Confidence Intervals Comparisons for the Slope in Linear Measurement Error Models

Abstract

This paper investigates the interval estimation for the slope parameter in linear measurement error model under the identifiability assumption that the ratio of error variances is known. Comparisons of different confidence intervals based on coverage probability and expected length are discussed. A compendium of existing results and new results are presented. Theoretical results are further compared and supported by simulation studies. Some real data examples are also analyzed. (This is joint work with Jia-Ren Tsai and Hans Schneeweiss).