Age Influences on Validity and Reliability

As a developmental stage, late life encompasses more than 45 years, from roughly ages 65 to 110. Not surprisingly, there is an extraordinary amount of heteroCopyright © 1998 by Academic Press. All rights of reproduction in any form reserved.

Assessment of Mental Health in Older Adults geneity in this population. Some elders are physically and cognitively very healthy; others develop disabling chronic illnesses quite early on. Thus, it is often very difficult to make generalizations about "the elderly.'' Not surprisingly, gerontologists have subdivided this developmental stage into three groups: the young-old, whose ages range from 65 to 79; the old-old (80-99); and the oldest-old, or centenarians. Others differentiate between optimal aging, in which there is little decrement or even improvement in some functions; normal aging, in which there are some decrements for which the elderly can readily compensate to maintain adequate psychosocial functioning; and impaired aging, marked by declines in physical and cognitive function.

Thus, it is very important to understand the position along these continua of the elder or sample of elders to be assessed. In general, in the United States, the young-old are relatively healthy and it is likely that assessment techniques used in younger populations are quite adequate for this population. Indeed, if one attempts to use instruments developed for impaired elders in the ordinary young-old population, one rapidly runs into ceiling effects—nearly all elders will score in the top range, rendering criteria for predictive and discriminant validity nearly useless. In other words, if there is no variance on an instrument, it cannot be used to correlate with other measures or to distinguish between groups.

In contrast, for frail elders, who are more likely to be in the old-old age group, the use of standard instruments may pose a problem in both the reliability and validity of the data. Cognitively impaired elders may become confused when confronted with typical Likert scaling, and dichotomously scaled instruments may have more reliability and validity. (We have found that even elders in good condition generally dislike and mistrust the Procrustean bed of fixed response formats, and often need to be cajoled into translating their phenomenological experience into admittedly arbitrary numbers.) In addition, frail elders may have poor attention spans, requiring the administration of brief forms of standard instruments and/or multiple testing sessions over several days. Although elders in general respond as accurately on surveys as younger populations, it is unlikely that cognitively impaired elders can do so, and interviews are more likely to yield valid information.

Elders with visual impairments may have difficulty in reading questionnaires, requiring the use of larger fonts. In addition, we have found that scantron sheets which use relatively pale type faces with poor contrast (e.g., lavender script on cream-colored paper) are con-traindicated with elders who have acuity problems.

Individuals with motor impairments, such as tremors associated with Parkinson's disease or severe arthritis in the hands or wrists, may have difficulty in filling out questionnaires, and will require longer periods of time to complete them. For elders with severe forms of these illnesses, scantron forms are virtually impossible. Some researchers have switched to computer presentations of instruments which can aid in overcoming such sensory and motor deficits.

In general, we have found that frail elders do best in interviews in which the required responses to questions are available in both verbal and visual forms. If Likert scales are necessary for some instruments, then response cards, written in large fonts, which elders can hold and point to responses, are very helpful. Given that cognitively impaired elders often have difficulty in switching tasks, changing response cards is a good way of signaling that one task is done and that attention needs to be refocused on another.

However, interviews conducted in home settings may pose a special problem in assessing the elderly. In our experience, it is very difficult to interview just one member of an elderly dyad, especially in long-term married couples. Such couples may learn to compensate for memory problems by consulting with each other, and typically the non-target spouse will respond to questions, making accurate assessment of the target individual problematic. Thus, we have found it necessary to physically separate couples, either by giving the non-target elder an instrument to complete in another room, or by using pairs of interviewers to conduct simultaneous interviews, again in separate rooms.

Biofeedback Mastery

Biofeedback Mastery

Have you ever wondered what Biofeedback is all about? Uncover these unique information on Biofeedback! Are you in constant pain? Do you wish you could ever just find some relief? If so, you are not alone. Relieving chronic pain can be difficult and frustrating.

Get My Free Ebook

Post a comment