Is In-Person Interviewing Necessary to Adequately Survey Persons with Disabilities?
Eric Grau, Mathematica Policy Research
Keywords: disability, interview mode, bias
In-person interviews are generally viewed as necessary to survey persons with disabilities since sample members’ impairments may preclude them from completing an interview by telephone or mail. In addition, persons with disabilities are often low-income and highly mobile, thus tracking and locating this population can be challenging without a field component. The National Beneficiary Survey (NBS), a multi-wave survey of Supplemental Security Income and Social Security Disability Insurance beneficiaries sponsored by SSA, was conducted by telephone with in-person follow-up for telephone nonrespondents. To assess the impact of adding in-person interviewing on bias in survey estimates, we used paradata and administrative data to compare attributes of respondents and nonrespondents 5 months and 60 days into field efforts to no field effort. We also report on an experiment which compared the quality of data collected by telephone and in-person. While we find some evidence of bias when comparing no field effort to some field effort, cases receiving 60 days were comparable to those receiving 5 months. We found only modest differences in data quality between telephone and in-person modes.