سال انتشار: ۱۳۷۹
محل انتشار: پنجمین کنفرانس آمار ایران
تعداد صفحات: ۹
Alastair Scott – Dept. of Stat., University of Auchkland, Private Bage, Auckland, Newzland.
J.N.K. Rao – Dept. of Math. And Stat., Carleton University, 1125 Colonel By Drive, Ottawa, ON K1S 5B6, Canada.
Most sophisticated statistical techniques were developed under the assumption of random sampling, and extending the range of application of these techniques to data from complicated sample surveys is notoriously difficult and cumbersome. Hinkins, Oh and Scheuren (1997) suggested avoiding some of the difficulties by analyzing a subsample with simpler structure than the full sample from which it is drawn, and then resampling repeatedly to increase efficiency. A similar idea has been suggested by Hoffman and Weinberg (1998) for handling clustered data in biostatistics. We look at some theory behind this idea and explore the limitations and applications of the method.