Chapter 5
The Relationship of Intelligence and Adaptive Behavior

Determining whether a person has mental retardation involves complex decisions that integrate information on current intellectual functioning and adaptive behavior. Information about each of these core dimensions is always incomplete and dependent on imperfect measures of the underlying constructs. Judgment is therefore necessary when making decisions about how best to assess intellectual and adaptive functioning and in interpreting the results; this chapter provides guidance for those judgments. However, the guidance cannot take the form of absolute decision rules that replace judgment about the appropriateness and meaning of evaluation results. For this reason, high standards and much preparation are needed for the profes-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits Chapter 5 The Relationship of Intelligence and Adaptive Behavior Determining whether a person has mental retardation involves complex decisions that integrate information on current intellectual functioning and adaptive behavior. Information about each of these core dimensions is always incomplete and dependent on imperfect measures of the underlying constructs. Judgment is therefore necessary when making decisions about how best to assess intellectual and adaptive functioning and in interpreting the results; this chapter provides guidance for those judgments. However, the guidance cannot take the form of absolute decision rules that replace judgment about the appropriateness and meaning of evaluation results. For this reason, high standards and much preparation are needed for the profes-

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits sional personnel making diagnostic decisions, including thorough knowledge of mental retardation as a diagnostic construct. This chapter discusses the diagnostic implications of the preceding chapters on intellectual functioning and adaptive behavior, as well as a review of the literature on the relationship between measures of intellectual functioning and measures of adaptive behavior. That fundamental relationship has significant implications for the discussion of how diagnostic decisions are made by combining information across multiple domains of functioning, from multiple sources, and from multiple methods of gathering information. A principle of convergent validity will emerge in this discussion as critical to a sound diagnosis of mental retardation. DIAGNOSTIC CONSTRUCT OF MENTAL RETARDATION Diagnostic constructs have two key components: conceptual definitions and classification criteria. Both are critical to understanding the meaning of the diagnostic construct. The four conceptual definitions of mental retardation discussed in this report (see Chapter 1) do not suggest explicit classification criteria. Although there is controversy regarding some features of these diagnostic systems (MacMillan et al., 1993, 1995; Reiss, 1994), the conceptual definitions differ little among the diverse organizations involved. It should also be noted that many other organizations and agencies establish conceptual definitions and classification criteria for mental retardation, including the Social Security Administration (SSA) and state departments of education. Broad consensus exists throughout the developed world about the basic features of the conceptual definition of mental retardation: it involves significant limitations in the core dimensions of intellectual functioning and adaptive behavior. Most national and worldwide diagnostic systems use the term “mental retardation,” and nearly all suggest that deficits in adaptive behavior arise because of limited intellectual functioning. The greatest variations in conceptual definitions and terminology occur in the legal requirements for classification of stu-

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits dents as mentally retarded in the special education system of each state’s department of education (Denning et al., 2000; Patrick & Reschly, 1982; Utley et al., 1987). Across the 50 states and the District of Columbia, different terminology is used (e.g., mental retardation, mental disability, significantly limited intellectual capacity), along with widely varying classification criteria. Nearly all states, however, define a disability based on deficits in the dimensions of intellectual and adaptive functioning. Controversies regarding mental retardation diagnostic systems arise most often regarding classification criteria, that is, how the conceptual definition of mental retardation is operationalized. Classification criteria vary significantly regarding the cutoff scores that are adopted to determine which cases meet or do not meet diagnostic eligibility criteria. Higher cutoff scores, of course, increase the population with a diagnosis of mental retardation, and lower cutoff scores decrease it. A little-appreciated influence is the joint effect of IQ and adaptive behavior cutoff scores on diagnostic decisions. As is shown later, the combined effects of different cutoff scores can drastically alter the number of people who can be considered for a diagnosis of mental retardation. Classification criteria also vary regarding the use of composite and part scores as well as the number of part scores that may be used from measures of adaptive and intellectual functioning. Classification Criteria for Intellectual Functioning The cutoff scores for measures of general intellectual functioning are better established than the cutoff scores for measures of adaptive behavior. There is broad consensus in the major diagnostic systems that performance on the intellectual dimension must be approximately two or more standard deviations below the population mean, which translates into an IQ score of 70 or less on measures with a mean of 100 and a standard deviation (SD) of 15. The degree of flexibility around the cutoff score of 70 varies among diagnostic systems; some allow a range bounded by one standard error of measurement, which

