are Accommodated Tests Comparable to Original Tests?

Regardless of the type of accommodation, a keyissue when considering whether to grant an accom-modation is whether the accommodated test willmeasure a construct different from the one the orig-inal test was designed to measure. Some psychome-tricians attempted to answer the question ofconstruct comparability (or phrased another way,how comparable are scores from original andaccommodated tests?) statistically. In most cases,the responses of students with and without disabili-ties to test questions have been analyzed. Factoranalysis, analysis of differential item functioning,and analysis of test score reliability have been con-ducted on both original and accommodated tests.For example, Tippets and Michaels (1997) foundthat the factor structures of several accommodatedtest forms of the Maryland State PerformanceAssessment were very similar to the factor struc-tures of the original test forms. Also, several studieswere conducted on original and accommodatedversions of the Scholastic Aptitude Test and theGraduate Record Examination (Rock, Bennett, &Jirele; 1988; Rock, Bennett, & Kaplan, 1987;Willingham, W.W., Ragosta, M., Bennett, R.E.,

Legal and Psychometric Policy Considerations

27Journal of Special Education Leadership 12(2) • November 1999

Braun, H., Rock, D.A., & Powers, D.E., 1988). Ingeneral, these studies found comparable levels ofscore reliability and item quality, and comparablefactor structures across original and accommodatedversions of these tests. However, commentaries onthis research (e.g., Geisinger, 1994) and the authorsof these studies point out that research in this areais extremely limited. There are often too few stu-dents with disabilities to conduct these types ofstudies, and so students with very different typesof disabilities are often aggregated into a singlegroup for analysis. Nevertheless, these studies illus-trate promising approaches for evaluating accom-modated tests. They also address Standard 14.6 ofthe Standards for Educational and PsychologicalTesting, which states, in part, “When feasible, thevalidity and reliability of [accommodated tests]should be investigated and reported by the agencyor publisher that makes the modification…”(AERA, APA, NCME, 1994; p. 80). Clearly, thevalidity of scores from accommodated versions oftests, and the comparability of scores from stan-dardized and accommodated tests, are areas inneed of much more research.

It is interesting to note that the problem of “con-struct equivalence” across original and accommo-dated tests also applies to tests that are translated oradapted from one language to another. Althoughnon-English language versions of a test are commonlyused throughout the U.S. to assess the academicskills of students operating in a language other thanEnglish, limited English proficiency does not qualifyas a “disability.” Nevertheless, the methodologicaltechniques used to evaluate different language ver-sions of a test (e.g., Sireci, Xing, & Fitzgerald, 1999)are fully applicable to the problem of comparingaccommodated versions of a test to the original version.

Summary

Consideration of psychometric standards of testquality require that the objectives measured by thetest be considered when making decisions aboutaccommodating standardized educational tests.Phillips (1994) used a psychometric framework topose five critical questions to be considered beforegranting an accommodation. She asserted that if anaccommodation alters the construct measured,

provides scores that are not comparable to scoresfrom the original version, or would also benefitnondisabled students, then the accommodationshould not be provided (p. 104). However, she alsoadded, “To avoid litigation when in doubt, the testadministrator may want to err on the side of grant-ing the accommodation whenever feasible” (p. 104).We consider this to be good advice, particularlywhen there is evidence that an accommodated test ismeasuring the same construct as the original test.

Conclusions: Some Suggestions ForTesting Students With DisabilitiesThe legal issues raised by federal statutory and con-stitutional provisions, as well as the measurementconcerns raised by psychometricians, present a com-plex situation for educators of students with disabili-ties. Based on our review of federal legislation, courtcases, and the psychometric literature, we providesome suggestions to help educators develop legallyand psychometrically sound testing programs.• School districts should provide adequate notice

of testing requirements to students with disabili-ties. The Ambach case suggests that three yearsconstitutes adequate notice. Such notice isneeded to ensure that IEPs are aligned with theeducational objectives tested.

• Test accommodation requests made by disabledstudents should be reviewed on an individualbasis. As the First Circuit explained in Wynne,accommodations deemed to be reasonable in oneset of circumstances may not be reasonable inanother. A court will most likely uphold the test-ing decisions of a school district, if that districtcan show that it has taken the special circum-stances of a student into account.

• Students should be informed if scores fromaccommodated tests will be flagged. Students(and parents, if necessary) should be required tosign consent forms indicating they are aware thatthe scores will be flagged, and agree to havetheir scores released.

• High stakes educational tests must demonstratecontent validity. Tests should be constructed sothat all content is appropriate for all test-takers,including those with disabilities. Any test con-tent or item format that is not essential from acontent validity perspective, and that may inter-

Legal and Psychometric Policy Considerations

28Journal of Special Education Leadership 12(2) • November 1999

fere with the performance of students with dis-abilities, should be eliminated (Sireci & Mullane,1994).

• Studies should be conducted on accommodatedtests (and on alternative assessments) to evaluatehow well the accommodated test represents theeducational objectives measured (content valid-ity studies), and to determine the comparabilitybetween scores on the original and accommo-dated versions of the test (construct validitystudies). The results from these studies should bedocumented and available to interested parties.

• Students with disabilities should be included inthe pretesting of test items and item formats. Theperformance of students with disabilities onthese preliminary test forms should be consid-ered when selecting items for the final, opera-tional tests.These suggestions will not solve all problems

encountered by educators involved in the testing ofstudents with disabilities. However, by consideringthese issues before instituting policies regarding theassessment of students with disabilities, leaders inspecial education can avoid many of the problemsthat cause students with disabilities to challenge atest in court.

ReferencesAERA: American Educational Research Association,

APA: American Psychological Association, &NCME: National Council on Measurement inEducation (1994). Standards for educational and psy-chological testing. Washington, D.C.: AmericanPsychological Association.

Allison v. Dept. of Corrections, 94 F.3d 494 (8th Cir. 1996).ADA: Americans with Disabilities Act of 1990, 42 U.S.C.S.

§ 12101 et seq. (1999). Board of Education of Northport-East Northport Union Free

School District v. Ambach, 60 N.Y.2d 758, 457 N.E.2d775, 469 N.Y.S.2d 669 (1983).

Brookhart v. Illinois State Board of Education, 697 F.2d 179(7th Cir. 1983).

Debra P. v. Turlington, 644 F.2d 397 (5th Cir. 1981).Dixon v. Alabama State Board of Education, 294 F.2d 150 (1961). FERPA: Family Education Rights and Privacy Act, 20

U.S.C.S § 1232g (1999), 42 U.S.C.S. § 1983 (1999).Geisinger, K.F. (1994). Psychometric issues in testing

students with disabilities. Applied Measurement inEducation, 7, 121-140.

Goals 2000: Educate America Act, 20 U.S.C.S. § 5801 et seq. (1999).

Hall v. Davis, 425 U.S. 693 (1976).Heubert, J. & Hauser, R. (Eds.) (1999) High stakes: Testing

for tracking, promotion, and graduation. Washington,D.C.: National Academy Press.

IDEA: Individuals with Disabilities Education Act, 20U.S.C.S. § 1401 et seq. (1999).

Jensen v. Reeves, 1999 U.S. Dist. LEXIS 6229 (D.Utah 1999). Mayer, K.S. (1998). Flagging nonstandard test scores in

admissions to institutions of higher education.Stanford Law Review, 50, 469-520.

Messick, S. (1989). Validity. In R. Linn (Ed.), Educationalmeasurement, (3rd ed., pp. 13-103). Washington, D.C.:American Council on Education.

