Current Topic...

HBOK 1-056

Title Confidence Intervals for the Coefficient of Variation when Random Effects are Selected from Finite Populations
Author(s) Richard K. Burdick and Shea Watrin
Source Quality Engineering
Topic Data Analysis
Abstract The conceptualized infinite population from which random effects are sampled in many practical applications does not provide a good description of the data. When random effects are sampled from finite populations of known size, an assumed random effects model provides confidence intervals on functions of the variance components that are too wide. On the other hand, fixed effects models underestimate true variability. The problem is discussed here in a biomedical application by comparing alternative methods for modeling variation in the process.
Access Restrictions ASQ members and journal subscribers
Link for .PDF
Link for HTML None
Reference Code 1-056

Add a Comment