Testing of risk difference under inverse sampling

摘要

The inverse sample is the current hot issues abroad, in the continuous sampling process. Inverse sampling is considered to be a more appropriate sampling scheme than the usual binomial sampling scheme when the subjects arrive sequentially and when the underlying response of interest is acute. In this article, we study various test statistics for testing rate ratio in case-control studies under inverse sampling. We use the way of Fisher-score to get the variance of interest parameter. Then we get the type I error and power by Monte Carlo simulation. In general, Score statistic is the best, which can get the small typeI error and large power.