Murphy, K. (1999, February 23). Lawsuit spotlights battleover basic skills testing; exams unfair for studentswho rely on aids to overcome problems of disabilities,critics say. The Los Angeles Times, p. A1.

Phillips, S.E. (1994). High-stakes testing accommodations:validity versus disabled rights. Applied Measurement inEducation, 7, 93-120.

Pullin, D. & Heaney, K. (1997). The use of “flagged” testscores in college and university admissions: Issuesand implications under Section 504 of theRehabilitation Act and the Americans with DisabilitiesAct. Journal of College and University Law, 23, 797-828.

Rock, D.A., Bennett, R.E., & Jirele, T. (1988) Factor struc-ture of the Graduate Record Examination’s generaltest in handicapped and nonhandicapped groups.Journal of Applied Psychology, 73, 382-392.

Rock, D.A., Bennett, R.E., & Kaplan, B.A. (1987). Internalconstruct validity of a college admissions test acrosshandicapped and nonhandicapped groups.Educational and Psychological Measurement, 47, 193-205.

Scarpati, S. (1991). Current perspectives in the assessmentof the handicapped. In R.K. Hambleton & J.N. Zall(Eds.), Advances in educational and psychological testing(pp. 251-276). Norwell, MA: Kluwer Academic.

Scarpati, S. & Sireci, S.G. (1998, August). Including students with disabilities in state and district tests:Perceptions of accommodations and score validity. Paperpresented at the annual meeting of the AmericanPsychological Association, San Francisco, CA.

Section 504 of the Rehabilitation Act of 1973, 29 U.S.C.S. §794(a) (1999).

Shepard, L.A. (1993). Evaluating test validity. Review ofResearch in Education, 19, 405-450.

Sireci, S.G. (1998). The construct of content validity. SocialIndicators Research, 45, 83-117.

Legal and Psychometric Policy Considerations

29Journal of Special Education Leadership 12(2) • November 1999

Sireci, S.G. & Geisinger, K.F. (1998). Equity issues in employment testing. In J.H. Sandoval, C. Frisby,K.F. Geisinger, J. Scheuneman, & J. Ramos-Grenier(Eds.), Test Interpretation and diversity (pp. 105-140).Washington, D.C.: American PsychologicalAssociation.

Sireci, S.G. & Mullane, L.A. (1994). Evaluating test fairnessin licensure testing: the sensitivity review process.CLEAR Exam Review, 5 (2) 22-28.

Sireci, S.G., Xing, D., & Fitzgerald, C. (1999, April).Evaluating adapted tests across multiple groups:Lessons learned from the information technologyindustry. Laboratory of Psychometic and EvaluativeResearch Report No.349, Amherst, MA: School ofEducation, University of Massachusetts.

Thurlow, M.L., Elliott, J.E., & Ysseldyke, J.E. (1998). Testingstudents with disabilities: Practical strategies for comply-ing with district and state requirements. Thousand Oaks,CA: Corwin Press.

Thurlow, M.L., Hurley, C., Spicuzza, R., & El Sawaf, H.(1996). A review of the literature on testing accommoda-tions for students with disabilities. Minnesota Report No. 9. Minneapolis, MN: National Center on Educational Outcomes.

Tippets, E. & Michaels, H. (1997, March). Factor structureinvariance of accommodated and non-accommodated perfor-mance assessments. Paper presented at the annualmeeting of the National Council on Measurement inEducation, Chicago, IL.

Title I of the Improving America’s Schools Act of 1994, 20U.S.C.S. § 6301 et seq. (1999). U.S. CONSTITUTION

Amendment XIV, § 1.Title 42 U.S.C.S. sec. 1983 (1999).Willingham, W.W., Ragosta, M., Bennett, R.E., Braun, H.,

Rock, D.A., & Powers, D.E. (Eds.). (1988). Testinghandicapped people. Needham Heights, MA: Allyn & Bacon.

Wynne v. Tufts University School of Medicine, 976 F.2d 791(1st Cir. 1992).

Wynne v. Tufts University School of Medicine, 932 F.2d 19(1st Cir. 1991).

About the AuthorsPreston C. Green, Ed.D., Esq., and Stephen G. Sireci,Ph.D., are assistant professors in the School ofEducation, 159 Hills-South, University ofMassachusetts at Amherst 01002. E-mail: Preston C.Green, Ed.D., Esq. at [email protected] andStephen G. Sireci, Ph.D. at [email protected]

Legal and Psychometric Policy Considerations

Begin Early, End Well Strategies to Improve Results for Students With Disabilities

Thomas F. Hehir, Ed.D.Harvard University & Education Development Center, Inc. of Newton, MA

• All infants and toddlers with disabilities must receive appropriate interventions.

• Preschool programs must prepare students to be successful in the primary grades.

• It is important to intervene early with children experiencing significant reading and behavior problemsin school.

• All students must be provided with meaningful access to the general education curriculum.

• All students with disabilities must complete high school.

31Journal of Special Education Leadership 12(2) • November 1999

Recently, I was on a flight in which I was seatednext to a young college man who was traveling

to Washington, D. C. to rendezvous with a youngwoman he had met during spring break in Florida.Asking me if I was from Washington, he was inter-ested in my suggestions for a good restaurant wherehe could impress his new girlfriend for under $20. Idrew a blank. During our conversation he informedme that he was finishing college in June and that hewas quite proud of that fact because he had a learn-ing disability. “I am not ashamed of that. I tell every-one. It’s part of who I am.” Well, I of courseinformed him of what I do for a living and the con-versation proceeded from there. I asked him whenhis parents first knew he had a learning disability.He informed me that he had been identified at agefour because of significant problems with speechand language, and he had received special educationsupport throughout school and was provided vari-ous accommodations during college. The determina-tion of this young man to defy the low expectationsthat so often plague students with disabilities wasevident in a letter he subsequently sent me seekingmy advice as he contemplates employment andgraduate school. “I say this as one who was told totake five or six years to complete high school andthen maybe I could go on to a two-year program. Isaid, ‘no way.’ I’m graduating with my class andgoing on to college!”

This young man’s successful educational experi-ence is shared by more and more students served

under IDEA, as the number of young people withdisabilities going on to higher education has tripledsince 1990 (American Council on Education, 1992).Even though we have made great progress, all stu-dents with disabilities do not share the experience ofthis young man. Too many drop out of school, closeto double the rate of their nondisabled counterparts,or leave school without the skills and knowledgethey need to be successful (Seventeenth AnnualReport to Congress, 1995). Increasingly, we knowfrom research and the experience of implementingIDEA what we need to do to improve educationalresults for students with disabilities. I would like toshare five strategies, which the research would indi-cate are important to improve educational results forstudents with disabilities.

Ensure that All Infants and Toddlers with Disabilities ReceiveAppropriate InterventionsIncreasingly educators know the importance of earlyintervention and how the early years can impact onthe success of children in school. This is particularlytrue for students with disabilities. The first years oflife are critical to intellectual, language, sensory, andmotor development. IDEA, Part C, funds early inter-vention programs in all the states. Though not typi-cally run by local school districts, public schoolshave a great stake in the effectiveness of these pro-grams. For instance, deaf children need to start

32Journal of Special Education Leadership 12(2) • November 1999

school with vocabularies commensurate with theirhearing peers if they are going to achieve decent lev-els of literacy. However, we know from research thatmost do not, and the reading levels of most deaf stu-dents are very low. We also know that the optimaltime for language development is in the first yearsof life. Deaf children typically don’t develop lan-guage well in the early years because they have lim-ited access communication from their parents.