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits TABLE 5-1 Proportions of People with Scores At and Below Different Cutoff Scores on a Normally Distributed Characteristic Cutoff Score Percentage Meeting or Exceeding Score Below 70 2.28 70 and below 2.68 Below 75 4.75 75 and below 5.48 Below 80 9.18 80 and below 10.20 Below 85 15.87 85 and below 17.62 translates to about 4 points for measures with reliabilities above r = .90. Other systems make a general statement that IQ can be approximately 70 to 75. The difference between a cutoff score at or below IQ ≤ 75 and a cutoff score at or below IQ ≤ 70 is dramatic, as shown in Table 5-1. Twice the proportion of people have scores at or below 75 (5.48 percent) than have scores at or below 70 (2.68 percent). In other words, a seemingly trivial change of five points on the intellectual dimension doubles the number of people from the given population that are potentially eligible for consideration on that dimension. Before raising an alarm with this statistical information, however, it should be noted that no prevalence study of people identified as having mental retardation has ever approached the level of 5 percent of the general population, at least in part because of the necessity of a concurrent deficit in adaptive behavior. More commonly, investigations have yielded a prevalence of 1 to 1.5 percent. The broad consensus that exists on the classification criteria for the intellectual dimension does not exist in the special education rules adopted by the states. Current state criteria on a cutoff IQ score for the intellectual dimension vary from a low of about 69 to a high of 80

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits (Denning et al., 2000). Furthermore, Patrick and Reschly (1982) found that the stringency of the IQ criterion is not always related to the prevalence of students classified as having mental retardation and placed in special education. State-to-state variations in special education rules regarding conceptual definitions and classification criteria for mental retardation lead to large differences in prevalence and many inconsistencies in the diagnosis of mental retardation between such agencies as SSA and the public schools. The use of school data is discussed later in this chapter. Classification Criteria for Adaptive Behavior In most diagnostic systems, the classification criteria for adaptive behavior are not developed as well or as clearly as those for intellectual functioning. Two elements are particularly relevant: the degree of difference from normal or average performance that is required to determine that a limitation in adaptive functioning exists—that is, the cutoff score—and the number of domains or areas in which limitations may be observed. Each of these elements has a significant influence on the number of people who might be considered for a diagnosis of mental retardation. Cutoff Scores As noted earlier, there is far less agreement on the appropriate cutoff score(s) for adaptive behavior measures than there is for measures of intellectual functioning. Precise cutoff scores generally have not been specified in diagnostic systems, primarily because of the lack of confidence in adaptive behavior measures and the availability of multiple instruments that may be used interchangeably or somewhat idiosyncratically. A selection of quotations suggests the wide range of views: “If an adequate standardized instrument were available for the measurement of adaptive behavior, the upper limit of Level - I could presumably be set, as with the Measured Intelligence dimension, at greater

