Psychological and Educational Assessment of Deaf and Hard-of-Hearing Individuals

The issue

braden-crop

Psychological and educational tests are designed to measure something. That “something” may be academic achievement, intelligence, depression, or something else. Unfortunately, tests also presume skills that are not related to what they are measuring and that can end up changing the meaning (validity) of the test result. For example, tests designed to measure intelligence, achievement, affect, or adjustment also measure some unintended things (e.g., reading ability, knowledge of the world) when used with test-takers who are deaf or hard of hearing (DHH). As an example, many physicians look for children begin to speak and the size and sophistication of their spoken vocabularies, as indicators of intellectual development. In general, smarter children talk earlier, know more words, and use more sophisticated language; severe delays often indicate abnormal intellectual development. However, using linguistic milestones as an indicator of intellectual development presumes that children have equal access to language. If they have a hearing loss, and they don’t have other access to language models in the home on a consistent 24/7 basis, the interpretation of delayed language milestones as an indicator of delayed intellectual development is invalidated. Thus, the issue for valid assessment of DHH test-takers is the ability to answer the question: “Does hearing loss affect test scores in ways independent of what is being measured, and if so, how?”

What we know

We know that most tests that are used in schools, clinics, and other settings presume that test-takers will have normal access to spoken language and educational and vocational opportunities (nearly all of which are mediated through spoken language), and will have no other problems that will interfere with their ability to respond to the test. We also know that these conditions are almost never true for individuals with hearing loss. Therefore, we know that we must either use tests that reduce these limitations or modify how we interpret the results in light of the test-taker’s hearing status in order to ensure we do not misinterpret (invalidate) the test results.  We also know that when we carefully interpret common tests into sign language (e.g., with back-translation procedures) or otherwise adapt their presentation using the test-taker’s primary mode of communication, the tests results are as reliable as they are from hearing test-takers. We also know that, although they may be equally accurate, they often do not have the same meaning.

If we intend to measure constructs like intelligence or general knowledge in DHH test-takers, and we either do not use language-reduced tests or fail to consider the role of hearing loss when interpreting the test results, we can make catastrophically bad decisions. We also know that careful selection of tests to reduce language demands (in testing and test content) is likely to produce valid results. Finally, we also know that just standardizing a test on DHH test-takers will not necessarily guarantee a test will be valid; instead, we have statistical tools to help us know whether items or test scores are producing results that are valid or invalid. Unfortunately, we rarely do the research needed to apply these tools to reasonable samples of DHH test-takers.

What we don’t know

We don’t know to what degree our efforts to tailor assessments to the unique needs of DHH test-takers might distort our conclusions. For example, if we only use nonverbal measures of intelligence, are our conclusions about general intelligence valid—or have we limited what we measure to exclude important domains associated with intellectual ability? More importantly, we do not know whether the results of educational and psychological tests add any value to decisions about programs or interventions. We assume that they do; that is, we assume that educational interventions designed to capitalize on cognitive strengths and weaknesses work better than interventions that are not—but we have little evidence to support those beliefs. Our practice of, and faith in, psychological and educational tests exceeds the knowledge base needed to support them.

Implications

We need to be thoughtful when selecting, administering, and (most importantly) interpreting the results of educational and psychological tests given to all test-takers—but especially those who have hearing loss. The potential for harm resulting from inappropriate assessment is significant—but so too is there potential harm in not using well-structured tests as part of an assessment. Some behaviors may be misinterpreted as evidence of other problems, and it is only through careful assessment conducted by a knowledgeable tester that we can best ensure that we enhance the welfare of all DHH test takers.

Posted on April 1, 2015 by
Jeffrey Braden
College of Humanities and Social Sciences
North Carolina state university
jpbraden {at} ncsu.edu

Further reading