Studies show that deaf children born of deaf par-ents, a minority of deaf children, have significantlyhigher levels of vocabulary because they typicallyhave a system of communication with their parents,sign language, whereby they can have access to lan-guage development (Lane, 1990). Further, it has beendocumented that deaf children who have high levelsof ASL ability when they start school achieve higherlevels of English literacy regardless of IQ (Strong &Prinz, 1997). This is why intervention is so importantfor these children and their families as early as possi-ble, a communication system between the child andthe parent must be established, and intensive workshould be done to enhance vocabulary development.

Effective intervention must be based on compre-hensive assessment and screening programs. It shouldbe noted that there is a bill in Congress to providesupport to states to provide universal infant hearingscreening. This bill and similar bills at the state levelsthat seek to identify infants and toddlers with disabili-ties as early as possible should be strongly supportedby the special education community. More informa-tion about this legislation can be obtained through theAmerican Speech, Language, and Hearing Association(ASHA), Rockville, Maryland.

The same principle of the importance of earlyintervention applies to other children with disabilities.Early speech and language therapy can correct prob-lems that will be more costly to correct later. And, theoptimal time for certain physical and occupationaltherapies is during the early years when developmentis most rapid. School district administrators shouldtake a deep interest in the early intervention programsin their community. Well functioning early assessmentand intervention programs will result in studentsready to learn when they reach school and candecrease the level of need for special education andrelated services later on. Each state and many localcommunities have Interagency Coordinating Councils,

which oversee these programs and are an excellentvehicle by which school districts can become involvedin this important program.

Make Sure Preschool Programs Prepare Students to Be Successful During the Primary GradesOver the past ten years the number of preschoolersserved under IDEA has doubled. Approximately fortypercent of students served by special education areidentified before they enter school. Given theincreased recognition of the importance of early inter-vention for students at risk of failure in school, this isa positive trend. However, simply providing thesestudents with programs is not enough. Programsmust be of high quality and prepare these youngstersto be successful in school. Recently, the Department ofEducation sponsored an important research synthesisconducted by the National Academy of Sciences,Preventing Reading Difficulties in Young Children(Snow, 1998). This study showed clearly that many ofthe difficulties that students encounter in learning toread can be prevented or significantly amelioratedthrough high quality early intervention programs.Basic phonemic awareness (understanding that sim-ple words are composed of sound segments) or theability to rhyme are skills frequently lacking in chil-dren at risk of having difficulty learning to readincluding children with disabilities. These skills canbe taught before children enter school. Further, manychildren need vocabulary and language enhancement.Other children need to develop social and behavioralskills, which will enable them to start school success-fully. The interventions we deliver in our preschoolprograms must be based on accurate assessment ofchildren’s needs. Quality preschool programs requirewell-trained staff to prepare students to learn in theinclusive environments in which more and more dis-abled students are educated today.

Intervene Early with Children ExperiencingSignificant Behavior and ReadingDifficulties in the Primary GradesOver fifty percent of students served by special edu-cation are not identified until they are in school(Nineteenth Annual Report to Congress Report onthe Implementation of IDEA, 1996). The primary

Strategies to Improve Results for Students with Disabilities

reasons these students are referred are due to prob-lems learning to read or problems with behavior.Typically these students do not get referred to specialeducation until third or fourth grade and sometimeseven later. Some children referred at this time do nothave disabilities, but have been poorly taught. Thosewho have disabilities frequently have had very nega-tive experiences in the primary grades and importanttime has been lost. General education and specialeducation need to work closely together to assurethat all children experiencing significant problemslearning to read or behaving in school receive effec-tive interventions in kindergarten and first and sec-ond grade to turn these problems around.

General education and special education need to

work closely together to assure that all children expe-

riencing significant problems ... receive effective inter-

ventions in kindergarten and first and second grade ...

One researcher interviewed the kindergartenteachers of students who were later identified asemotionally disturbed. (Walker, Severson, & Fell,1995) In virtually every case the kindergarten teacherhad been deeply concerned with the child’s behav-ior. These kids weren’t typical kindergartners in theeyes of their former teachers, but were quite atypi-cal. Unfortunately, most students with emotionaldisturbance do not receive special education servicesuntil they are in the fifth grade or later. And, whenthey do receive special education, the programs theyreceive often are ineffective. A study conducted byOSEP documented that less than twenty percent ofthis population had behavior goals on their IEPs,and close to fifty percent of these students drop outbefore graduation (Wagner, Blackorby, Cameto,Hebbler, & Newman, 1993). Children exhibitingatypical behavior in kindergarten or the first gradeshould be given help then, not later. Kids with sig-nificant behavior problems don’t typically grow outof their problems. The undesirable behaviors aremore apt to increase if left unaddressed. The soonerwe start providing these children with positivebehavioral interventions, the greater the likelihoodwe will be able to change these children’s behavior.

The same urgency that applies to students experi-encing significant behavior problems also applies to

those experiencing difficulty learning to read. Large-scale studies done by the National Institutes ofHealth estimate that upward of twenty percent ofyoung children experience significant difficulty learn-ing to read (Lyon, 1999). There are a variety of rea-sons for this. Some have been poorly taught. Othersstart school without a firm language foundation.Some have learning disabilities. Regardless of thecause, difficulty learning to read in the primarygrades is highly predictive of poor overall schoolresults (Lyon, 1999). These children need more inten-sive support sooner rather than later. For studentswith learning disabilities more explicit instruction inphonemic awareness may be needed. Many of thesestudents may always need more help in reading,writing, and spelling activities. Unfortunately, stu-dents with learning disabilities do not typicallyreceive services from special education until thirdgrade or later, often after the student has experiencedsignificant school failure. Indeed, the current way inwhich most states define learning disabilities followsa federal definition that requires the student demon-strate a significant discrepancy between intelligenceand performance in school. This assumes failure. Thisdefinition is currently under review by the U. S.Department of Education, and IDEA ’97 allows for amore flexible definition of “developmental delay.”Definitions aside, the sooner we provide these stu-dents with the help they need the more successfulthey will be, and the likelihood of expensive out-of-district placements will decrease. President Clintonin his FY 2000 budget has proposed a fifty milliondollar program to provide LEAs with grants to pro-mote earlier intervention for students with signifi-cant reading and behavior problems. This program,if funded by Congress, will provide innovativeschool districts with the resources to intervene whenit is obvious children need help and at the timewhen interventions are more likely to be effective.

The need to provide earlier intervention for stu-dents experiencing significant behavior and or read-ing problems is one of the most difficult butimportant challenges special education administra-tors’ face today. Current practice and, at times, inap-propriate intervention is failing too many children.Yet in order for us to reach children sooner, we can-not depend on special education alone. Childrenwith disabilities are not the only children experienc-ing these problems in the primary grades. General

33Journal of Special Education Leadership 12(2) • November 1999

Strategies to Improve Results for Students with Disabilities

34Journal of Special Education Leadership 12(2) • November 1999

education and special education must work togetherto address these issues with comprehensive school-wide approaches that will assure that all children getthe interventions they need when they need them.

