Access to Assessment? Legal and Practical Issues Regarding Psychoeducational Assessment in Children With Sensory Disabilities

Assessment Practices With Students With Visual or Hearing Impairments

It could be argued that standardized psychoeducational assessment measures are simply not suited for use with individuals with sensory impairments and thus should not be used. Indeed, test manuals do strongly caution against the use of common assessment batteries with individuals with sensory impairments (e.g., Wechsler, 2004). However, there is evidence that these measures are commonly used in such a manner. For example, a major guidebook (Flanagan & Kaufman, 2009) on administering the Wechsler Intelligence Scale for Children, Third Edition (WISC-III; Wechsler, 2004) includes suggestions for administering the test to examinees with sensory disabilities. Although the authors of the guidebook indicate that there are situations in which certain subtests should not be administered, the fact that such information is available—and does not de facto rule out administering the assessment—suggests these are strategies practitioners may use when administering the WISC-III to children with sensory impairments.

In addition, surveys of practitioners indicate that it is relatively common for practitioners to use major standardized psychoeducational assessments, such as the WISC-III, with examinees who have sensory impairments. For example, J. C. Miller and Skillman (2003) surveyed assessors who work with examinees with visual impairment to assess their use of and satisfaction with different intellectual and cognitive assessment measures in this population. They found that the most commonly used tests were indeed those that are popular for use with the general population (e.g., the Wechsler and Woodcock–Johnson batteries). However, they also found that, despite their frequency of use, satisfaction with these measures when used with examinees with visual impairment was generally mediocre to low. Conversely, cognitive measures developed specifically for use with examinees who are blind or have low vision were rarely used among the practitioners; however, practitioners who did use them reported being extremely satisfied with these assessment measures.

Furthermore, in 1993, Weaver and Bradley-Johnson reported that the WISC-III (Wechsler, 1991) was the second most preferred test among school psychologists who worked with examinees who were deaf and hearing impaired. However, 30% of the 371 respondents indicated concerns about inadequate testing materials, 27% reported difficulties with communication during test administration, and 7% reported concerns about providing fair and equitable assessment (Weaver & Bradley-Johnson, 1993). In addition, Krouse and Braden (2011) contacted practicing psychologists and retrieved a sample of 128 scored WISC-IV protocols that had been administered to children who were deaf or hard of hearing between October 2003 and April 2007. Taken together, these studies suggest that standardized psychoeducational assessments, such as the WISC-IV, are being regularly used by psychologists who assess individuals with sensory impairments. However, the results also suggest that, as with J. C. Miller and Skillman’s (2003) participants, they do not find these assessments fully satisfactory.

These findings beg the question of why examiners who work with individuals with visual or hearing impairment would select assessments that have been not normed with their examinee population and that they generally find unsatisfactory. It is possible that this could represent an availability issue. Cognitive assessment can be a financially driven area of psychological practice (Dial & Dial, 2010). Because these tests can cost several hundred dollars per kit (see www.pearson.com for pricing on the Wechsler tests and other psychoeducational assessments), agencies and training programs may choose the kits they purchase judiciously. People with visual impairment make up a relatively small fraction of the population; thus, producing and purchasing assessments for examinees with visual impairment may not be financially viable (Dial & Dial, 2010).

Although they may be inaccessible to some individuals with disabilities, assessments developed for the general population, such as the Wechsler tests (Wechsler, 1991, 2004, 2008) are so widely researched and used that they are typically considered the “gold standard” for assessment (Dauterman & Suinn, 1966; Dial & Dial, 2010; J. C. Miller & Skillman, 2003). Examiners may hesitate to use less well-known tests that used smaller norming samples and less rigorous standardization procedures and thus may avoid using tests that have been specifically developed for individuals with visual impairment due to psychometric limitations of such tests (Dial & Dial, 2010). As B. D. Miller (2008) discussed, even the few mainstream tests that include samples of examinees with hearing impairments include very small (e.g., n = 35) and homogeneous samples that are inadequate for true norming. More research is needed to understand the test selection criteria examiners use with selecting assessments to use with examinees with sensory disabilities (J. C. Miller & Skillman, 2003).

