Feasibility of Utilizing Cluster Sampling in the Navy: A Demographic Comparison of Cluster and Stratified Random Samples

摘要

Traditional mail surveys sent to randomly selected members have been the predominant data collection mechanism to assess the attitudes and opinions of U.S. Navy personnel. Over the past ten years, response rates on large-scale Navy surveys have steadily declined. For example, the Navy Equal Opportunity/ Sexual Harassment (NEOSH) Survey, which is mailed biennially to a Navy-wide sample, dropped from 60 percent in 1989 to 30 percent in 1999. On-site survey administration at selected Navy fleet concentration areas has been suggested as a possible solution; however, this sort of convenience sampling is not scientific and cannot be generalized to the Navy's population. If a methodology could be utilized that linked targeted command survey administration with representative, scientific sampling, many of the shortfalls in current Navy survey administration could be addressed. The purpose of this study was to determine the feasibility of utilizing cluster sampling to develop a statistically representative, scientifically valid sample at randomly selected Navy commands. This study statistically compared the demographics of a cluster sample to a stratified random sample.