You are here

Survey Methodology1. Introduction

Survey Methodology
1. Introduction

Prior research has shown the important role that transportation plays in linking persons with disabilities to a range of community resources and services.[1] Full participation in society, including school, work, and other activities, requires access to effective modes of transportation for all persons, especially those with a disability. Economic independence and self-sufficiency are possible only when transportation is accessible, affordable, and available, close to home and to the many destinations that people need to reach.[2]

As one of twelve operating administrations within the U.S. Department of Transportation (DOT), the Bureau of Transportation Statistics (BTS) has a compelling interest in the transportation needs of persons with disabilities. One role of the BTS is to provide transportation data to decision-makers and other stakeholders, and its mission is to collect data, and compile, analyze, and publish statistics.

1.1 Goals of the Survey

In terms of national transportation statistics, one area that suffers from a critical lack of information is transportation use by persons with physical, mental or emotional disabilities. Little is known about the ability of the nations transportation system to fulfill the unique transportation needs of persons with disabilities. For this reason, the BTS initiated the 2002 Transportation Availability and Use Study to fill this information gap and allow the Government and other researchers to analyze the data, report the results, and help provide an informed basis for the development of transportation policies and programs.

In particular, the goals of the survey are to:

  • Identify the various modes of transportation used by persons with and without a disability, including pedestrian walkways, automobiles, public transit, and specialized transportation programs;
  • Describe the levels of satisfaction and problems that persons experience when using these travel modes;
  • Permit comparisons of transportation use patterns, satisfaction levels, and problems, between persons with and without a disability;
  • Provide data on access to an automobile (including vehicles adapted for use by persons with disabilities), driving behavior, accidents, and self-imposed limitations on driving;
  • Allow for the analysis of particular target groups of interest, such as the elderly, low-income persons, working-age adults, and children with and without disabilities; and
  • As an overall goal, create an information resource for transportation planners to use when developing national, state, and local policies and programs for persons with disabilities.

According to the Census Bureau, approximately 19 percent of the U.S. resident (non-institutionalized) population age 5 and above have a disability.[3] The disability prevalence rate among children under 5 years of age is approximately 3 percent.[4] In addition, analysis of the 1995 National Health Interview Survey on Disability (NHIS‑D), using disability measures that were similar to the 2000 Census, indicated that among households that report having persons with a disability, 79 percent contained only one person with disability, 18 percent contained two persons with disability, and the remaining 3 percent contained three or more persons with a disability.[5] This information allowed the study team to estimate how many households would need to be screened in order to locate one person with a disability for the survey. Based on this analysis, approximately 30 percent of all households were identified as having at least one person with a disability. Allowing for a 10 percent underreporting rate, the study team assumed a 27 percent rate when estimating the required sample size for the survey (see Section 2.1, Sample Design, below).

This study employed a screening questionnaire, followed by an extended survey instrument for the selected persons. Persons of any age (including children) were eligible to be interviewed. Proxy interviews with a knowledgeable respondent were required for persons under age 16 years and for persons who were unable to complete the interview for themselves due to the severity of their impairments. The survey employed data collection methods that ensured the greatest possible participation regardless of the respondents' limitations. This included Internet and mail versions of the questionnaire, a Spanish version of the questionnaire, and the use of interpreters. The study was conducted by Westat, a private survey research firm in Rockville, Maryland.

The survey gathered information in the following topical areas:

  • Frequency of travel outside the home, including trip purpose, mode of transportation, frequency of use for different modes, need for assistance, and satisfaction with transportation services;
  • Availability of paratransit (door-to-door service) and respondent use of paratransit;
  • Personal motor vehicle ownership, use, and safety issues, including vehicles modified for use by persons with disabilities;
  • Experiences when using various modes of travel, including difficulties with public and private transportation; and
  • Respondent demographics (gender, age, income, ethnicity, race, disability, living arrangements, employment status, school attendance, and education level).

The target for completed interviews was 4,000 persons: 2,000 with disabilities and 2,000 without disabilities. The study included both people with and without disabilities to allow comparisons between the two groups for various transportation uses and problems, and to determine which situations were unique to one group versus the other. With this information, the BTS and others can make recommendations to improve transportation for persons with disabilities.

