COMPARATIVE BIAS BETWEEN SAMPLING FRAMES FOR FARM SURVEYS
提出了一种量化不同农场抽样框架相对偏差的方法，通过随机调查比较了四种抽样框架的代表性，帮助研究者选择最合适的框架并更准确理解研究结果。
This paper proposes a methodology for quantifying the relative bias inherent in various farm sampling frames. It can be applied to any area or region for which a significant sample of farms is required. A random survey examined the relative representativeness of selected farm characteristics by National Farmers' Union of Scotland (NFUS) and Scottish Landowners' Federation (SLF) membership lists, the British Telecom Yellow Pages farm business listings, and the Scottish Office Agriculture and Fisheries Department June census returns on farm holdings. It is suggested that a methodology to assess comparative bias should be employed prior to any extensive work requiring the use of a sampling frame for farm holdings or businesses, to ensure that the most appropriate framework is chosen and a greater understanding of research findings achieved.