Abstract:
|
Survey costs are an important aspect of the total survey error perspective. Surveys either attempt to minimize error for a fixed cost, or obtain a specified level of error at any cost. Decisions about how to balance costs and errors are critical for the survey's success. There is a large survey literature on various sources of error - sampling, coverage, nonresponse, measurement. But the literature on costs is smaller and more fragmented, with costs reported differently across similar studies. This paper has two purposes: 1) to explore methods used for monitoring and estimating survey costs, and 2) to propose a set of standards for cost reporting and steps to help implement these standards. The goal is to improve the reporting of costs across studies so that survey methodologists will be able to make informed decisions about cost-error tradeoffs.
|