Estimation and Use of Standard Errors of Latent Class Model Parameters
论证了在寻找简约模型时标准误的用处，并提供了使用MLLSA中Goodman迭代比例拟合算法估计所有参数标准误的方法。
Only recently have latent class models been used effectively to analyze marketing data, though they have been popular for more than a decade in the social sciences. Most research reported in the literture does not include the standard errors of the estimates of the latent class model parameters. The author argues for the usefulness of standard errors while exploring for parsimonious models. He provides an approach to estimating standard errors of all parameters as estimated by the iterative proportional fitting algorithm of Goodman implemented in MLLSA.