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits than minus one Standard Deviation from the population mean” (Heber, 1961, p. 61). “If more precise instruments were available for the measurement of Adaptive Behavior, and general norms could be precisely stipulated, the upper limit could presumably be set at minus two standard deviations from the population mean” (Grossman, 1973, p. 19). “It seems impractical at this time to suggest fine gradations that can be achieved with accuracy, and, in the final analysis, clinical judgment is needed to arrive at an estimate of adaptive behavior level. . . . Standardized scales, supplemented by clinical judgments whenever possible, should be applied in making diagnoses” (Grossman, 1983, p. 46). “Despite increased emphasis on adaptive skills in the definition, there has been virtually no support for the use of a single global score or age equivalent index to operationalize adaptive skill limitations. There are a number of reasons why a global score and precise cutoff point would not be productive” (American Association on Mental Retardation, 1992, p. 42). “The second criterion for diagnosing a person as having mental retardation is that the individual have limitations in two or more adaptive skills. This part of the diagnosis is more substantive and subjective and requires clinical judgment that takes into account environmental demands and potential support systems” (American Association on Mental Retardation, 1992, p. 49). The American Psychological Association Division 33 (Editorial Board, 1996) mental retardation diagnostic system is the one exception to the general trend in diagnostic systems of avoiding precise specification of adaptive behavior cutoff scores to define mental retardation eligibility. The Division 33 scheme is explicit in recommending the use of a “comprehensive, individual measure of adaptive behavior” (p. 13) and in specifying precise cutoff scores: For adaptive behavior measures, the criterion of significance is a summary index score that is two or more standard deviations below the mean for the appropriate norming sample or that is within the range of adaptive behavior associated with the IQ range sample in instrument norms (Editorial Board, 1996, p. 13).

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits In its next paragraph, the Division 33 discussion allows for part scores, but the criterion for eligibility is that “two or more of these scores lie two or more standard deviations below the mean” (p. 13). The Division 33 system also adopted the now rather standard criterion of intellectual functioning at two or more standard deviations below the mean. The consequences of these requirements on the number of people currently considered for a diagnosis of mental retardation or on the number of persons considered in the future have not been addressed. The results of previous studies had suggested that a stringent criterion for adaptive behavior plus the usual criterion for intellectual functioning led to a sharply reduced number of people eligible to be considered for a mental retardation diagnosis (Heflinger et al., 1987; Reschly, 1981a). However, these studies used a particular measure of adaptive behavior that had a very low correlation with measures of intellectual functioning. The classification criteria governing diagnosis of mental retardation for special education services by state departments of education generally do not provide guidance regarding the use of adaptive behavior composites, part scores, or cutoff scores to determine eligibility. It is not surprising that the use of an adaptive functioning criterion in the schools is inconsistent and unpredictable (Reschly & Ward, 1991). Moreover, enormous variations exist across the states and, in some instances, across local school districts within states. Adaptive Behavior Domains Diagnostic systems are either silent on the appropriate number of adaptive behavior domains, or they adopt widely varying schemes. The most recent classification system of the American Association on Mental Retardation (AAMR) specifies 10 adaptive skills areas without any explanation of how that number was determined or why some domains were included and others excluded (American Association on Mental Retardation, 1992). In contrast, the American Psychological Association Division 33 diagnostic system is generally consistent with the factor analytic literature (see Chapter 4 and later discussion in this

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits chapter) in specifying the use of a composite score, recognizing that a general adaptive behavior factor exists, and the use of a limited number of part scores. The appropriate number of domains in the assessment of adaptive behavior depends on the instrument, age level, and other considerations. Some useful guidance on the number of meaningful domains is provided by factor analytic studies; however, different factor methods yield different results, so such studies are rarely definitive. Most adaptive behavior scales yield a general factor, regardless of the number of domain or subdomain scores, if the analytic method permits the emergence of such a factor (e.g., Harrison & Oakland, 2000a; McGrew & Bruininks, 1989). Typically, one or more group factors also emerge, particularly if (a) confirmatory factor analytic procedures are applied, (b) items reflect diverse areas of functioning, (c) sufficient floors and ceilings are provided, (d) broad age ranges are included, and (e) individuals from the moderate and mild levels of mental retardation, as well as people with borderline and normal levels of functioning, are included in the sample. Across all ages, McGrew and Bruininks suggested the possibility of four or five group factors. This literature is discussed in detail in Chapter 4. Beyond its theoretical importance, the appropriate number of adaptive behavior domains has a very practical significance: it can have a tremendous influence on the number of people who may be diagnosed as having mental retardation and therefore are eligible for Supplemental Security Income (SSI) and Disability Income (DI) benefits. At the request of the committee, Thompson (2001) ran a series of Monte Carlo simulations to address this effect. She found that the number of adaptive behavior domains on which deficits must be shown had a marked effect on identification rates, with more individuals being identified as having mental retardation if only a single adaptive behavior domain had to meet a defined cutoff score than if two or more domains had to meet a cutoff score. Furthermore, the number of domains on which deficits could be measured had a modest but significant effect on identification rates: more individuals will be diag-

