ABSTRACT
This chapter discusses the standard statistical methods to quantify the estimation error, and how we can make precise statements about it. It suggests quoting the standard error besides the estimator itself. The chapter considers the confidence intervals in a way symmetric around the estimator. The idea was to find a short range of values of the parameter which is in some sense consistent with the data. Sometimes, however, we are more interested in an upper (or lower) bound on the parameter. The chapter talks about a confidence bound for the unknown parameter. Confidence regions are a generalization of confidence intervals for multiparameter case. The corresponding Bonferroni’s confidence region could then be possibly reduced in size without losing the required confidence level. Sometimes one can utilize specific features of the model at hand to construct confidence regions in other ways.
