School Psychology Quarterly, Vol. 21, No. 4, 2006, pp. 359-374

The Solomon Effect in Learning DisabilitiesDiagnosis: Can We Learn from History?

Stefan C. DombrowskiRider University

Randy W. Kamphaus, Melissa Barry, Amber Brueggeman,Sarah Cavanagh, Katie Devine, Linda Hekimoglu, and Sarah Vess

University of Georgia

The Individuals with Disabilities Act (Individuals with Disabilities Education Im-provement Act; IDEIA, 2004) has been reauthorized, and new parameters for defin-ing learning disabilities (LD) have been established that provide more flexibility forcorresponding state and local regulations. The field now has a unique opportunityto shape the practice of LD diagnosis and should consider important conceptual,theoretical, empirical, economic, legal, and practical issues related to LD diagno-sis. This article highlights five key recommendations for the diagnostic definition oflearning disabilities: (1) the definition needs to be unambiguous; (2) it must be uni-versally accepted across professions, researchers, and governmental entities; (3) itmust incorporate clearly defined subtypes of learning disabiI ities; (4) it must be em-pirical ly supported; and (5) it must point to valid, reliable, and cost-effective pro-cedures for the identification of children with and without learning disabilities.Consideration of these points may help the field to avoid repeating past mistakesand returning to the folly of poor LD diagnostic practice.

The Individuals with Disabilities Education Act (Individuals with Dis-abilities Education Improvement Act; IDEIA, 2004) has beenreauthorized, and revised parameters concerning learning disabilities(LD) diagnosis have been established. One of the more dramatic, yet de-fensible, changes to IDEIA legislation was to eliminate reliance on theIntelligence (IQ)-Achievement discrepancy model as the basis for LDdiagnostic decisionmaking. LD diagnosis may now be predicated upona comprehensive evaluation in which "a variety of assessment tools and

Address correspondence to Stefan C. Dombrowski, Rider University, 2083 LawrencevilleRoad, Lawrenceville, NJ 08648; E-mail: sdombrowski@rider.edu.

359

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

360 DOMBROWSKI ET AL.

strategies" are used "to gather relevant functional, developmental, and aca-demic information" [Section 614 (b)(6); IDEIA, 2004]. The federal IDEIAlanguage also includes reference to the use of response to scientific, re-search-based instruction: "In determining whether a child has a specificlearning disability, a local education agency may use a process that determines ifthe child responds to scientific, research-based intervention . . . " [Section 614(b)(6); IDEIA, 2004]. Using the federal regulations as a backdrop, variousstate departments of education and legislative bodies will be faced withthe task of reformulating their particular LD diagnostic regulations. Inturn, this will shape the LD classification approach that will ultimatelybe adopted by states and school districts for years to come.

The federal regulations are fairly open-ended, allowing state legisla-tures flexibility in establishing corresponding LD diagnostic parame-ters. The field, therefore, has a critical opportunity to pause and reflectupon the accumulated research and practical realities related to LD di-agnosis so as to avoid repeating past mistakes. Stanovich (1986) used abiblical reference (i.e., the "Matthew Effect"1) to make the case againstthe use of IQ tests and therefore the IQ-Achievement discrepancy modelfor reading disability diagnostic decisionmaking. Following this meta-phorical precedent, and in light of the newly revised IDEIA LD diagnos-tic language, it seems appropriate to allude to what we shall call the"Solomon Effect."

The Solomon Effect refers to a pithy maxim from the Old Testamentchapter of Proverbs attributed to King Solomon. The Solomon Effect2 en-courages us to learn from history so that we avoid repeating it. In a poi-gnant, and perhaps crude, metaphor, Proverbs 26:11 cautions us to becircumspect about revisiting a past practice that has been documentedas ineffective: "Just as a dog returns to its vomit, so does the fool return to

1. The Matthew Effect is a biblical reference to the notion of the rich getting richer and thepoor getting poorer. Stanovich (1986) explained this metaphor in the context of good andpoor readers. He explained that students who are strong readers will expand their fund ofgeneral information, increase their vocabularies, and improve their verbal comprehen-sion. In turn, this will lead to increased IQ test scores. Stanovich described the opposite ef-fect in children who struggle with reading. Over time, poor reading skills will result indepressed IQ test scores, resulting in the reduced likelihood that the child will experience adiscrepancy and therefore an LD classification.

2. The Solomon Effect is to be distinguished from Solomon's Choice. Solomon's Choice isa literary allusion that refers to the way in which King Solomon arrived at a decision regard-ing whether a child belonged to one of two women. When both women approached himclaiming to be the child's mother, King Solomon could not distinguish who indeed was thebiological mother. Thus, he declared that he would just cut the child in half. When one of thewomen viscerally protested and stated that the other woman could have the child, Solomondiscerned that the child belonged to the woman in whom strong emotions were evoked andwho agreed to spare the life of the child by allowing it to go with the other woman.