The study actually completed 5,019 interviews, 2,321 with persons with disabilities and 2,698 without. The data files and documentation include many different disability measures, allowing analysts to construct their own definition of disability using the multiple items in the survey. The disability questions were taken from Census 2000 long form, as well as from the Americans with Disabilities Act language, and a question on special education was added by the project team. See the attached Computer Assisted Telephone Interviewing (CATI) questionnaire, items B2a.-B2e., for the specific wording of these questions. For additional information about the development of the Census disability questions, see the following Web sites:, and

1.2Constructed Variables

The study team constructed a separate variable, CDISABLD, that identified if a respondent reported one or more of the several Census disabilities (questions C3A-C4D). The purpose of this variable is to assist users who may want to compare the results of this survey with Census 2000 data, according to a common set of disability items. As a caveat, the 2000 Census collected disability information from persons who were 5 years of age and older, while this survey included persons of all ages. Therefore, when comparing the results of this survey with the 2000 Census disability data, users should select only persons who are 5 years of age or older.

In addition to the 2000 Census disability items, this survey asked two questions about a disability related to the Americans Act (ADA) (questions B2a. and B2b.), and one question about the receipt of special education services, which are designed for children with disabilities (question C5.). A second constructed disability variable, TDISABLD, identified if a respondent reported any of the ADA items, the special education item, or the Census disability items. The purpose of this composite measure is to give the user a variable that identifies respondents reporting any of the disabilities in this survey.

Also, the "Other, Specify" responses that could not be coded as an existing response item were placed into newly constructed categories consistent with the new information. Unique responses were left in the "Other" category. For ease of identification, the new categories appear in the codebook below the "Other" response. In addition, a MODE variable was constructed to identify if the extended survey was completed by CATI, Internet, or mail. There is a constructed age variable as well (AGE) which reconciles any differences between the screener and the extended reports of the respondent's age.

Sample weights were developed for this data set, including the final full sample weight (RAKEDW00), and 80 replicate weights (RAKEDW01-RAKEDW80), as described in Section 4., Weighting and Variance Estimation, below.

1.3Caveats on the Interpretation of Data

The public use data file consists of individual-level (person) records. For this reason, the counts produced from this data set are person-level rather than household-level figures. For example, the variable, TOTVEHI, identifies the number of household vehicles. However, one may not sum this count to determine the total number of vehicles in the country. This household vehicle count would require the use of household weights, which are not available for the extended questionnaire component of the survey. Only person weights are included in the file. Instead of a vehicle count, one may count the number of persons who live in households with a specified number of vehicles, (e.g., no vehicles, one vehicle, etc.).

Also, because this is a person file, with only individual-level weights, many of the household-level screener variables at the beginning of the survey are not applicable at the person level and, therefore, do not appear in the public use data file. This includes the two ADA disability questions (B2a. and B2b). However, if either of these ADA items was the sole basis for identifying a person with a disability for an extended interview, this was reflected in the construction of the TDISABLD variable. A similar situation occurred with the CDISABLD variable, where respondents were selected for an extended disability survey based on a positive screener response to one or more of the Census disability questions. The screener responses were used in these cases because these Census items were missing from the extended interview file. This occurred when there was a telephone disability screener response and a mail extended interview, and the extended interview Census disability questions were not answered by the respondent.

The written documentation is organized according to the following sections:

  1. Introduction, including background information, study goals, and survey content areas;
  2. Design and methods, including a description of the sample design, survey operations, and quality control methods;
  3. Response rates; and
  4. Weighting and variance estimation.

The documentation also includes:

The codebook, listing the variables for the questions and response items;

Tables of frequencies and percents with standard errors; and

Copies of the questionnaires, including the CATI, Internet and Mail versions.

The accompanying CD includes the documentation and data files in SAS 8.0, SPSS 10, Excel, and comma-delimited ASCII formats.

[1] Burkhardt, J., et al. Mobility and Independence. Ecosometrics, Inc., for the National Highway and Traffic Safety Administration and the U.S. Department of Health and Human Services, Washington, DC, 1998.

[2] Kerschner, H, and Aizenberg, R. Supplemental Transportation Programs for Seniors. The Beverly Foundation, Pasadena, CA. and the AAA Foundation for Traffic Safety, Washington, DC, 2001.

[3] U.S. Census Bureau, Census 2000. Table DP-2. Profile of Selected Social Characteristics: 2000.

[4] National Center for Health Statistics. Health United States, 2002. Hyattsville, MD: 2002.

[5] 1994/95 National Health Interview Survey on Disability, original tabulations from public use data files.