Observed confidence levels for regression parameters
Poliak, Cathy D.
MetadataShow full item record
Regression models are used in various subjects such as economics, chemistry and engineering. Suppose a sample is taken from a population in which regression parameters are of interest and the parameter space has been divided into a countable set of possible overlapping subsets. The problem of regions is concerned with which of the subsets contains the true parameter. To solve this problem a measure of confidence is applied to these regions. This paper develops a measure of confidence for regions of a parameter space for regression parameters. Some literature that investigates the problem of regions and the development of the confidence levels are discussed in this paper. We will also look at the asymptotic properties of these measures. Then the methods are applied to examples.