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

THE SOLOMON EFFECT IN LD DIAGNOSIS 361

his folly." Let us not return to the folly of poor LD diagnostic practice. In-stead, let us learn from history to ensure that the definitional, practical,taxonomic, and measurement problems of the past are not repeated.

These aspects of learning disability diagnosis must be faced squarelyto develop diagnostic approaches that are universally acceptable(Shepard, 1989). Within this article, we present an overview of some ofthe myriad issues confronting the field. In particular, we recommendthat the field attend to five key issues when establishing a diagnostic def-inition of learning disabilities: (1) The definition needs to be unambigu-ous; (2) it must be universally accepted across professions, researchers,and governmental entities; (3) it must incorporate clearly defined sub-types of learning disabilities; (4) it must be empirically supported; and(5) it must point to valid, reliable, and cost-effective procedures for theidentification of children with and without learning disabilities. Thefield must attend to these issues to avoid the hasty adoption of an LD di-agnostic approach that may be en vogue today, but deemed ineffectivetomorrow following subsequent research.

THE DEMISE OF THE DISCREPANCY:A HISTORICAL PERSPECTIVE

Throughout the past century, researchers have investigated children'sdifficulty with learning to read and write, despite no discernable sign ofcognitive or sensory disability (Hinshelwood, 1917; Orton, 1925). To-day, these children would likely be diagnosed with LD. The early labelsascribed to these children were medically oriented (e.g., brain injured,perceptually impaired, dyslexic, and neurologically impaired) and con-sistent with the clinical venues in which they were being served. As edu-cational settings became increasingly responsible for children withlearning difficulties, Samual Kirk, a professor of special education, in-troduced the term "learning disabilities" in an effort to move the fieldaway from a medically oriented conceptualization:

A learning disability refers to a retardation, disorder, or delayed de-velopment in one or more of the processes of speech, language,reading, writing, arithmetic, or other school subjects resulting froma psychological handicap caused by a possible cerebral dysfunctionand/or emotional or behavioral disturbances. It is not the result ofmental retardation, sensory deprivation, or cultural andinstructional factors (Kirk, 1962, p. 263).

Kirk's definition had a significant influence on subsequent generationsof LD definitions, including two that are widely used today to diagnose

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

362 DOMBROWSKI ET AL.

children: Those of the U.S. Office of Education (IDEIA) and the Diagnos-tic and Statistical Manual of Mental Disorders, 4th Edition (DSM-IV; Amer-ican Psychiatric Association, 1994). Kirk's initial definition did notcontain specific discussion of a discrepancy between intellectual abilityand achievement (Kirk & Bateman, 1962). Instead, his definition made ageneral reference to a psychological disorder resulting from cerebraldysfunction (Mercer, Forgnone, & Wolking, 1976).

Kirk's initial definition was still medically oriented. Thus, to moveaway from a medically oriented LD definition that was less education-ally relevant, the field adopted the discrepancy model as a primary de-fining characteristic. Barbara Bateman (1965), one of Kirk's students,was the first to provide an LD definition that contained reference to adiscrepancy between ability and achievement. This definition wasthought to have greater educational relevance since it was less neurolog-ically based and more parsimonious. Rutter and Yule (1975) and Yule(1973) wrote the first articles that provided an empirical basis for theIQ-achievement discrepancy, and their research influenced the field'sconceptualization of LD. However, numerous studies have challengedRutter and Yule's conclusion that an IQ-achievement discrepancymodel can be validly used for LD diagnosis (Aaron, 1997; van der Wissel& Zegers, 1985).

Although the discrepancy model was adopted in order to resolveproblems with LD diagnosis, the effect was to exchange one definitionalproblem for another (Aaron, 1997; Dombrowski, Kamphaus, &Reynolds, 2004; Lyon, 1996; Siegel, 1999). The discrepancy model hasmade the LD definition just as educationally problematic as prior medi-cally oriented definitions. After more than four decades of use and accu-mulated research evidence against the use of the discrepancy model, thefield is poised to discard it for purposes of LD diagnosis, but the LD fieldstill needs to resolve how it will define and diagnose LD.

THE "DEFINING PROBLEM"

Although there have been great gains in research and service for chil-dren with learning disabilities, there are many unresolved issues in defi-nition, assessment, diagnosis, and treatment. In order to solve some ofthese issues, it will be imperative to achieve consensus in the field re-garding an operational definition of an LD. The convening of expertpanels, such as the ten-organization LD Roundtable (U.S. Office of Spe-cial Education, NJCLD, 2002), has made some progress in this regard butquestions remain to be investigated. Another approach to reaching con-sensus is to examine the empirical research on types or subtypes of LD as

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

THE SOLOMON EFFECT IN LD DIAGNOSIS 363

a guide to defining the core features of the disorder and its variousclinical presentations.