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits nosed as having mental retardation if deficits can be found in 1 or 2 out of 9 or 10 domains, than if deficits are found in 1 or 2 of only 4 domains of adaptive behavior. In other words, it is easier to qualify for a diagnosis of mental retardation if there are more domains in which deficits can be shown. It is important to remember that expectations about adaptive behavior and competence vary by sociocultural group, settings, and age levels. The conceptual definition and descriptions of adaptive behavior in the 1983 AAMR manual (Grossman, 1983) have been particularly instructive in this regard because in it different competencies were associated with broad age ranges (e.g., preschool, childhood, adolescence, adult). The committee considered the interaction of age-based expectations and adaptive behavior domains, including the current domains identified in the SSA listings (see Table 5-2). Clearly, the current SSA scheme recognizes different domains at different age levels and is similar in most respects to the adaptive domains discussed in Chapter 4. In other respects, however, the SSA domains are inconsistent with findings from factor analytic results, the descriptions of adaptive behavior in authoritative sources, and the content of current adaptive behavior inventories. This led to the committee’s recommendation, presented in Chapter 4, for revising the SSA adaptive behavior domains (see right side of Table 5-2). The SSA domain of concentration, persistence, and pace is not assessed by most adaptive behavior instruments, although these skills could be part of the work attitudes and skills domain recommended for adults. The current SSA domains do not include some that are prominent in current conceptions and measures of adaptive behaviors, particularly the self-help and communication domains in the preschool years. For childhood and adolescence, the SSA scheme does not include motor/mobility, communication/functional academics, or daily living skills, although the latter may be covered by SSA in the domain of personal functioning. And, the communication/functional academics and work attitudes and skills domains are missing from the SSA adult domains. The adaptive behavior areas specified in Table 5-2 are

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits TABLE 5-2 Adaptive Behavior Domains in Current Social Security Administration Regulations and Committee Recommendations Age Social Security Administration Domains Age Committee Recommendations Birth-2 1. Motor (fine/gross) 2. Social Birth-4 1. Motor 2. Social 3. Self-help 4. Communication 3-17 1. Personal 2. Social 3. Concentration/persistence/pace 5-17 1. Motor/mobility 2. Social 3. Communication/functional academics 4 Daily living skills 18+ 1. Daily living 2. Social 3. Concentration/persistence/pace 18+ 1. Motor/mobility 2. Social 3. Communication/practical cognitive skills 4. Daily living skills 5. Work the ones recommended by the committee to be adopted by SSA to guide decisions about diagnoses of mental retardation (the recommendation itself appears in Chapter 4). Most current adaptive behavior measures have domains that are similar to the domains recommended by the committee; however, no scale is perfectly matched to these domains. Moreover, essential content, such as functional academic skills involving basic literacy, temporal relationships, and quantitative concepts, appears on most scales, but in different domains. For example, the area of functional academic skills is a separate domain in the Adaptive Behavior Assessment Scales (ABAS—Harrison & Oakland, 2000a) and the Comprehensive Test of Adaptive Behavior (CTAB—Adams, 2000), but it is spread over