Regardless of why the tests were selected, however, it is important to note that even though the majority of respondents to J. C. Miller and Skillman’s (2003) survey reported using assessments designed for the general population, they did not report that they found them satisfactory for use with individuals with visual impairment. Thus, the mere use of such measures with population should not be interpreted as an endorsement of their validity or acceptability. As with the testing of individuals who are deaf or have hearing impairment, the limitations of these tests collide with the perceived need to use some sort of standardized assessment to measure the intellectual abilities of individuals with sensory impairment (Dial & Dial, 2010; Metz et al., 2010), despite the clear or apparent biases of the tests or lack of accessibility of entire subtests or subscales.

Biased and Unbiased Assessment in Educational Placement

Prior to the passage of the Education for All Handicapped Children Act in 1975, the first federal law guaranteeing that all children with disabilities receive a free and appropriate public education and the predecessor of IDEA (2004), the United States Supreme Court had already addressed the issue of bias in educational assessment. In the case of Diana v. State Board of Education (1970), the court ruled that schools could not use intelligence tests that were administered in English or that were culturally biased in order to classify Mexican American students or others with limited English proficiency as having intellectual disability (then referred to as “mental retardation”). This ruling clarified that testing used to place children in special education programs or to classify them with intellectual disability must be accessible to the student being tested and not overly biased by cultural or linguistic factors.

The landmark case of Larry P. v. Riles (1979, 1984) extended the implications of Diana v. State Board of Education. Larry P. addressed the use of intelligence tests—which were, at the time, normed only with Caucasian samples—to classify African American children as having intellectual disability. The judge ruled that many items on these tests relied on experiences, knowledge, and cultural norms more likely to be experienced and shared by middle- and upper-middle class Caucasian children than children of other racial, ethnic, and socioeconomic backgrounds and therefore were culturally biased. The Larry P. ruling led to a complete ban on the use of standardized intelligence testing with African American children in California (Larry P. Task Force, 1989). School psychologists were instead encouraged to use a variety of other assessment methods (e.g., ecological assessment and observation, psychological processing assessment, neuropsychology assessment, and review of school records) to determine whether African American students in California qualified for special education services (Larry P. Task Force, 1989). The underlying idea behind the ban on intelligence testing following the Larry P. case was that intelligence tests were strongly biased in favor of one cultural group and against others, thus implying that they were testing culture perhaps as much as they were testing intellectual ability (Larry P. Task Force, 1989).

Although subsequent court cases have not upheld the strict ban on intelligence testing implemented after Larry P. v. Riles, which has since been overturned, concern about cultural and linguistic bias in psychoeducational testing remains a prominent feature in legal and judicial proceedings. PASE v. Hannon (1980) involved parents of African American children who sued so that their children could be given intelligence tests. The plaintiffs in PASE argued that because African American students lacked access to intelligence testing, they were unjustly denied needed special education services due to a lack of evidence of disability. Thus, the courts moved away from the notion that intelligence testing was inherently and perhaps inevitably biased against non-Caucasian or low socioeconomic status populations and toward the idea that all students should be assessed using measures that are not culturally biased and that have been adequately normed on students of diverse racial and ethnic groups and socioeconomic status (Slippery, 1981). In other words, the standard encouraged by the courts has gone from simply excluding minority groups from intelligence testing to ensuring that access to intelligence testing includes all individuals in a way that is not biased against certain groups (Slippery, 1981).

Implications of Unbiased Testing for Students With Sensory Disabilities

To date, little has been written on the rights of students with sensory disabilities regarding fair and unbiased psychoeducational testing. Although the examiner is instructed to inquire whether the student needs and is wearing corrective lenses or hearing aids (see sample protocols in Wechsler, 2004, 2008; Woodcock et al., 2001), these guidelines assume that students’ sensory impairments are correctable. The publishers provide no official guidelines regarding examinees who have hearing or vision impairments that cannot be fully corrected (Wechsler, 2004, 2008; Woodcock et al., 2001). The only advice given is simply to question the validity of the assessment (Flanagan & Kaufman, 2009). This raises an interesting question—if test results are considered invalid if the student has a correctable but uncorrected sensory impairment, are they also considered invalid if the individual has an uncorrectable sensory impairment?

