Inference for Marginal Means in Contingency Tables with Conditional Independence
研究在给定C层时A与B独立的列联表中，如何估计A的边际均值，给出了多种估计量及其方差公式，并用于假设检验和置信区间，适用于疫苗评估等场景。
Abstract Consider a three-way contingency table with factors A, B, and C. Assume factors A and B are independent for each given level of factor C. Inference procedures concerning the marginal means of the levels of factor A are studied. Maximum likelihood estimators, unbiased estimators, and uniformly minimum variance unbiased estimators are obtained. Exact variance formulas for the estimators are obtained. Large sample theory is developed and used for testing and confidence intervals. A variety of sampling models and generalizations are studied, including the case in which the number of levels of factor C is infinite. An example concerned with vaccine evaluation is given.