论文部分内容阅读
It is commonly believed that the effect of ignoring covariance structure is mainly on the lost of efficiency.In this talk, I will use some numerical outcomes and examples to illustrate potential concerns when one ignores the correlations in longitudinal covariates measurements.The less known fact is the serious level of biases that could be induced by ignoring existing correlations.This motivates a procedure which aims at careful selection of measurement error models.