For example, subtyping research of the last several decades suggeststhat there are not seven distinct types of LD as defined in the 1977 regula-tions (i.e., oral expression, listening comprehension, written expression,basic reading skill, reading comprehension, mathematics calculation,and mathematics reasoning) (Kamphaus, 2001; Lyon, Fletcher, &Barnes, 2003). Instead, Lyon et al. (2003) identified four broad subgroupsof LDs: reading disorders, mathematics disorders, reading-mathemat-ics disorder, and disorders of written expression. Reading disorderswere further subcategorized into disorders of word recognition,comprehension, and fluency subtypes.

Alternatively, Siegel (1999) proposed two subtypes of LD by conclud-ing:

Over the past 30 years, it has become clear that there are two majorclusters of learning difficulties. The most commonly known is read-ing disability, sometimes called dyslexia. There is no difference inmeaning between the terms dyslexia and reading disability. Anequally prevalent but less commonly known disability is arithmetic(mathematics) disability, sometimes called nonverbal learning dis-ability, developmental output failure, writing-arithmetic disability,or visual-spatial disability. Although there are some similarities be-tween the two major clusters, they are dissimilar enough to be con-sidered distinct entities. Some individuals have difficulties onlywith writing and/or spelling. As these written language problemsusually occur in the context of problems with reading and/or arith-metic and mathematics problems, the existence of a separate writtenlanguage disability has not been clearly established nor is there aclear definition of it, especially in the adult population, (p. 306).

Overall, recent subtyping research is most consistent with the DSM tri-partite definition of reading, mathematics, and writing disabilities, sug-gesting that parsimony should be favored in the diagnostic process byfocusing more on the assessment of this smaller number of broadacademic areas.

The subtyping research is valuable for addressing at least one crucialpart of the LD definitional challenge, the core "symptoms" or "con-structs" constituting the disorder. Other issues remain including thelevel of impairment necessary to be considered LD, the chronicity ofproblems, and the various "rule outs" (e.g., lack of adequate instruc-tion). Science has shown potential for clarifying some aspects of the defi-

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

364 DOMBROWSKI ET AL.

nition of LD and more science should be supported and funded to fleshout the remaining components of the definition.

IMPORTANCE OF CLASSIFICATION: WHY PERSEVERE?

Continuous controversy regarding the definition and diagnosis of learn-ing problems has led individuals and organizations to argue for a newclassification approach that alleviates ambiguity among different pro-fessions, creates more valid procedures for diagnosis, and facilitates ac-curate research (Adelman & Taylor, 1986; Dombrowski et al., 2004).Therefore, it is imperative that the field understand the inherentpurpose of classification.

Classification is a fundamental, innate process in all scientific fieldsbecause it guides our search for information or truth (Wiggins &Schwartz, 1994). Methods of classification create a common nomencla-ture that enhances communication and leads to better understanding ofscientific phenomena (Blashfield & Draguns, 1976). Systems of classifi-cation structure the phenomena in a given field, yet are subject to changeover time with scientific advancement (Keogh, 1983). Professional disci-plines that have service responsibilities, such as school psychology andspecial education, may emphasize the importance of classification forservice delivery, but overlook other important theoretical and empiricalconsiderations (Adelman, 1992; Keogh, 1983). This potential oversight isproblematic because the development of scientific theories would be dif-ficult without a classification system that provides the language todescribe phenomena in a given field.

The nomenclature undergirding the classification of LD should ide-ally be shared among psychologists, psychiatrists, teachers, administra-tors, researchers, and any other professionals who work with LD popu-lations. Problems arise because no singular definition of LD satisfies therequirements of these diverse disciplines (Adelman & Taylor, 1986).Consequently, researchers from different backgrounds apply classifica-tion models that they have learned from their respective fields(Blashfield, 1993; Tucker, Stevens, & Ysseldyke, 1983). For instance,some professionals think that the LD definition should be tied to biologi-cal causes, such as neurological dysfunction, while others think that thedefinition should remain broad so that children with learning disabili-ties that are not as well understood still have their educational needs met(Adelman & Taylor, 1986). Ultimately, accurate classification of the con-struct is necessary to move the LD field forward (Stanovich, 2005). AsShepard stated in her classic 1989 work, "In a research context, a consci-entious effort to make a valid categorization is essential, if knowledge isto be advanced" (p. 547).

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

THE SOLOMON EFFECT IN LD DIAGNOSIS 365

The mental retardation (MR) field suffered the same problem about150 years ago when the disorder was not adequately differentiated fromindividuals with psychotic disorders who were thought to be possessedby demons. The treatment of choice was to imprison and essentiallyabuse the individual with torture-like "treatments" in order to exorcisethe demons (Kamphaus, 2001). It was in this context that the French phy-sicians Itard and Seguin worked to develop a definition of MR and mea-surement tools, the forerunners of modern intelligence tests, to demon-strate that MR was a separate condition warranting more humane and"educational" treatments. In the case of MR, progress in definition andmeasurement was crucial for moving the field forward, much in thesame way that improved diagnostic standards for attention-deficit/hy-peractivity disorder moved forward the research and treatment of thatdisorder in the last two decades.