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits at least two domains in the Vineland Adaptive Behavior Scales (VABS—Sparrow et al., 1984a) and the Scales of Independent Behavior-Revised (SIB-R—Bruininks et al., 1996). Valuable information from each of these instruments on the functional academic skills area is available, but direct translation of the available scores to a decision about performance in this area is difficult. Similar relationships exist between the available scores from instruments and the recommended areas in Table 5-2. Decisions about performance in each of these areas therefore need to be based on the results of adaptive behavior instruments, to the extent that one or more instruments are appropriate for a given client, and a broad variety of other information. RELATIONSHIP OF ADAPTIVE BEHAVIOR AND INTELLECTUAL FUNCTIONING The relationship between measures of adaptive behavior and intellectual functioning is highly variable and has multiple influences. Correlations between adaptive and intellectual functioning have varied in published studies from near zero (no relationship) to nearly 1.0 (perfect relationship). The strength of this relationship is important because it influences diagnostic decisions significantly. Variables That Influence Correlations A comprehensive review by Meyers et al. (1979) summarized data on the correlations of measures of adaptive and intellectual functioning. This section is informed by that review as well as by data published in test manuals over the past 20 years. Newer adaptive behavior scales generally conform to the generalizations made by Meyers et al. in 1979. Scale Content A major influence on the relationship of adaptive and intellectual functioning is the content of the measures, particularly the adaptive

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits ior in one or more settings, review of records reflecting previous and current performance, and interviews with the client and knowledgeable others. Disability determination examiners should be familiar with and utilize each of the methods in developing a well-informed perspective on the functioning levels of clients considered for a mental retardation diagnosis. Direct testing of clients’ adaptive behavior is a rarely used technique, even though direct testing of general intellectual functioning is the norm. Yet, at least some direct testing is appropriate with adaptive functioning. For example, functional academic skills, such as basic literacy, understanding temporal relationships, and quantitative concepts, are crucial to adaptive functioning for children, adolescents, and adults. Adults who cannot tell time or meet time-related work obligations are at a significant disadvantage in coping with everyday demands. Information from third-party respondents on these skills may or may not be accurate, especially as these skills relate to everyday functioning. Some adaptive behavior measures suggest establishing conditions under which behaviors can be “tested” if third-party respondents are unable to report their actual observations of the behavior (Adams, 2000). Interviews with third-party respondents by using standardized adaptive inventories is the most common method for collecting adaptive behavior information. For children, the third party is most often a parent or a teacher. While third-party interview is not used extensively in intellectual assessment, the results of IQ tests should be further evaluated through interviews with the client and significant others to determine if the observed test performance is consistent with day-to-day functioning. In addition to third-party respondents, interviews with clients and other parties are components of a comprehensive adaptive behavior assessment. One adaptive behavior instrument has been normed with adults using a self-report format; however, no data were provided in the manual contrasting the self-report and third-party respondent results for persons with mild mental retardation (Harrison & Oakland, 2000a). The accuracy of individuals with mild mental retardation in reporting their own adaptive behavior on this instru-

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits ment is not yet clear; nevertheless, unstructured and structured interviews with clients are necessary for determining if adaptive behavior deficits exist. Observations of clients’ intellectual functioning and adaptive behaviors can be systematic and structured or informal and anecdotal. Both techniques yield valuable information for understanding overall functioning. Opportunities for systematic observation of clients’ adaptive behaviors are limited by resource constraints and the near impossibility of conducting highly structured observations in all relevant settings and at the times that are appropriate. Moreover, many adaptive behaviors that are crucial to adequate functioning do not occur frequently, making systematic observation even more difficult. Therefore, even informal and anecdotal observations from different people and across different settings are valuable to an overall decision about adaptive behavior and should be obtained to the extent feasible. Review of records is another data collection method with strong applicability to the determination of intellectual and adaptive behavior deficits. School records are especially useful if evaluators understand the nature of mental retardation, classification practices in schools, and subtle indicators of low functioning in classrooms and schools. However, school records indicating either a diagnosis of mental retardation or the absence of one cannot be used as a definitive indication of intellectual and adaptive behavior status. Records from agencies other than schools can also be useful in determining adaptive behavior deficits. Medical, social service, and legal sources may yield further information that is useful in making judgments about deficits. Further discussion of the use of records from schools and other agencies in order to make diagnostic decisions appears below. Sources of Information Judgments about intellectual and adaptive functioning should be based on multiple sources of information including, at a minimum, the individual client and significant others such as (depending on age) parents, teachers, peers, neighbors, and family members. The kind and