Provide Students with Meaningful Access to the General CurriculumIn all of our efforts to improve educational resultsfor students with disabilities, we must begin withthe assumption that students graduate from ourschools with the skills and knowledge they will needto lead full, productive lives. Therefore, studentswith disabilities need to be learning the same con-tent in school as their nondisabled peers.Unfortunately, many students with disabilities havebeen placed in special education classrooms, whichhave not been teaching the same content as has beentaught in the regular classrooms. Many studentswith disabilities at the secondary level have beendiscouraged from taking challenging academicclasses (Wagner, et al., 1993). Further, until recentlymost students with disabilities were not included indistrict or state level testing and accountability sys-tems. The result has been that students with disabili-ties often leave school without the skills andknowledge they need to compete and, thus, toomany are under-employed or unemployed with lim-ited access to higher education.

IDEA ’97 sought to change this situation withspecific requirements that students with disabilitieshave access to the general education curriculum andthat they be included in local and state assessmentprograms. States must also establish performancegoals for their special education programs that atminimum address how well students with disabili-ties are progressing in the curriculum.

What does access to the curriculum mean?Basically it means that educational planning for stu-dents with disabilities needs to have as its foundationthe curriculum being taught within the district andthat curriculum be presented at a level and in a man-ner in which students with disabilities can acquire thecontent (McLaughlin, 1999). Clearly many studentswith disabilities will need accommodations, modifica-tions, and supports in order to access subject matter.For instance, a high-school-aged student with learningdisabilities who does not read at grade level may needthe text modified to his or her reading level in a

physics course. The important point is that this stu-dent’s disability should not preclude this student fromaccess to physics or any other subject.

In order for students with disabilities to gainaccess to the curriculum, special and general educators need to work closely together to plan inadvance the accommodations and modifications thestudent will need as well as the support the teachersand student should expect. The IEP should reflectthese understandings with active participation of theparents and, as appropriate, the student. IDEA ’97includes these important concepts and also allowsfor the more flexible use of federal special funds toencourage collaborative approaches between specialand general education.

A critical aspect of assuring access to the curricu-lum is the need to assess how well students areacquiring the skills and content being taught.Unfortunately, very few special education programshave been managed based on results data. Specialeducation has often been more focused on processoversight. This was probably necessary in the earlyyears of P.L. 94-142 implementation as we wereputting systems in place to educate students withdisabilities. However, focusing on process oversightis not enough. We now need to engage appropriateongoing assessments to make sure these programsare working for students, and when they are not weneed to adjust our approaches accordingly.

All Students with Disabilities Must Complete High SchoolStudents with disabilities have almost twice the rateof dropping out of high school as their nondisabledpeers. This is particularly true of students withlearning disabilities or emotional disturbances. Also,the consequences of dropping out of school are moresignificant for students with disabilities. Few everreturn to school and more experience unemploy-ment, early unwed maternity, and problems with thelaw (Wagner, et al., 1993). On the other hand, stu-dents with disabilities who stay in school and geteffective programs are moving on to higher educa-tion and employment at increasing levels (Harris, L.& Associates, 1986, 1994, 1998).

There are characteristics of some programs, identi-fied through research, that increasingly are associatedwith the likelihood that students with disabilities will

Strategies to Improve Results for Students with Disabilities

35Journal of Special Education Leadership 12(2) • November 1999

stay in school and will have better results once leavingschool. One of these factors is access to vocational edu-cation. Students with disabilities who have access togood vocational programs are more apt to stay inschool and enjoy higher incomes once they leave(Wagner, et al., 1993). This is not to say that all stu-dents with disabilities should be in vocational pro-grams, but, for many, these programs have beenshown to be effective. Another characteristic that hasbeen shown to improve results is the availability ofcounseling services (Wagner, et al., 1993). Studentswith disabilities often encounter difficulties negotiat-ing the high school environment. If they have a personto go to for help in solving these problems, and to attimes advocate for them, they are more likely to stayin school (Larson, 1992). Also, integration into generaleducation programs is related to improved educa-tional results, but only if students receive the accom-modations, modifications, and supports that they need(Wagner, et al., 1993).

Research has also identified another factor isassociated with increasing the chances of studentswith disabilities dropping out: transferring studentsfrom one program to another. Students with disabili-ties are more apt to be transferred from program toprogram when they are encountering difficulties.Though school personnel often make these transferswith the best intentions, attempting to find a pro-gram that works, these actions increase the probabil-ity students will drop out. (Larson, 1992) A betterapproach in many of these situations would be tomodify the program the student is currently in tobetter meet his or her needs.

The need to improve high school options for stu-dents with disabilities is reflected in a new require-ment in IDEA ’97: the requirement to begintransition planning no later than the age of 14. Animportant aspect of this process is the requirement toconsider the course of study the student will be pur-suing in high school. Given the risks these studentsface in high school, good planning at this stage iscrucial. Part of this planning process involves requir-ing students to increasingly take responsibility forthemselves and as appropriate make critical deci-sions about their academic programs. Empoweringstudents in this way has been shown to be effective.Some schools, such as J.E.B. Stuart High School inFalls Church, Virginia, have even experimented withstudent-led IEP meetings for older students.

SummaryAll of the strategies and approaches presented in thisarticle require us to move away from traditionalnotions of special education as separate and distantfrom the general education program and move towarda recognition that special education and general edu-cation must work together to improve results. Many ofthe effective practices discussed in this paper can beeffective for other students as well. When special edu-cation and general education work together the educa-tion of all children improves. The wisdom of a greatAmerican, Helen Keller, is appropriate here, “Alonewe can do so little, together we do so much!”

ReferencesAmerican Council on Education. (1992). College freshmen

with disabilities: A statistical profile. Washington, D.C.:Heath Resource Center.

Harris, L. & Associates. (1986, 1994, 1998). Survey ofAmericans with disabilities. Washington, D.C.: NationalOrganization on Disability.

Lane, H. (1990). Bilingual education for ASL-using children. In M. Garretson (Ed.), Eyes, hands, voices:Communication issues among deaf people. A DeafAmerican Monograph, 40 (1,2,3,4,), 79-85.

Larson, K. (1992). Explaining the link between disabilities inurban youth, dropping out, and integration into society. Apaper presented to the OSEP Spring LeadershipConference, Washington, D.C.

Lyon, G.R. (1999). The NICHD research program in readingdevelopment: a summary of research findings. New York:The National Center on Learning Disabilities.

McLaughlin, M.J. (1999). Access to the general educationcurriculum: Paperwork and procedure on redefining“special education.” Journal of Special EducationLeadership, 12, 9-14.

Snow, C. (Ed.). (1998). Preventing reading difficulties inyoung children. Washington, D.C.: NationalResearch Council.

Strong, M. & Prinz, P. (1997). A study of the relationshipbetween American Sign Language and English liter-acy. Journal of Deaf Studies and Deaf Education, 2.

U.S. Department of Education (1995). Seventeenth annualreport to Congress on the implementation of theIndividuals with Disabilities Education Act. Washington,D.C.: U.S. Government Printing Office.

U.S. Department of Education (1996). Nineteenth annualreport to Congress on the implementation of theIndividuals with Disabilities Education Act. Washington,D.C.: U.S. Government Printing Office.

Strategies to Improve Results for Students with Disabilities

36Journal of Special Education Leadership 12(2) • November 1999

Wagner, M., Blackorby, J., Cameto, R., Hebbler, K., & Newman, C. (1993). The transition experience of young people with disabilities. Palo Alto, CA:SRI International.

Walker, H.M., Severson, H.H., & Fell, E.G. (1995). The earlyscreening project: A proven child find process. Longmont,CO: Sopris West, Inc.