Diana v. State Board of Education (1970) made it clear that it is unreasonable to test students in media that they cannot understand due to noncognitive factors. If it is considered unacceptable to test students in a language other than their primary or home language, it can be argued that it is also unacceptable to test someone who is blind on their ability to identify pictures or someone who is deaf on their ability to hear sounds. Doing so would be essentially testing the student’s (noncognitive) disability and not his or her intelligence or achievement; this could be seen as akin to testing an examinee’s English language proficiency as opposed to their intelligence.

Accommodations and Test Validity

In other testing situations, individuals with disabilities may be accommodated through alterations to the test or testing situation that accommodate for the functional impact of the person’s disability (Americans With Disabilities Act, 1990). However, intellectual and achievement tests are standardized under tightly controlled and universal procedures and materials (Flanagan & Kaufman, 2009); thus, any alteration to the testing procedures or materials may alter the validity of the test and compromise the results (Flanagan & Kaufman, 2009). Even seemingly “common sense” accommodations, such as reading written test items to an examinee with visual impairment or providing a deaf or hard-of-hearing examinee with a sign language interpreter, could constitute a breach of standardization and invalidate the test. Although some guidelines for accommodating individuals with physical and sensory disabilities during testing situations do exist (see Hill-Briggs, Dial, Morere, & Joyce, 2007), the authors of the guidelines caution that the impact or lack thereof of these accommodations on test validity have not been empirically or legally established, and examiners should take caution in ensuring that accommodations do not alter the difficulty or targeted outcome of a given test (Hill-Briggs et al., 2007). These cautions may be particularly relevant in educational settings, given the often complex legal concerns and guidelines surrounding issues of educational assessment and special education placement and services (IDEA, 2004).

Students Who Are Deaf or Have Hearing Impairment

Students With Visual Impairment

Like students who are deaf or have hearing impairment, students with visual impairment also may be unable to access standardized psychoeducational testing (Dial & Dial, 2010). Major intelligence and achievement tests assume normal vision. Accordingly, they require that examiners ensure that the student is wearing corrective eyewear, if needed, during testing (Flanagan & Kaufman, 2009). However, students who qualify as having visual impairment under IDEA (2004) have either no vision or severely impaired vision that cannot be adequately corrected using corrective lenses. Whereas concerns about the accessibility and validity of psychometric testing for children who are deaf or have hearing impairment center on the language-based components of the said tests, examinees with visual impairment are typically able to complete language-based subtests, as many verbal subtests are presented orally (Wechsler, 2004), and subtests that rely on reading can be produced in Braille without concerns about cross-language translation (Flanagan & Kaufman, 2009).

Assessment Measures and Strategies Used by Examiners

As mentioned above, one strategy for intellectual assessment with individuals who are deaf or have hearing impairment has been to use nonverbal intelligence tests, such as the UNIT. Although they are imperfect alternatives to a full intellectual battery, such as the WISC-IV, these tests have the advantage of having been normed on relatively large and representative samples and are typically accepted as a complementary assessment in situations in which a child’s test results may be confounded by their English language proficiency or lack thereof (Fives & Flanagan, 2002).

Conversely, examiners who are assessing students who are blind or have visual impairment do not have the option of any truly well-standardized and common intellectual assessment measures (Dial & Dial, 2010). Interestingly, Dial and Dial (2010) noted that although a number of intelligence tests for individuals who are blind or have visual impairment have been developed, most of these tests have small, nonrepresentative norming populations, and few are available for use by practitioners today. Some of these assessments are discussed below. As a result, many examiners instead use intelligence tests designed for the general population (J. C. Miller & Skillman, 2003), which are readily available and have large norming samples and strong psychometric properties (Wechsler, 2004, 2008).