The LD field has stood at this crossroads for some time now(Stanovich, 2005) and has not moved forward with new ways of classify-ing the disorder. As has been the case with other disorders of childhood,a classification breakthrough portends much faster scientific progresstoward the goals of better prevention and intervention for children withLD. Classification is a central, not peripheral, issue. Any new approachto LD classification must be evaluated within the context of the extantLD diagnostic literature base to avoid being construed aspseudoscientific (Stanovich, 2005).

PREVALENCE AND ECONOMIC ISSUES

Recent estimates are that 2.9 million students receive services for a spe-cific LD in the U.S., which accounts for approximately 6.0% of total pub-lic school enrollment (National Center for Education Statistics, Statisticsof Public Elementary and Secondary School Systems, 2003; U.S. Depart-ment of Education, Office of Special Education and Rehabilitative Ser-vices, 2003). Between 1991-1992 and 2001-2002, there has been a 28.5%increase in the number of children being served through the Individualswith Disabilities Education Act (IDEA) for specific learning disabilities,and these children account for 50% of all children being served in specialeducation (U.S. Department of Education, Office of Special Educationand Rehabilitative Services, 2003). However, the number of studentsserved varies greatly by state and ethnicity. Some ethnicities, includingAmerican Indian/Alaska Native and Hispanic students, areoverrepresented (U.S. Department of Education, Office of SpecialEducation Programs, Data Analysis Systems, 2002).

The prevalence of LD also varies by country. A multinational study in-volving data from 21 countries revealed wide discrepancies in the prev-

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

366 DOMBROWSKI ET AL.

alence of LD (Organisation for Economic Co-operation and Develop-ment, 2004). This study found that Finland (9.56%) had the highestpercentage of children identified as having LD relative to all other coun-tries, including the U.S. (7.01%). Conversely, Luxembourg (0.53%) hadthe lowest percentage. In this study, significant effort was undertaken toincrease cross-national agreement on the definitions of learning disabil-ity categories; however, individual countries had varied definitions,which hampered cross-national comparisons. Regardless of nationalLD prevalence rate, there has been a consistent increase in the identifica-tion of LD (Lyon, 1996; Terman, Larner, Stevenson, & Behrman, 1996).This increase has also occurred at the local, state, and national levelwithin the U.S. (Lyon, 1996).

There are numerous proposed explanations for the increasing preva-lence of LD diagnosis. These include financial incentives to increase di-agnosis to qualify further for funding, social and political pressures toapply the less stigmatizing diagnosis of LD to children who may evi-dence mental retardation, increasing numbers of specialists who diag-nose and treat children with LD, and inadequate preparation of teachersto effectively handle individual differences in learning abilities withinthe general education classroom (Lyon, 1996; Hocutt, 1996). Additionalexplanations include more accurate epidemiological research; increasedrecognition of LD in girls; increased societal awareness of the character-istics of LD; and increased understanding of the potential negativeconsequences that mild difficulties can have on future learning (Lyon,1996).

The attempt to determine the true prevalence of LD is hindered by ad-ditional problems. A major point of contention has been the lack of a uni-fied, operational definition of LD that has led to misinterpretation andmisidentification of LD in the U.S. (Dombrowski et al., 2004). Ostensiblya national diagnostic algorithm, the LD discrepancy model has been in-consistently applied and is often ineffective (Aaron, 1997; Dombrowskiet al., 2004; Shepard, 1983; Overton, Fielding, & Simonsson, 2004). Forexample, Bocian, Beebe, McMillan, and Gresham (1999) found that only39% of students deemed eligible actually met criteria for a significantdiscrepancy between intelligence and achievement, and the decisionsfor eligibility were inconsistent. Instead, the eligibility decisions werebased on observations and children's educational need for assistance.The over-identification of children of specific ethnic groups, such asHispanic children, is another area of concern. Clinicians often ignore theexclusionary clause in the definition of LD, which prohibits categorizinga student as LD if the apparent difficulties are due to cultural or eco-nomic disadvantage (Fletcher & Navarrete, 2003). In fact, one survey ofprofessionals revealed that fewer than 50% of school psychologists reg-

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

THE SOLOMON EFFECT IN LD DIAGNOSIS 367

ularly considered the exclusionary clause in their diagnosis, and 37% in-dicated that they usually ignored or tried to get around the clause (Har-ris, Gray, Davis, Zaremba, & Argulewicz, 1997). Finally, greateravailability and specificity of services for children with disabilities is as-sociated with lower prevalence and more specific diagnosis of LD, sug-gesting that misdiagnosis may serve the purpose of providing somekind of service to children in need (Shepard, 1983). These inconsistentand idiosyncratic approaches to diagnosis, along with the varyingdefinitions of LD, impede the field's ability to determine true LDprevalence rates.

