ABSTRACT Partial replacement of units in repeated surveys increases the efficiency of the estimation of the population mean. The composite estimator with constant coefficients, based on the recursive ratio, is useful in surveys with many variables. The mean square error of this estimator is obtained for an arbitrary rotation scheme. Comparisons indicate that it is more efficient than the sample mean for various rotation schemes. Simulations show that it performs better than other composite estimators in surveys with many variables changing differently over time.

中文翻译：

---

基于递归比的任意旋转方案的复合估计器

摘要 在重复调查中部分替换单位提高了总体均值估计的效率。基于递归比的具有常数系数的复合估计量在具有许多变量的调查中很有用。该估计器的均方误差是针对任意旋转方案获得的。比较表明，对于各种轮换方案，它比样本均值更有效。模拟表明，在许多变量随时间变化不同的调查中，它的性能优于其他复合估计器。