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits amount of information gathered from different people will vary significantly across clients. In some cases, judgments must be based primarily on an interview with a single third-party respondent and on observations or interviews with the client, while in other cases there will be multiple sources of information. SSA disability examiners have to make judgments about the sufficiency of the information in deciding whether to actively seek additional sources of information. Settings The client’s functioning across different settings is also relevant to decisions about intellectual and adaptive behavior deficits. The settings that are most relevant depend on the client’s developmental level. For preschoolers, the relevant setting is the home and, depending on the client, day care or preschool settings. For children between ages of about 5 and 18, the school and home settings are crucial for nearly all clients, as are skills in meeting expectations as they age for roles in the neighborhood and the community. Deficits that are apparent only in a single setting generally should not be the basis for a determination of an adaptive behavior deficit. Diligence in collecting and examining information from multiple settings is very important. The literature refers to so-called six-hour retarded children (President’s Committee on Mental Retardation, 1970), described as having performance deficits only in school settings and coping adequately in home and community settings. “Six-hour retarded children” were assumed by many to blend into the normal adult population without significant adaptive limitations. Studies of young adults who clearly met this conception yielded a very different picture (Koegel & Edgerton, 1984). Contrary to the assumptions, as young adults, these children had enormous difficulties in coping with everyday demands and avoiding being exploited by others due to their functional limitations in practical cognitive and other adaptive skills. Current classification and placement practices make it less likely that children will be identified in this way in schools (see later discussion), but many of them do have significant problems coping with the

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits everyday demands involving social relations with peers and negotiating the cognitive demands required for personal and social adaptation outside school. Careful analysis of how children and adolescents perform in school, home, and other settings must be made in order to come to the most accurate diagnosis. Examiner Qualifications Individuals making mental retardation diagnoses must meet high standards regarding professional preparation and relevant experience. SSA disability examiners should have in-depth and up-to-date knowledge in the following areas: mental retardation theory, research, treatment, and best practices; mental retardation diagnostic construct; measurement of intellectual and adaptive functioning; assessment principles and best practices; mental disorders theory, research, and best practices; purposes and practices of multiple agencies, such as schools, law enforcement, and health care; knowledge of human development; and assessment of the individual’s strengths and limitations in the context of multiple environments, including family, work, and community. Convergent Validity Convergent validity is an application of the concept of the multitrait/multimethod examination of the validity of measures of psychological constructs (Campbell & Fiske, 1959). The committee recommends the principle of convergent validity as a means for SSA examiners to make sense of all the information evaluated for diagnostic decisions about mental retardation. In clinical practice, information is collected and evaluated over broad domains of functioning, using multiple methods of gathering data, multiple sources for that information, and multiple settings (e.g., Gresham, 1991). If the information is generally consistent with a particular diagnostic decision, such as mental retardation, that decision is made based on this confirmation. If, however, there are several elements in the overall array of information that

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits are inconsistent with such a diagnosis, the decision is not confirmed or—and this is important—further investigation is undertaken to explain discrepancies. Inconsistent Information Many individuals with legitimate diagnoses of mental retardation will present clinically with one or more elements of information that are inconsistent with the diagnosis. Many times, on further examination, the inconsistent information will derive from a mistake or distortion of measures of adaptive or intellectual functioning—for example, the existence of high scores from an adaptive behavior inventory with little or no ceiling for a young adult. In other cases, a particular respondent or performance in a specific setting may not be consistent with the diagnosis of mental retardation. Inconsistent information must be investigated thoroughly in order to avoid the harmful consequences of false positive or false negative decisions. Further investigation may take many different forms, including more extensive interviews with the client and significant others, additional assessments of adaptive or intellectual functioning, or follow-up contacts with personnel in other agencies who may or may not have made a diagnosis of mental retardation. Information from Other Agencies As noted previously in this report, mental retardation diagnoses are made for many purposes by many different agencies, including schools, law enforcement, and health care agencies. Different agencies use different diagnostic systems and classification criteria, making diagnostic disagreements among agencies perfectly legitimate. Moreover, the standards used by some agencies, public schools in particular, differ markedly from state to state and, occasionally, from district to district within states. Several sources of information confirm the increasing reluctance of school officials to make a diagnosis of mental retardation. First, the

