statistically valid” cannot legitimately be made. In other words, a probability sample and its results are always “valid.” Because of differences in the choice of a design, the level of available resources, and the method of estimation, however, some procedures lead to higher precision (smaller confidence intervals) than other methods.