About the AuthorThomas F. Hehir, Ed.D., formerly was the director ofthe Office of Special Education Programs, Office ofSpecial Education and Rehabilitative Services, U. S.Department of Education, Washington, D.C.Currently, Dr. Hehir is an instructor in the School ofEducation, Harvard University, Cambridge, MA anda Distinguished Scholar, Education Development

Center, Inc., 55 Chapel Street, Newton, MA 02158. E-mail: [email protected]

Additional NoteEach of the goals highlighted in this article will bedelineated by nationally recognized authors in abook forthcoming from the Council of ExceptionalChildren with support from the U. S. Department ofEducation. For further information about the book,contact the Council of Exceptional Children, 1920Association Drive, Reston, VA 20191-1589 or call (888) 232-7733, TTY (703) 264-9446.

Strategies to Improve Results for Students with Disabilities

Inclusive Assessments: A Means to an EndSuperintendent Commentary

Edward Lee Vargas, Ed.D.Ysleta Independent School District, Texas

37Journal of Special Education Leadership 12(2) • November 1999

In an era of standards-based reform and high stakestesting, community expectations for higher levels

of achievement and accountability for all students’learning are driving everyone’s agenda. This shouldmean that school districts, schools, and classroomsare acting in ways that produce much higher levelsof achievement than before for each and every stu-dent, particularly those placed in special educationprograms. As superintendent of a large urban dis-trict, there is no doubt in my mind that we muststart with the belief and expectation that every singleone of our students can and will achieve at muchhigher levels. The extent to which one sees this as aproblem or an opportunity is determined in largepart by the beliefs and skill levels of the person con-fronted with this challenge. Further, it is not justone’s individual beliefs and skills, but the beliefs andskills of his or her associates. Superintendentsshould surround themselves with skilled specialeducation administrators to more effectively meetthese expectations with opportunities.

Superintendents who work closely with theirspecial education directors can more effectively andefficiently reach into those classrooms with the mostdiverse of student learners to first inclusively assesseveryone. Inclusive assessment findings can then beused to establish the baseline from which toimprove, raise expectations, and create a sense ofurgency to implant more effective research-basedmethods that actually increase student performanceresults as an ongoing process of continuousimprovement. Superintendents and special educa-tion directors together can then raise the bar and theachievement of all students, particularly those who

have traditionally not performed at high levels suchas students in special education, students with lim-ited English proficiency, and students from socio-economic circumstances that place them at adisadvantage. After all, superintendents and specialeducation directors have so much in common. Theyboth, in all practicality, run parallel systems withtheir own policies, rules, regulations, processes,funding streams, compliance, accountability system(IEPs), parent involvement requirements, and part-nerships with community and agency groups to helpservice students and achieve educational goals. As aformer special education director in Seattle,Washington, that is what I did. As a superintendentfor one of the largest urban districts in Texas, that iswhat I do now. My superintendent in Seattleincluded me directly in every district and schoolreform initiative and because of that, I made surethat special education was a part of each one ofthose initiatives. The same was true when I was anassistant superintendent in Santa Ana, California.The result was that all students, particularly those inspecial education, benefited from every one of thosechange initiatives.

Superintendents who work closely with their spe-

cial education directors can more effectively and

efficiently reach into those classrooms with the

most diverse of student learners to first inclusively

assess everyone.

38Journal of Special Education Leadership 12(2) • November 1999

Superintendents are ultimately responsible forensuring higher levels of achievement for all stu-dents. “All” includes those students who areexcluded by their IEP committee from school anddistrict assessments and those who are not. Higherstandards and truly inclusive assessments providesuperintendents and special education directors withan opportunity to work more closely together thanever before and to reexamine each programs’ effec-tiveness and then do something about it. Theyshould use the assessment data to ask a critical ques-tion of each and every school level program, “Is thisparticular program effectively raising academicachievement, improving the social-emotional growthand development for each student, and enhancingthe school climate for all?” If the answer is yes, itshould be retained or expanded. If the results showpotential, it should be modified so that the answer isyes. If the answer to the question is no, it should beeliminated, and fast. Effective inclusive assessmentpractices can help a superintendent drive ineffectiveprograms out and more effective programs in.

Effective inclusive assessment practices can help a

superintendent drive ineffective programs out and

more effective programs in.

When superintendents understand and useinclusive assessment practices, they can better lever-age needed reforms, determine their effectiveness forall students, including those placed in special educa-tion, and use them as an effective strategy forincreasing accountability for student achievement.Expanding the Individualized Education Plan con-cept to every single classroom and school in terms ofsetting specific, measurable, and time-framedachievement level targets for every student can

increase accountability, particularly when the class-room and school level results are directly tied to theperformance of the school’s traditionally lowest per-forming groups. School systems that anchor theirdistrict and school achievement levels in this way tospecial education students and other student groupswho have traditionally not achieved at high levels,can then in reality demonstrate the overall effects oftheir reform efforts. In other words, before victorycan be declared in the school reform movement,improvement must be demonstrated for every sub-group of students and at substantially higher levels.If all truly means all, then this is the right thing todo, and special education directors should helpsuperintendents raise the bar and the achievement ofall, starting with every special education student.

Superintendents who work closely with theirspecial education directors and administrators canfind themselves at an advantage in today’s highstakes testing environment. Knowing what specificreform initiatives work and do not work for all stu-dents starts with assessing all students. Effectiveinclusive assessment programs will arm the superin-tendent with data and information for improveddecision making to increase effective educationalpractices and jettison ineffective ones.Superintendents and special education directors whowork together can use inclusive assessments as anopportunity to make increased student achievementa districtwide reality, not for some, not even formost, but for all.

About the AuthorEdward Lee Vargas, Ed.D., is the superintendent for the Ysleta Independent School District, 9600 SimsDrive, El Paso, Texas, 79925-7225. E-mail:[email protected]

Inclusive Assessments

CASE IN POINTIncluding Students with Disabilities in Large-Scale Testing One Case at a Time

David J. Chard, Ph.D. University of Texas at Austin

39Journal of Special Education Leadership 12(2) • November 1999

At Buena Vista Elementary School in central Texas,Ms. Guadalupe Martinez* is busy planning liter-

acy lessons for her third grade class. This is Ms.Martinez’ eighteenth year of teaching and her sixthyear at Buena Vista. She takes pride in her profes-sion and understands the importance of teachingthird grade in Texas, the year that children must takethe Texas Assessment of Academic Skills Test(TAAS). Ms. Martinez remembers the early years ofthe statewide accountability system and how sheand her colleagues feared it would negatively trans-form the school’s environment. Once she understoodthe value of assessing a school’s performance onimportant educational outcomes, she grew less fear-ful. Now, however, growing pressures to include allof her students in the testing process have causedsome of the fear to return. Last year, Ms. Martinezexempted two students with disabilities from takingthe test. The first child, Brian, has a severe readingdisability. He demonstrates difficulties with manipu-lating the sounds in words and decoding even sim-ple words. Consequently, he receives intensive smallgroup instruction from a special education teacher ina resource room setting in addition to the literacyinstruction he receives from Ms. Martinez. The sec-ond child, Lucia, has severe physical disabilities. Sheis unable to perform any fine motor skills such asholding a pencil, and she uses an electric wheelchairto move about the classroom. Cognitively, Lucia isdeveloping typically, she is able to answer questionsverbally about stories she reads and engages inclassroom discourse successfully. Ms. Martinez hadserious reservations about whether Brian or Luciawould be able to take, let alone pass, the TAAS. Herconcern about Brian was that the test would be inac-

cessible to him because of his lack of reading fluencyat the level required and in the time allotted. ForLucia, her concerns focused on the girl’s inability torespond to the paper/pencil format of the test.Despite her concerns, Ms. Martinez has been advisedby the Special Education Director of her school dis-trict that next year students like Brian and Lucia willbe included in the TAAS testing or in an alternatetest unless a rationale is provided for why theyshould not be tested.