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits TABLE 5-5 Changes in Prevalence of Learning Disabilities and Mental Retardation over the 1976-1999 Period in Public School Special Education Programs   1976-77 1998-99 Change % Change Learning disabilities 797,213 2,861,333 2,064,120 260% Mental retardation 969,547 613,207 –356,340 –36% NOTE: All data are from the 1978 and 2000 Annual Reports to Congress by the Office of Special Education Programs, U.S. Department of Education. prevalence of mental retardation in schools has declined substantially over the last 25 years. It is extremely unlikely that all of the decline is attributable to a truly lower prevalence of mild mental retardation. The decline in mild mental retardation is paralleled by a corresponding and substantially greater increase in learning disabilities (see Table 5-5). According to the Office of Special Education Programs child count data (U.S. Department of Education, 2000), prevalence of learning disabilities has increased by 260 percent while mental retardation prevalence has declined by 37 percent since the 1976-1977 academic year, when these data were first collected. These changes are even more impressive because they occurred during a period when children and youth with moderate, severe, and profound mental retardation gained access to the public schools for the first time in many states and districts. Although the child count data do not differentiate levels of mental retardation, it is highly likely that the decline in mild mental retardation has been even greater than the overall decline in mental retardation, simply because those with more severe mental retardation are more obviously impaired in many areas and are therefore more likely to be correctly diagnosed with mental retardation. The declining prevalence of mental retardation in the public schools is even more complex because it varies significantly across the states. The mental retardation prevalence among states varied by a

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits factor of three in the 1998-1999 data; that is, the highest prevalence reported by any state was three times higher than the lowest state-reported prevalence. Moreover, 28 states reported public school prevalence of less than 1 percent of the student population, meaning that some students with mild mental retardation are either not being placed in special education or are in special education because of other diagnoses, such as learning disability. Studies in California confirm both the reluctance of school professionals to confer the diagnosis of mental retardation and the willingness to use other diagnoses for children with characteristics that meet mental retardation classification criteria (MacMillan, Gresham et al., 1996). This leads to concerns about the continued viability of the mild mental retardation diagnostic construct in special education and schools (MacMillan & Reschly, 1996; MacMillan, Siperstein, & Gresham, 1996). The California studies indicate that staffing teams simply refused to diagnose students as having mild mental retardation even when IQ, achievement, and adaptive behavior data clearly pointed to that diagnosis. The degree to which these results generalize to other states is unknown, although the large decline in children diagnosed as having mild mental retardation suggests that the reluctance among California school psychologists and special educators to use this diagnosis may exist in many other places as well. The presence or absence of a diagnosis of mental retardation from another agency, especially from public schools, should neither confirm nor disconfirm an SSA diagnosis of mental retardation. Information from other agencies should be evaluated by SSA examiners, but should not be regarded as definitive. Although the official diagnoses used in other agencies are often not applicable to SSA eligibility determination, such information as direct measures of skills and records reflecting overall adjustment can be highly useful. Attempts should be made to obtain records from other agencies with that information interpreted by persons familiar with the functioning of the agency. Information from schools that is particularly relevant to mental retardation diagnoses includes measures of skills such as standardized test results,