The true prevalence of learning disabilities has significant implica-tions not only for the validity of the diagnosis of LD, but also for the abil-ity to serve children with educational needs. Special education is costly.Expenditures for students with specific learning disabilities amountedto $10,558 per pupil in 1999-2000, which was 1.6 times the expenditurefor a regular education student (Chambers, Shkolnik, & Perez, 2003).Currently, special education is funded through federal, state, and localgovernments, with the bulk of responsibility falling on state, and to alesser degree, local authorities (Parrish & Chambers, 1996; Terman et al,1996). This funding paradigm is problematic. While the diagnosis of spe-cial education disabilities is permitted by federal law to be uncapped,the federal cost to underwrite services for children with disabilities iscapped. Consequently, local school districts bear the burden of a signifi-cant percentage of the cost of special education. Therefore, the presentway in which special education is financed places an undue burden onlocal school systems.

Furthermore, the cost of assessment to establish student eligibility forspecial education raises concerns. The cost of assessment, estimated at$800 to $8,000 per child (U.S. Department of Education, President'sCommission on Excellence in Special Education, 2005), consumes a largepercentage of the funding for special education. Because it is often unre-lated to treatment, further testing is still required for intervention plan-ning (Fletcher & Navarrete, 2003; Lyon, 1996; Reschly, 1996). If assess-ment practices were more efficient, a greater percentage of the fundingcould be used for interventions. The over-referral for special educationassessment poses additional cost burdens (Reschly, 1996). If a more effi-cient special education assessment process were established, using anappropriate pre-referral process, then fewer students would ultimatelybe referred for assessment.

Therefore, the heightened demand for LD services and high cost ofthose services raise questions regarding diagnosis, the process of deter-mining eligibility, and the effectiveness of treatments. Most studentswho are determined to be eligible will continue to receive services until

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

368 DOMBROWSKI ET AL.

they exit school, such that costs will continue to rise if more students areadmitted into than exited from LD services (Terman et al., 1996).

AN OVERLOOKED LEGAL CONSIDERATION

It is important to consider the legal implications of any diagnostic ap-proach for LD. In addition to the IDEIA, the Americans with DisabilitiesAct (ADA; United States Department of Justice, 1996) has the capacity toshape LD diagnosis. Some of its fundamental tenets must be considered.The ADA has already begun to exert its influence on the field of learningdisabilities at the collegiate level, and its language is broad enough alsoto apply to children enrolled in primary and secondary public schools(Siegel, 1999). The ADA applies to anyone, either documented or re-garded as, having a physical or mental impairment that substantiallylimits one or more major life activities. Based on this definition, it can beargued that the ADA provides protection for those diagnosed with LD.

Increasingly, both public and private institutions providing instruc-tion or examination services are being held accountable for any per-ceived discrimination toward individuals with a documented LD. ADAcases such as Guckenberger v. Boston University demonstrate that the legalconfusion surrounding learning disabilities must be a concern when es-tablishing new diagnostic criteria (Siegel, 1999; Wolinsky & Whelan,1999). In the Guckenberger case, the plaintiffs, Guckenberger et al., andthe defendant, Boston University, disagreed about who can be appropri-ately diagnosed with an LD (Siegel, 1999). This case highlighted the exis-tence of ambiguous LD diagnostic guidelines and provided for the pos-sibility that the judicial system, through the lens of ADA, will determinethe diagnostic standard at the collegiate level. As a consequence, thisopens the possibility that the rule of ADA law could extend to primaryand secondary public school settings and affect diagnostic practice if apoorly defined LD diagnostic definition persists.

The ADA, like the IDEIA, offers a number of protections. First, it pre-vents discrimination against a person with a disability who wishes toparticipate in a service, program, or activity. The ADA also requires theprovision of programs and services in an integrated setting, unless sepa-rate or different measures are necessary to ensure equal opportunity.Third, the ADA eliminates unnecessary eligibility standards or rulesthat deny individuals with disabilities an equal opportunity to enjoytheir services, programs, or activities unless these standards are clearlynecessary. Finally, the ADA requires reasonable modifications in poli-cies, practices, and procedures that deny equal access to individualswith disabilities, unless a fundamental alteration in the program wouldresult. The "reasonable modifications" include the controversial accom-

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

THE SOLOMON EFFECT IN LD DIAGNOSIS 369

modations that receive much attention in the LD literature, such asaccess to textbooks on tape, extended time for exams, help of a notetaker,and oral examinations.

It is necessary to clearly define LD so that the field rather than thecourts will determine how the construct should be defined and diag-nosed. Further, a clear definition will serve to curb costly future litiga-tion expenses that arise when a definition and diagnostic standard is un-clear and open to interpretation from multiple perspectives.

PSYCHOLOGICAL MISMEASUREMENTAND ETHICAL STANDARDS

Measurement and ethical standards must be considered prior to adopt-ing a new LD diagnostic model. An important responsibility of a practi-tioner is to determine the validity and reliability of any assessmentmethod based on the Standards for Educational and Psychological Test-ing (1999) developed jointly by the American Educational Research As-sociation (AERA), the American Psychological Association (APA), andthe National Council on Measurement in Education (NCME) (AERA,APA, & NCME, 1999). Certain standards pertain specifically to learningdisabilities and responsible test use. For example, Standards 4.19 and4.20 recommend caution when interpretation involves using one ormore cut scores and requires that empirical evidence of the cut score'svalidity be provided. Standard 13.7 recommends that multiple sourcesof information be considered when making a placement decision thatwill have a major impact on the student.