Ms. Martinez, like many parents, teachers, andschool leaders, is concerned about recent changes tospecial education law that require all students withdisabilities be included in local and state large-scaleassessments. Her concerns include the following: • If students with disabilities are not included in

large-scale assessments, the test results willinform us only of the suitability of educationalprogramming for students who are not in specialeducation;

• The high stakes nature of many large-scaletests may put undo burden on students withdisabilities;

• It is unclear how the results from large-scale testswill influence the graduation status of studentswith disabilities; and

• It is unclear how the results for students withdisabilities will be interpreted in terms of hold-ing a school’s special education programmingaccountable. These concerns warrant further attention as they

render important implications for students like Brianand Lucia and their parents, teachers, and adminis-trators. The purpose of this commentary is toaddress each of Ms. Martinez’ concerns and to

*pseudonym

40Journal of Special Education Leadership 12(2) • November 1999

remind the reader of the case-by-case nature of thedecisions to be made regarding including studentswith disabilities in large-scale assessment systems.

Similar concerns have been voiced when otherbold changes have been proposed in special educa-tion. To put these recent changes into context, a bitof history may be helpful. In 1975 advocacy groupsstaged a “sit-in” in the offices of the then Secretaryof the Department of Health, Education, and Welfaredemanding publication of the much awaited regula-tions that would finally enforce the Education for AllHandicapped Children’s Act (EAHCA; 34 U.S.C.1400 et seq.). Parents, advocates, and educators couldhardly have envisioned the effects their work thenwould have on the culture and day-to-day opera-tions of our public schools today. Of course, at thetop of this list is that all students with disabilities areentitled to a free, appropriate public education in theleast restrictive environment. Successive reauthoriza-tions of special education law, now the Individualswith Disabilities Education Act (IDEA), have refinedand extended the act’s provisions to reach a broaderrange of students and to make educational program-ming for students with disabilities as strong as thatprovided to their nondisabled peers.

As state and local education agencies havemoved toward the use of educational standards andlarge-scale testing for the purposes of educationalaccountability, inclusion of students with disabilitiesin these testing systems has become an importantgoal. The underlying motivation here is to identifyareas in which a school’s performance is acceptableand areas in which a school needs improvement. Asa result of the 1997 reauthorization of IDEA and itsattendant regulations, all students must be includedin these systems by July 1, 2000. There may be manybenefits to testing all students, but the one thatstands out is that a school will have a more accuratepicture of its teaching and learning performance ifall students are tested (Thurlow & Johnson, thisissue). If children receiving special education ser-vices are exempted from the accountability system,as they have been historically, the test results willinform us only of the suitability of educational pro-gramming for students who are not receiving specialeducation. Consequently, improvement plans thatstem from the results of large-scale assessments willnot address the needs of all children.

The potential risks of including students withdisabilities in large-scale testing efforts may warrantconcern. Certainly, there is the possibility that the“high stakes” nature of many statewide tests willput undo burden on students with disabilities. Forexample, in Texas, if a school has a minority population that exceeds ten percent of the totalschool population, the aggregate test score for thatsubpopulation must meet or exceed ninety percentcorrect in order for the school to reach recognizedstatus. If students with disabilities are included inlarge-scale assessments and a school’s performancesuddenly does not meet expectations, the school’scommunity may be tempted to blame students withdisabilities for the poor performance. While notdeliberate, this kind of “scapegoating” is damagingfor the school community. Additionally, there aremany unanswered questions about how the resultsof testing students with disabilities will be inter-preted. For example, Thurlow & Johnson (this issue)and Green & Sireci (this issue) have both outlinedthe legal status of how statewide test results mayinfluence whether a school promotes a student fromgrade to grade and even the graduation status ofstudents with disabilities. The graduation status ofstudents with disabilities is not unique to the dis-course on assessment, but many states are now link-ing student graduation with successfully passing thestatewide assessment (e.g., Oregon). Undoubtedly,how state testing will be linking to the graduation of students with disabilities warrants continued discussion and research.

Many states are introducing a policy of linkingfailure on the statewide assessment of literacy tograde retention. This policy is designed to improveinstruction in reading for all children and to provideadditional instructional resources when childrenmake inadequate progress. It is not clear how stateswill use assessment results for students with disabil-ities in terms of grade retention and holding schoolsaccountable for the performance of special educationprograms. The U.S. Supreme Court argued in Boardof Education v. Rowley (1982) that students werereceiving educational benefit under IDEA if theywere passing from grade to grade. The simplicity ofthis ruling belies the complexity of the decisions thatmust be made to determine if a student has acquiredthe necessary knowledge and skills to pass to the

Including Students with Disabilities in Large-Scale Testing

41Journal of Special Education Leadership 12(2) • November 1999

Including Students with Disabilities in Large-Scale Testing

next grade level. Little is known about how muchprogress we should expect students with disabilitiesto make from year to year in critical knowledge andskill areas. Consequently, if a student does not per-form to a set criterion on the large-scale assessment,should this determine whether the student passes tothe next grade level, be an evaluation of the perfor-mance of the special education program, or both?

Depending on the IEP Significant changes to special education law haveoften led parents and schools down a long roadmarked by debate and negotiation. The final mile-post in the road has been, and remains, the individu-alized education program (IEP). Early in specialeducation history, the IEP was a document designedby a team of professionals and the student’s parentsthat defined the special education and related ser-vices a student was to receive. The function of theIEP team has expanded significantly to include deci-sions, among others, about a student’s access to thegeneral education curriculum. While policymakersrhetorically state that all students with disabilitieswill be included in large-scale assessment efforts, inreality the IEP team is now also responsible for mak-ing this decision on an individual basis for all stu-dents with disabilities. Many IEP teams embark onthis task without criteria to guide their decisions.Certainly, as we examine the new requirements ofIDEA related to large-scale testing, we must bear inmind that the power of special education past andpresent lies in the IEP. IDEA does not require that allchildren be provided a free appropriate education inthe general education classroom but rather that it isin the least restrictive environment. This languageallows IEP teams to make decisions in the best inter-est of the child with whom they are primarily con-cerned. Similarly, IDEA does not require that allchildren receive the same educational programmingbut depends on the professional judgment of IEPteams to determine what is appropriate. As with allthe critical elements of the special education process,inclusion of a student with disabilities in state anddistrict testing will also depend on the professionaljudgment of the IEP team.

Fortunately, as IEP teams add the decision ofhow to include a student with disabilities in theassessment system to the list of other decisions they

must make, they will be informed. It is incumbenton state and local policymakers to establish whatwill be acceptable accommodations, modifications,and alternate forms of large-scale assessments forstudents with disabilities. Each of these optionsleads to unique implications regarding the validityof the assessment. For example, an accommodationdoes not alter what is being assessed (e.g., readingcomprehension) but simply how the student’sresponses to the assessment are gathered (e.g., oralvs. written). In contrast, a modification may changewhat is being assessed (e.g., reading comprehensionwhen the student reads the passage versus listeningcomprehension when the teacher reads the passage).Important ongoing research on various accommoda-tions, modifications, and alternate test forms(Tindal, Helwig, & Hollenbeck, this issue) shouldinfluence state and local policies which in turn willprovide IEP teams with a basis upon which to maketheir decisions.