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits teacher-assigned grades, history of retention in grade, curriculum track pursued, and participation in special education programs, regardless of diagnosis. CONCLUSIONS AND RECOMMENDATIONS Broad consensus exists about the appropriate cutoff criteria for intellectual ability in mental retardation diagnosis: A cutoff score of approximately two standard deviations below the mean is well accepted in most settings. A similar consensus does not exist, however, regarding the appropriate cutoff for adaptive behavior. The use of a stringent adaptive behavior cutoff like that used for intellectual functioning would sharply reduce the number of people with IQs below 70 eligible to be considered for a diagnosis of mental retardation. On the basis of the committee’s knowledge of individuals with mental retardation as well as the relevant research literature, this outcome is undesirable. We, therefore, propose formal adaptive behavior assessment as part of a comprehensive evaluation for individuals with or suspected of having mental retardation and cutoff scores that are more lenient than those widely used for intellectual functioning. Diagnostic decision making in mental retardation needs to be based on a comprehensive evaluation that uses multiple methods of collecting data from multiple sources across multiple settings. We support a principle of convergent validity as a means to interpret a broad variety of information. As discussed in Chapter 3, in the rare case in which a composite IQ is suspected to be spurious, the composite score should be ignored and either an appropriate part score (as described in Chapter 3) or other methods should be used to confirm or disconfirm a diagnosis of mental retardation. Diagnostic decisions should always be based on the preponderance of evidence, not just one numerical score. Finally, the need for more research, particularly on the measurement of adaptive behavior, is crucial to improving decisions about mental retardation eligibility. Adaptive behavior assessment is not as well

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits developed as intellectual assessment, although improvements over the last decade have occurred. On the basis of its review, the committee makes the following recommendations. Recommendation: A diagnosis of mental retardation must be based on high-quality assessments of intellectual and adaptive functioning that meet the following criteria: A broad variety of information on adaptive behavior and intelligence should be collected, including data on performance in different settings, from different sources, and using varying methods. Comprehensive, multifactored measures of intelligence and adaptive behavior should be used in mental retardation eligibility determination. Brief, unidimensional measures or short forms of comprehensive tests should not be used. The principle of convergent validity shall be applied in eligibility decisions about mental retardation eligibility. Information that is inconsistent with a diagnosis of mental retardation should be recognized, evaluated, and explained in the overall diagnostic decision. Assessments must be conducted by people with appropriate education and experience for the kind of instrument used and the nature of the eligibility decision to be made. People conducting intellectual assessments must meet publishers’ requirements for Class C instruments. Measures of adaptive and intellectual functioning should be carefully selected and interpreted in order to minimize the negative effects of low validity, low reliability, floor and ceiling effects, and steep item gradients. The norms for measures of adaptive and intellectual functioning must be suitably contemporary. Use of outdated norms or previous editions of recently restandardized measures is not

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits acceptable. The norms for intellectual measures should be no older than 12 years because of the deterioration of the normative standards over time. Decisions about mental retardation eligibility should be made by people with appropriate preparation in the areas of mental retardation and other disabilities and disorders, measurement of intellectual and adaptive functioning, knowledge of human development, and the influence of context on behavior. The committee’s recommendations reflect concerns about the quality of the available evidence as well as the necessity to provide reasonable guidance to people making eligibility decisions regarding a diagnosis of mental retardation. The committee concludes that more research on the measurement of adaptive behavior with children and adults is urgently needed, including investigation of classification agreement. The following recommendation reflects these concerns. Recommendation: Federal agencies, including the Social Security Administration (SSA), should fund studies to evaluate the accuracy of program eligibility decisions and foster research on adults with mental retardation, including their adaptive behavior. The research funding should include investigations of multimethod techniques for the assessment of job-related skills, social adaptation, health, and well-being. In addition, relevant epidemiological studies and research on the accuracy of diagnosis of mild mental retardation are essential to inform policy and decision making. SSA should evaluate the consequences of implementing the committee’s recommendations in the context of public policies and economic conditions, reporting findings to the public within five years. Since improved accuracy in eligibility determination depends more on improved measures of the key dimensions of mental

OCR for page 208
Mental Retardation: Determining Eligibility for Social Security Benefits retardation than on adjusting cutoff scores, the committee recommends research on improving measures, especially adaptive behavior assessment, and on methods to combine information on adaptive and intellectual functioning in making eligibility decisions based on a diagnosis of mental retardation. SSA should make available for use by legitimate researchers tapes of Supplemental Security Income and Disability Insurance program utilization, comparable to public-use tapes available for Medicaid program utilization. SSA should link its data on individual benefit awards to other agency data on health care and service costs for those same beneficiaries. SSA should examine data on eligibility determination procedures across its 10 districts, in order to discover if implementation of classification policies is consistent or varies regionally.