Adaptation of Subtests Developed for the General Population

Some effort has been made to modify inaccessible subtests on the WISC-IV and other similar tests, including the adult version of the WISC-IV, the Wechsler Adult Intelligence Scale, Fourth Edition (WAIS-IV; Wechsler, 2008), so that they are accessible to examinees with visual impairment. These alternate or modified subtests generally attempt to access spatial skills and problem solving and processing through tactile means (Dial & Dial, 2010), such as haptic tests that target pattern identification (J. Miller et al., 2007) or tactile version of block design (Dauterman & Suinn, 1966; J. C. Miller & Skillman, 2008). There is not yet a “gold standard” tactile assessment measure for use with individuals with visual impairment, and the assessment measures that have been developed for this population are frequently unavailable for purchase by practitioners (Dial & Dial, 2010) and are thus rarely used (J. C. Miller & Skillman, 2003). Thus, given the wide array of possible assessments, we have chosen to focus generally on some examples of two common forms of tactile subtest modification—tactile block design and haptic matrices tests.

Implications and Possible Solutions

Although formal, standardized intelligence and achievement testing are still considered key components of current psychoeducational assessment practices (see Dial & Dial, 2010), other supplemental sources of information should be used to provide a more comprehensive picture of a student’s overall academic and functional performance and needs. For example, behavioral observation can provide valuable information on a student’s actual day-to-day performance in a classroom setting (Larry P. Task Force, 1989), and the current IDEA amendment (2004) required that a functional behavioral assessment be included in any psychoeducational evaluation that involves concerns about emotional or behavioral functioning. This indicates a legal acknowledgment of the prominent place that observation should have in assessment in general. Likewise, Dial and Dial (2010) emphasized the importance of behavioral observation in addition to psychoeducational testing when evaluating children or adults who have visual impairment.

As J. C. Miller and Skillman (2008) and Dauterman and Suinn (1966) both noted, observing how an individual examinee interacts with a given task (e.g., corrections or error patterns) can provide valuable information in regard to their approach to challenging tasks and problem solving. In addition, examiners should consider whether examinees with sensory impairments have difficulty with particular subtests or items due to sensory deficits and record this information on the test protocol (see Wechsler, 2004, 2008; Woodcock et al., 2001) and take this into account when interpreting the results (Flanagan & Kaufman, 2009).

In addition to observation, professionals who are assessing students with sensory disabilities should obtain information on the student’s academic and psychosocial functioning from multiple sources. These may include reports from vocational or service agencies (see Dauterman & Suinn, 1966), teacher or parent report forms (e.g., the Behavioral Assessment Scale for Children [BASC-2]; Reynolds & Kamphaus, 2006), or comprehensive, standardized evaluative batteries (e.g., the Comprehensive Vocational Evaluation System; Dial et al., 1990), which are specifically designed for use with individuals who are blind or have visual impairment, and collect information on both functional status and skills as well as containing an accessible cognitive assessment battery.

When assessing students with sensory impairments, practitioners should give increased consideration to data from supplemental sources that assess skills and domains covered by the inaccessible portions of intelligence and achievement tests. For example, because the verbal skills of a child who is deaf or hard of hearing and communicates via sign language cannot be assessed through a standardized intelligence assessment, information about the student’s social and functional communication skills should be gathered via other sources (Metz et al., 2010). Similarly, observation of a student with visual impairment may focus more on problem-solving and processing skills (Dial & Dial, 2010), cognitive domains that are typically assessed through the visually loaded PRI and PSI scales.