Responsible test use, however, is more sophisticated than implied bythe self-evident position of the aforementioned standards. SamuelMessick (1989), whose ideas have subsequently been incorporated intothe present standards, argued that test use and interpretation shouldrely not only on its evidential basis and construct validity, but also mustinclude a focus on its consequential basis. For instance, what are the so-cial consequences of test use and are they detrimental to a particulargroup? Has the examiner adequately separated value judgments fromempirical evidence when making the diagnosis of LD? In fact, the mostrecent test standards have adopted the concept of consequential validitywith advice to investigate the source of any such unintended conse-quences (AERA et al., 1999). This is very important when the conse-quence of such a decision may be to deny a child of obviously neededservices while another "less needy" child may receive those services.

Research on LD has shown that many diagnoses of the disorder do notmeet current accepted standards for the diagnosis (Shepard, Smith, &Vojir, 1983) thus making it quite possible for services to be allocated in-

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

370 DOMBROWSKI ET AL.

accurately, an obviously problematic "consequence" of the LD diagnos-tic process. For example, the classic large-scale study by Shepard et al.(1983), found that many children classified as LD using the Coloradospecial education dataset did not meet any acceptable definition of LD,discrepancy model, or otherwise. Of greater concern was the findingthat obvious rule outs such as hearing impairment were also not at-tended to in the classification process. Of interest to psychologists wasthe finding that most school district personnel opted to use tests of lesserreliability and validity when given the option. Clearly, clinicians and re-searchers must take into account the consequential validity of theoutcomes when establishing a new diagnostic methodology.

CONCLUSION

Prior to adopting an LD diagnostic model, the issues discussed in this ar-ticle should be considered in order to ensure greater longevity for anynew diagnostic method. Moreover, the LD diagnostic approach that isultimately adopted should be revisited every 7 to 10 years. LD diagnosisrepresents greater than 50% of all special education diagnoses (U.S. De-partment of Education, Office of Educational Research and Assessment,1996), so the establishment of an appropriate, uniform, and objectivelydefined diagnostic algorithm is critically important. Without an objec-tively defined LD diagnostic algorithm, LD diagnosis may continue tobe idiosyncratic and haphazard. As a result, school districts may face po-tential increases in due process hearings and litigation. Of greater im-portance, leaving the approach to LD diagnosis open-ended (as ispresently codified in IDEIA, 2004) might have grave consequences forchildren. School districts and various state departments of education arestrongly encouraged to consider the following five points when estab-lishing a diagnostic definition of learning disabilities: (1) the definitionneeds to be unambiguous; (2) it must be universally accepted across pro-fessions, researchers, and governmental entities; (3) it must incorporateclearly defined subtypes of learning disabilities; (4) it must be empiri-cally supported; and (5) it must point to valid, reliable, and cost-effec-tive procedures for the identification of children with and withoutlearning disabilities. While more research is always desirable, there isample evidence of numerous missteps and blind alleys taken by the LDfield in the past, and there is no need to revisit them a la the "SolomonEffect" as we move forward.

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

THE SOLOMON EFFECT IN LD DIAGNOSIS 371

ReferencesAaron, P. G. (1997). The impending demise of the discrepancy formula. Review of Educa-

tional Research, 67(4), 461-502.Adelman, H. S. (1992). LD: The next 25 years. Journal of Learning Disabilities, 25,17-22.Adelman, H. S., & Taylor, L. (1986). The problems of definition and differentiation and the

need for a classification schema. Journal of Learning Disabilities, 19,514-520.American Educational Research Association (AERA), American Psychological Associa-

tion (APA), & National Council on Measurement in Education (NCME). (1999).Standards for educational and psychological testing. Washington, DC: AmericanEducational Research Association.

American Psychiatric Association (APA). (1994). Diagnostic and statistical manual of mentaldisorders (4th ed.). Washington, DC: Author.

Bateman, B. (1965). Learning disabilities: An overview. Journal of School Psychology, 3(3),1-12.

Blashfield, R. K. (1993). Models of classification as related to a taxonomy of learning dis-abilities. In G. R. Lyon, D. B. Gray, J. F. Kavanagh, & N. A. Krasnegor (Eds.), Betterunderstanding learning disabilities: New views from research and their implications for ed-ucation and public policies (pp. 17-26). Baltimore: Paul H. Brooks.

Blashfield, R. K., & Draguns, J. G. (1976). Toward a taxonomy of psychopathology: Thepurpose of psychiatric classification. British Journal of Psychiatry, 129,574-583.