SummaryIn the case of Brian, because his disability is in thearea of reading, a possible option to the readingtest in 3rd grade would be to have the test readaloud. In this circumstance, however, the constructvalidity would be sacrificed. Consequently, itwould be reasonable for the IEP team to recom-mend an alternate test. In this circumstance, even ifBrian performs well on the alternate test, who willdetermine whether his performance warrants apass to the following grade?

In contrast, in Lucia’s case, because of her physi-cal disability, her IEP team may recommend thatLucia take the same test as her nondisabled peersbut that she receive an accommodation. That accom-modation may be to read the text from a computerscreen rather than from paper. Again, furtherresearch and policy will clarify whether this isindeed an appropriate accommodation. For the nearfuture, IEP teams may have to base their decisionson a logical analysis and reconsider their decisionsonce research and policy are advanced. The require-ment that all students with disabilities be includedin the accountability systems now so common inpublic education is indeed bold. As Tindal, Helwig,& Hollenbeck (this issue) state, this change is goingto force us to “rethink our notions of ‘standardization’”

(p. 15). Indeed, if it is the case that large-scale sys-tems of accountability are having a positive effect onthe quality of general education in states like Texas,special education must also benefit. Whether indi-vidual children will be included in the testing is upto their IEP team to decide—one case at a time.

References Individuals with Disabilities Education Act, 20 U.S.C.S. §§

1401 et seq. (1999). Board of Education of Hendrick Hudson Central School District

v. Rowley, 50 U.S.L.W. 4925, 4932, 458 U. S. 176 (1982).

About the AuthorDavid J. Chard, Ph.D., is an assistant professor in theDepartment of Special Education, 408 SZB College ofEducation, University of Texas at Austin, Austin, TX78712. E-mail: [email protected]

Including Students with Disabilities in Large-Scale Testing

42Journal of Special Education Leadership 12(2) • November 1999

Assistant Professor—Tenure Track

Teach doctoral, educational specialist, and master’s students in areas involving:

Transcript evidence of graduate degrees in special education, human services-type administration from accredited Americaninstitutions required, and with the earned doctorate in a field related to the goals of the department. Demonstrated evidence ofadvanced sign communication skills. Verifiable evidence of three or more years of administrative line experience within a schoolor agency setting. Written evidence of scholarly publications and/or grantsmanship.

Send dossier to Chair, Department of Administration & Supervision, Gallaudet University, Washington, DC 20002-1695.

Enquiries encouraged at (202) 651-5525, Fax: (202) 651-5749 or e-mail: [email protected]. An EOE employer.

• Advanced seminars in special education policy formulationstrategies and procedural implementation methods, espe-cially in the area of deafness

• Advanced seminars in the administration of human services agencies at the local, state, or national levels, especially in the area of deafness

• Federal and state legislative enactments in special education

• Innovative curriculum initiatives in leadership, includingsummer institutes for visiting professionals, and the seniors’honors program for Gallaudet’s undergraduate students

43Journal of Special Education Leadership 12(2) • November 1999

Manuscript Guidelines and Editorial PoliciesThe Journal of Special Education Leadership, publishedby the Council for Administrators of SpecialEducation, seeks articles that capture an administra-tor’s attention by providing useful information thatstimulates new ways of thinking about managingand leading. Only articles that have been validatedand accompanied by accepted theory, research, orpractice are sought.

The Journal of Special Education Leadership’s goals are:1. To provide fresh ideas and perspectives,

grounded in recent advances in administrativetheory and research, on contemporary issues thatadministrators must face.

2. To become a primary source of useful ideas forthose who seek to educate present and futureadministrators of special education programs.

3. To become a forum through which practicingadministrators of special education programs canchallenge the meaningfulness of translations ofadministrative theory and research.

Contributors to each issue will include practicingadministrators, researchers, policymakers, or othersinterested in special education administration. Thepurpose of this arrangement is to encourage interac-tion among individuals within those roles in devel-oping articles. Interactions may include any of thefollowing: a jointly authored manuscript, an inter-view preceded or followed by commentary writtenby the interviewer, and a follow-up article that isspecifically linked to the theory and/or research article that provides examples from the field andimplications for administrators in similar situations.

A typical article might begin with either a briefcase illustrating the primary theme, or posing certainquestions and issues that special education adminis-trators need to address. A typical article will also sat-isfy the academic reader who seeks more than justopinions and wants to see a serious effort at connect-ing ideas to accepted theory and research.

With respect to style and format, manuscriptsshould:

• Be accompanied by a letter signed by theauthor(s),

• Have a separate title page that identifies theauthors (the names(s) of the author(s) should notappear anywhere on the manuscript, except onthe title page),

• Be written in clear, straightforward language,avoiding jargon and technical terms,

• Conform to APA format (see Appendix B of APA Publication Manual, 4th edition, 1994), particularly:- Entire manuscript is double spaced, with

margins.- All pages are numbered in sequence, starting

with the title page.- All references in text are listed and in complete

agreement with text citations.- All author identification information, including

professional title and affiliation, address, and phone number, is on the title page only.

- Cover letter states the manuscript is original, not previously published, and not under consideration elsewhere.

• Include at the beginning an Executive Overviewof 3-5 bulleted major points made in the article,

• Use subheadings but not the traditional onessuch as “Introduction”; use, instead, “The FutureChallenge” or “Do Seamless Delivery Systemshave a Future?”

• For the purpose of documentation, cite notes inthe body of the paper using superscript notenumbers, and

• Include a biographical sketch of each author thatincludes name, title, and place of employment.

Authors are encouraged to get feedback fromcolleagues and practitioners on early drafts. A papercan be improved dramatically when knowledgeablereviewers are asked for reactions in advance of submission.❒ Manuscripts should be double-spaced and no

more than 15 pages in length, including figures.When questions arise regarding issues of gram-mar or style, authors should refer to thePublication Manual of the American PsychologicalAssociation, 4th edition.

Call For Papers

44Journal of Special Education Leadership 12(2) • November 1999

The Journal of Special Education Leadership is published two times per year. The issues vary withsome being thematic. Each issue includes 4-5 articlesand 1-2 administrative briefs/technical notes.

Review ProcessSelection of manuscripts for publication is based on a blind peer review process. However, all manuscripts are screened first by the editor. Thosemanuscripts that do not meet the manuscriptrequirements, or that are not consistent with the purpose of the journal, are not forwarded for peerreview. The author is either notified that the man-uscript is not acceptable for the Journal of SpecialEducation Leadership, or requested to make changesin the manuscript so that it meets requirements.Copies of the manuscript are not returned to theauthor in either case.

Manuscripts that are consistent with the purposeof the journal are sent out for peer review. Reviewerswill not know the identity of the author.

Based on the blind reviews, the Journal of SpecialEducation Leadership editor will communicate theresults of that review to the author. The decision that is communicated to the author will be one of the following:• Acceptable, with routine editing• Acceptable, with revisions indicated by editor• Unacceptable

When a decision is made that a manuscript is unacceptable for the Journal of Special EducationLeadership, it may be recommended that it be sent to a journal of one of the CEC Divisions. This recommendation does not mean that the manuscriptwould be automatically accepted by a Division journal; the manuscript would have to go throughthe review process again.