Finally, it is important to note that some portions of standardized psychoeducational assessment may provide valid and valuable information regarding the intellectual functioning of students with sensory impairments. For example, the verbal subscales of the WAIS have been shown to have good psychometric properties in some samples of examinees with visual impairment (Dauterman & Suinn, 1966), and the PRI WISC-IV subscale (Krouse & Braden, 2011) and the UNIT (Krivitski et al., 2004) have been shown to have adequate psychometric properties in samples of children who are deaf or have hearing impairment. Nevertheless, professionals who conduct psychoeducational evaluations of students with sensory impairments should be continuously mindful of the serious issues of validity that exist with psychoeducational testing in these populations, especially when making decisions about placement and eligibility as they relate to cognitive functioning. Also, as Dial and Dial (2010) noted, level of disability, onset of disability, and other factors may further confound the psychometric validity and reliability of intelligence tests in students with sensory impairments. Even in the few situations in which norming data exists for students with sensory impairments, samples tend to be small and homogeneous (B. D. Miller, 2008), limiting generalizability.

Summary

There is a clear precedent in the law and courts regarding the necessity of unbiased psychoeducational assessment that does not discriminate on the basis of race, ethnicity, language, or culture (e.g., Diana v. State Board of Education, 1970; Larry P. v. Riles, 1979, 1984). Although the parallels between cultural and linguistic diversity and sensory impairment are not exact, psychoeducational assessments should also avoid bias on the basis of a vision or hearing impairment to provide meaningful, valid results. However, given the strict procedures surrounding the administration of intelligence and achievement tests (Wechsler, 1991, 2004, 2008; Woodcock et al., 2001), accommodating examinees with sensory disabilities may also risk invalidating the test.

Commonly used tests, such as the Wechsler and Woodcock–Johnson batteries, presume normal hearing and vision and are thus inaccessible to individuals with sensory impairments in their standardized forms (Flanagan & Kaufman, 2009). For examinees who are deaf or have hearing impairment, verbal subtests may be extremely problematic, due to their requirements for aural questions and oral responses (Maller, 2003; Metz et al., 2010). Similarly, examinees with visual impairment may experience difficulties with tasks that require visual coding, symbol writing, or the manipulation of nontactile objects, as in block design. Although current practice recommendations suggest eliminating entire subtests or indices from test administration due to the presence of a sensory impairment, this strategy also creates issues with validity, resulting in an incomplete profile of the examinee’s cognitive strengths and weaknesses.

Conclusion

Due to the numerous barriers that prevent many examinees with sensory impairments from receiving valid and unbiased psychoeducational assessment through standard batteries, questions of legal, ethical, and educational best practice are raised. Psychoeducational assessment, particularly intelligence testing, has long been a cornerstone of disability assessment in schools. However, for individuals with sensory impairments, best practice may involve de-emphasizing the prominence of standardized intellectual and achievement testing in psychoeducational evaluation.

Standardized assessments should be interpreted in the context of the student’s sensory disability and the limitations of the test or tests due to that sensory disability. Graduate students and professionals in school psychology should receive training on limitations and best practices when assessing students with sensory impairments, a topic which has been de-emphasized in the past (see Weaver & Bradley-Johnson, 1993). Simultaneously, other metrics, such as classroom observations and school performance data, provide additional context by which the evaluation team should view standardized assessment data. This may be seen as analogous to the Larry P. Task Force (1989), which called for a dynamic, multimodal assessment model in response to serious concerns about the validity and bias of standardized intelligence tests among African American children. Similarly, RTI may hold promise as a more fair means of assessment for students with sensory impairments, as the RTI-based model of learning disability assessment emphasizes curriculum-based monitoring and response to instruction over time instead of psychoeducational assessment scores (Fletcher et al., 2004). This decreases the emphasis on traditional standardized psychoeducational assessment and, thus, reduces the potency of the barriers associated with inaccessible tests.

As it now stands, modern intelligence tests have been normed on racial, ethnically, geographically, and socioeconomically diverse samples to address and ameliorate many concerns about test bias. Therefore, it is possible that individuals with sensory disabilities may one day have access to unbiased and fully accessible psychoeducational assessment. Until that point, researchers and policy makers should work to develop clear and evidence-based guidelines on best practices for psychoeducational assessment in students with sensory disabilities.