Bocian, K.M., Beebe, M.E., McMillan, D., & Gresham, F.M. (1999). Competing paradigms inlearning disabilities classification by schools and the variations in the meaning ofdiscrepant achievement. Learning Disabilities Research & Practice, 14,1-14.

Chambers, J., Shkolnik, J., & Perez, M. (2003). Total expenditures for students with disabili-ties, 1999-2000: Spending variation by disability; Report of special education ex-penditure project (SEEP). American Institutes for Research in the Behavioral Sciences,Palo Alto, CA. Center for Special Education Finance. Retrieved December 8,2004, fromw w w . e r i c . e d . g o v / E R I C D o c s / d a t a / e r i c d o c s 2 / c o n -tent_storage_01/0000000b/80/23/be/49.pdf.

Dombrowski, S. C, Kamphaus, R. W., & Reynolds, C. R. (2004). After the demise of the dis-crepancy: Proposed learning disabilities diagnostic criteria. Professional Psychology:Research and Practice, 35(4), 364-372.

Fletcher, T., & Navarrete, L. (2003). Learning disabilities or difference: A critical look at theissues associated with the misidentification and placement of Hispanic students inspecial education programs. Rural Special Education Quarterly, 22(4), 37-46.

Harris, J.D., Gray, B.A., Davis, J.E., Zaremba, E.T., & Argulewicz, E.N. (1997). Theexclusionary clause and the disadvantaged: Do we try to comply with the law? Jour-nal of Learning Disabilities, 21,581-583.

Hinshelwood, J. (1917). Congenital word blindness. London: HK Lewis.Hocutt, A. (1996). Effectiveness of special education: Is placement the critical factor? The

Future of Children, 6(1), 77-102.Individuals with Disabilities Education Improvement Act (2004). Retrieved August 31,

2005 from www.ed.gov/about/offices/list/osers/osep/index.htmlKamphaus, R. W. (2001). Clinical assessment of child and adolescent intelligence (2nd ed.).

Boston: Allyn & Bacon.Keogh, B. K. (1983). Classification, compliance, and confusion. Journal of Learning Disabili-

ties, 16,25.Kirk, S. A. (1962). Educating exceptional children. Boston: Houghton Mifflin.Kirk, S. A., & Bateman, B. (1962). Diagnosis and remediation of learning disabilities. Excep-

tional Children, 29(2), 73-78.Lyon, G.R. (1996). Learning disabilities. The Future of Children, 6(1), 54-76.

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

372 DOMBROWSKI ET AL.

Lyon, G. R., Fletcher, J. M., & Barnes, M. C. (2003). Learning disabilities. In EJ. Mash & R.A.Barkley (Eds.), Child psychopathology (2nd ed.). New York: TheGuilford Press.

Mercer, C. D., Forgnone, C, & Wolking, W. D. (1976). Definitions of learning disabilitiesused in the United States. Journal of Learning Disabilities, 9, 376-386.

Messick, S. (1989). Validity. In R. L. Linn (Ed.), Educational Measurement. (3rd ed., pp.13-103). New York: Macmillan Publishing Company.

National Center for Education Statistics, Statistics of Public Elementary and SecondarySchool Systems, 2003 (data 1989-90 through 2001-02). Retrieved December 8,2004,from http://nces.ed.gov/programs/digest/d03/tables/dt052.asp.

Organisation for Economic Co-operation and Development. (2004). Equity in education:Students with learning disabilities, learning difficulties, and disadvantages, statis-tics and indication. Danvers, MA: OECD.

Orton, S. (1925). Word blindness in school children. Archives of Neurology and Psychiatry, 14,581-613.

Overton, T., Fielding, C, & Simonsson, M. (2004). Decision making in determining eligibil-ity of culturally and linguistically diverse learners: Reasons given by assessmentpersonnel. Journal of Learning Disabilities, 37(4), 319-330.

Parrish, T., & Chambers, J. (1996). Financing special education. The Future of Children, 6(1),121-138.

Reschly, D. (1996). Identification and assessment of students with disabilities. The Future ofChildren, 6(1), 40-53.

Rutter, M., & Yule, W. (1975). The concept of specific reading retardation. Journal of ChildPsychology and Psychiatry, 16,181-197.

Shepard, L. (1983). The role of measurement in educational policy: Lessons from the identi-fication of learning disabilities. Educational Measurement: Issues and Practice, 2,4-8.

Shepard, L. (1989). Identification of mild handicaps. In R. L. Linn (Ed.) Educational measure-ment (3rd ed, pp. 545-572). New York: Macmillan Publishing Company.

Shepard, L., Smith, M., & Vojir, C. (1983). Characteristics of pupils identified as learningdisabled. American Educational Research Journal, 20,309-331.

Siegel, L. S. (1999). Issues in the definition and diagnosis of learning disabilities: A perspec-tive on Guckenberger v. Boston University. Journal of Learning Disabilities, 32(4),304-319.

Stanovich, K. E. (1986). Matthew Effects in reading: Some consequences of individual dif-ferences in the acquisition of literacy. Reading Research Quarterly, 21, 360-407.

