Introduction
Obtaining survey data has become a challenging task, as response rates have decreased over the years in the United States and Europe (Atrostic, Bates, Burt, & Silberstein, 2001; de Heer, 1999). Collecting data from hard-to-survey populations is even more difficult; they are either hard to reach or known for low cooperation rates (for a more extensive discussion, see Tourangeau, Chapter 1 in this volume).
Complete lists covering many hard-to-survey populations do not exist (Sudman & Kalton, 1986) and there is no simple method to define these groups (Lin & Schaeffer, 1995; Smith, 1983). Nevertheless, researchers have made an attempt to identify characteristics of typical nonrespondents (e.g., Caetano, Ramisetty-Mikler, & McGrath, 2003; Gannon, Northern, & Carroll, 1971; Shahar, Folsom, & Jackson, 1996). Many nonresponse characteristics found in these studies are sample-specific and therefore not useful for other investigations. While most surveys cannot produce response rates by population group, use of inclusion rates (i.e., a ratio of the estimate in a survey to an official estimate) can provide useful information (Griffin, 2012). Hard-to-survey groups also possess characteristics that have demonstrated barriers to participating in many studies.