Disability Surveys

The Value of Disability Surveys

The main value of disability surveys is that they are relatively inexpensive, unobtrusive, and accurate. Disability statistics can be produced in other ways: in a census that enumerates all persons in a population, not just a sample, or in administrative records that include all members of a population, such as the medical care system for elderly people in the United States. However, producing statistics from censuses and administrative records systems is more expensive, more burdensome, or less accurate than producing them from surveys.

Surveys can produce a variety of useful disability statistics. One important use of disability statistics is to formulate and evaluate disability policy. Proponents and opponents of disability policies in employment and transportation use survey statistics in debate. Disability statistics are also used by manufacturers and distributors of disability-related products and services (e.g., assistive devices, disability insurance plans), as they plan for production, advertising, and sales. Scientific researchers use disability statistics to investigate the causes and consequences of disability.

Design and Analysis of Disability Surveys

Samples for disability surveys are selected in several ways. In a scientific sample, the sample persons are selected at random with known probabilities of selection. A scientific sample can be selected from a preexisting list of persons; for instance, if most people have listed telephone numbers, a sample can be selected randomly from telephone directories. Also, a sample can be selected by randomly selecting residential areas, listing the addresses of dwellings in those areas, and then randomly selecting addresses from the list.

Samples for disability surveys may also be selected by establishing quotas for sample categories with selected characteristics. The quotas are set so that the resulting sample will have the same proportions of persons in each category as the population. For instance, if it is known from a census that 10 percent of a population is African American, then a sample of [Page 493]1,000 persons would be designed to include 100 persons in that category. Data collectors contact people in the population until the quota for each of the categories has been achieved.

While all disability surveys use standardized data collection techniques, they may use different modes of data collection. Frequently used modes are face-to-face interview, telephone interview, and self-administered questionnaires distributed by postal mail. Many survey organizations have begun to use self-administered questionnaires distributed by electronic mail or postings on the World Wide Web. Modes differ in quality of data and cost of administration: Generally, face-to-face interviews yield better quality but cost more, while electronically distributed questionnaires yield poorer quality but cost less. To optimize data quality and costs, disability surveys may use combinations of modes.

...