Stanovich, K. E. (2005). The future of a mistake: Will discrepancy measurement continue tomake the learning disabilities field a pseudoscience? Learning Disability Quarterly,28(2), 103-106.

Terman, D., Larner, M., Stevenson, C, & Behrman, R. (1996). Special education for studentswith disabilities: Analysis and recommendations. The Future of Children, 6(1), 4-24.

Tucker, J., Stevens, L. J., & Ysseldyke, J. E. (1983). Learning disabilities: The experts speakout. Journal of Learning Disabilities, 16, 6-14.

U.S. Department of Education, Office of Educational Research and Assessment. (1996). Di-gest of education statistics. Washington, DC: U.S. Government Printing Office.

U.S. Department of Education, Office of Special Education and Rehabilitative Services.(2003). 24th Annual Report to Congress on the Implementation of The Individualswith Disabilities Education Act (data 1976-77 through 2001-02). Retrieved Decem-ber 8,2004, from http://nces.ed.gov/programs/digest/d03/tables/dt052.asp.

U.S. Department of Education, Office of Special Education Programs, Data Analysis Sys-tem (DANS). (2002). Student characteristics (table data 1991-92 through 2000-01).

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

THE SOLOMON EFFECT IN LD DIAGNOSIS 373

Retrieved December 8, 2004, from www.ed.gov/about/reports/annual/osep / 2002 / section-ii.pdf.

U.S. Department of Education, President's Commission on Excellence in Special Educa-tion (2005). A new era: Revitalizing special education for children and their families. Re-trieved September 10, 2005, from www.ed.gov/inits/commissionsboards/whspecialeducation / reports / images /Pres_Rep.pdf.

United States Department of Justice. (1996, July 26). Americans with Disabilities Act of 1990.Retrieved December 3,2004, from www.usdoj.gov/ crt/ada/pubs/ada.txt.

U.S. Office of Special Education Programs, National Joint Committee on Learning Disabili-ties. (2002, July 25). Specific learning disabilities: Finding common ground. Re-trieved September 11,2005, from www.ldonline.org/news/commonground.html.

van der Wissel, A., & Zegers, F. E. (1985). Reading retardation revisited. British Journal ofDevelopmental Psychology, 3,3-9.

Wiggins, O. P., & Schwartz, M. A. (1994). The limits of psychiatric knowledge and the prob-lems of classification. In J. Z. Sadler, O. P. Wiggins, & M. A. Schwartz (Eds.), Philo-sophical perspectives on psychiatric diagnostic classification (pp. 89-103). Baltimore:Johns Hopkins Press.

Wolinsky, S., & Whelan, A. (1999). Federal law and the accommodation of students withLD: The lawyer's look at the BU decision. Journal of Learning Disabilities, 32,286-291.

Yule, W. (1973). Differential prognosis of reading backwardness and specific reading retar-dation. British Journal of Educational Psychology, 43,244-248.

Action Editor:

Stefan C. Dombrowski, Ph.D., is an Associate Professor and coordina-tor of the School Psychology program at Rider University inLawrenceville, NJ. His research interests focus on the impact of perinatalfactors on later educational and behavioral outcomes. He also conductsresearch on assessment-related issues as they relate to the developmen-tal well-being of children. Dr. Dombrowski lives in Cherry Hill, NJ withhis wife and son.

Randy W. Kamphaus, Ph.D., is a Distinguished Research Professor andHead of the Department of Educational Psychology and InstructionalTechnology at the University of Georgia. A focus on issues related toclinical assessment has led Dr. Kamphaus to pursue research in classifi-cation methods, differential diagnosis, test development and learningdisability assessment. He has served as a principal investigator on feder-ally funded research projects dealing with early intervention and pre-vention of behavior problems, classification methods, and aggressionreduction. He is a Fellow of the APA and Past-President of the Divisionof School Psychology for APA. Dr. Kamphaus has authored orco-authored 12 books, 4 psychological tests, and more than 70 scientific

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.

374 DOMBROWSKI ET AL.

journal articles and chapters. The Ph.D. School Psychology Program atthe University of Georgia is accredited by the APA.

Melissa Barry, Amber Brueggeman, Sarah Cavanagh, LindaHekimoglu, and Sarah Vess are pursuing their Ph.D. in school psychol-ogy at the University of Georgia. Katie Devine is pursuing her Ph.D. inclinical psychology at the University of Georgia. The Ph.D. School Psy-chology Program at the University of Georgia is accredited by the APA.

This

doc

umen

t is c

opyr

ight

ed b

y th

e A

mer

ican

Psy

chol

ogic

al A

ssoc

iatio

n or

one

of i

ts a

llied

pub

lishe

rs.

This

arti

cle

is in

tend

ed so

lely

for t

he p

erso

nal u

se o

f the

indi

vidu

al u

ser a

nd is

not

to b

e di

ssem

inat

ed b

road

ly.