Author Responsibilities FollowingPublication Acceptance

After a manuscript is accepted for publication inthe Journal of Special Education Leadership, the authoris responsible for completing the following:• Obtaining publication clearance, if needed,

for a manuscript first presented at a professionalmeeting;

• Acknowledging the funding agency for supported research;

• Verifying the authenticity of all quoted materialand citations and for obtaining permission fromthe original source for quotes in excess of 150words or for tables or figures reproduced frompublished works;

• Preparing camera-ready copies of all figuresincluded in the article;

• Assigning literary rights to CASE by signing aCopyright Transfer Agreement;

• Sending two (2) paper copies of the revised manuscript to the Journal of Special EducationLeadership’s Editorial Office; and

• Sending an exact copy of the revised manuscriptto the Editorial Office on a floppy disk (3 1/2”),with the document saved in WordPerfect,Microsoft Word, or WordPro format, if possible.(Acceptable alternatives are ASCII format, on aDOS or Mac platform, however these formats arenot preferable.)

Author Checklist

Before sending a manuscript, please complete theAuthor Checklist below. This will help ensure thatyour manuscript is not screened out or returnedbefore review.

❒ Manuscript is consistent with the purpose of the journal.

❒ Manuscript is no longer than 15 pages total.

❒ Manuscript conforms to APA format (seeAppendix B of APA Publication Manual, 4th edition, 1994).

Send 5 copies of manuscript to:Dr. Mary Lynn Boscardin, EditorJournal of Special Education Leadership175 Hills-SouthSchool of EducationUniversity of MassachusettsAmherst, MA 01003

Acknowledgment of receipt of your manuscript will be sent to you within 2 weeks. Review of yourmanuscript will occur within 6 weeks.

Call for Papers

One-Minute Academic Functional Assessment and Interventions“Can’t” Do It… or “Won’t” Do It?

by Joe Witt, Ph.D. and Ray Beck, Ed.D.

Save time spent trying strategies that don’t work. Using the One-Minute assess-ment process, you can determine when your students can do the work butwon’t, when they need more practice to become fluent, and when they can’tdo the work (need more teaching). Different intervention strategies are needed

for each of these situations, and this book provides dozens. Includes reproducibleforms, worksheets, and checklists. 212 pages. Grades K-12. CA119MIN. $19.50.

Conducting Functional Behavioral Assessmentsby Ron Nelson, Ph.D.; Maura Roberts, Ph.D.; and Deborah Smith, Ph.D.

“… a very practical guide to accomplishing what seems a complex and overwhelming task.”—Amy Donovan, NASP Communique’

Step-by-step procedures for gathering pertinent behavioral information helpyou develop a behavioral plan that matches interventions and supports to the function(s) of targeted behaviors. Examples of common classroom scenarios illustrate how to determine the specific antecedents and conse-quences of target behaviors. Four easy-to-use Data Collection Forms(Interview/Self Report; Observation and Analysis; Temporal Analysis andRanking; and Summary Analysis) help you collect and organize your data. 49 pages.Grades K-12. CA112SET. Book with 10 sets of four forms. $19.50.

Functional Assessment Intervention ProgramSoftware That Makes Conducting FBAs Easier

by University of Utah; Utah State University; and the Utah State Office of Education

Fast and easy to use, this software analyzes your responses to questions and sug-gests strategies. Questions help you consider the factors that may influence stu-dent behavior, including: physical and emotional condition, setting characteristics,persons present, and associated events. By confirming or discarding factors, yougenerate a report that contains confirmed factors, motivations, strategies, and alist of suggested interventions. For Windows® 95/NT and 3.1. Includes 20-pageUser’s Manual. Grades PreK-12. CA117FAIP. $69.

With MC/Visa: (800)547-6747 or www.sopriswest.comOr Fax your PO: (303) 776-5934

Helping You Meet the Needs of At-Risk Students

Working Together:

The Art of

Consulting &

Communicating

by Anita DeBoer, Ed.D.

Help your staff to work produc-tively in collaboration efforts withthis practical guide and inter-personal style questionnaire.Interactive strategies help you to develop the essential skills forsuccessful collaboration:

• Trust-building• Listening• Facilitating• Questioning• Communicating• Peer problem-solving

With these skills for successful collaboration, you and your staffcan act proactively—rather thanreactively—when difficult issuesarise. Techniques are provided forworking with the many inter-personal styles, and the 30-itemInterpersonal Style Questionnairehelps you determine primary styles,so that you can more easily refer tothe techniques most relevant toyour staff. Models are also includedfor successful consulting. 244 pages.Grades K-12. CA71SET. $47.50.

Working Together:

What Collaborative

Teaching Can

Look Like

by Susan Fister, M.Ed. and Anita DeBoer, Ed.D.

Help your teaching staff workmore productively as a team with this video and accompanying24-page manual. Designed for staffdevelopment, this video demon-strates an effective model of collab-orative teaching and promptsviewers to apply the model to theirown circumstances. Grades K-12.CA70VIDEO. $49.

Working Together:

Tools for

Collaborative

Teaching

by Anita DeBoer, Ed.D. and Susan Fister, M.Ed.

This great resource for team teach-ing makes implementing collabora-tive practices in the classroom easy.Models for successful collabora-tion; ideas on how to effectivelyplan for and schedule collaborativeteaching; tips on delivering qualityinstruction; and systematic meth-ods of evaluating collaborativeefforts are all included. Use aloneor in conjunction with WorkingTogether: What Collaborative TeachingCan Look Like. 138 pages. Grades K-12. CA70COTEACH. $19.50.

Foster Collaboration

Between General and Special Education Staff

Helping You Meet the Needs of At-Risk Students

With MC/Visa:

(800) 547-6747

or www.sopriswest.com

Or Fax your PO:

(303) 776-5934

IEP ConnectionsFor a Dynamic, Living,

and Practical Process

Steve Kukic, Ph.D. and Judy Schrag, Ph.D.

In collaboration with National Association of State Directorsof Special Education (NASDSE)

Meet IDEA Mandates~ and ~

Reenergize Your IEP Process

step-by-step guidelines and reproducible supportsprovide an easy-to-follow IEP framework that makes strong,positive connections among assessment, educational stan-dards, instruction, and coordinated service planning. WithIEP Connections, you’ll have everything you need to:

• Strengthen connections between families and schoolsand between IEP teams and general education staff.

• Determine appropriate curriculum and instructionalmodifications.

• Build trust among IEP team members.• Collaborate, solve problems, and resolve conflicts

effectively.• Engage in a successful decision-making process.• Involve students in the IEP process.

Four components incorporate the new IDEA mandateswhile keeping the focus on serving individual students:

• The Leader’s Guide provides a decision-making processfor determining the modifications that will connectIEPs to general education curriculum, methods forsolving problems, and more. Includes reproducibles.

• The Team Member’s Manual provides all the tools and information to fully participate in this reenergizedIEP process.

• The IEP Connections Video illustrates the ways in whichIEPs connect to assessment, standards, benchmarks,general education curriculum, instruction, coordinatedservice planning, communication with parents and students, and—most importantly—positive studentresults.

• The set of four Support Forms (50 of each) make implementation even easier:� Portfolio Conference/Assessment Guide� Teacher’s Self-Rating Scale� Team Member Self-Rating� Team Effectiveness Inventory

IEP Connections Kit (Leader’s Guide, Team Member’s Manual, Video).CA111SET. $145.

Complete set of the four Support Forms (50 of each). CA111FOR. $65.

Set of 10 IEP Team Member’s Manuals. CA111TMG. $49.

Professional Development is also available.

Order Now!With MC/Visa: (800) 547-6747 or www.sopriswest.com

Or Fax your PO: (303) 776-5934Helping You Meet the Needs of At-Risk Students

Grades K-12

Simple,

Back Cover:Use new Quark file