ABSTRACT
This chapter describes the nature of the sample and the survey instruments. It presents some background on the study areas, and methods used to conduct the statistical and geographic analysis. The key portion of the total design method is the sequence of mailings and follow-ups designed to increase response rate. The sample of homeowners to be surveyed was randomly drawn from a list of the entire population of detached, single-family, owner-occupied homeowners. A classic approach of mail and telephone surveys, with a prescription for increasing response rates, was developed by Don A. Dillman. The 13-page survey questionnaire was sent to 3,421 households previously identified through the random sampling method. The distance from each surveyed homeowner's geographic location to the nearest special studies zone was computed as the Euclidean distance. The decision to purchase earthquake insurance was hypothesized to be influenced by the relative location of the home to areas of risk, termed "adverse selection."
