Diagnostics and Robust Estimation When Transforming the Regression Model and the Response.
Abstract
In regression analysis, the response is often transformed to remove heteroscedasticity and/or skewness. When a model already exists for the untransformed response, then it can be preserved by applying the same transform to both the model and the response. This methodology which we call 'transform both sides' has been applied in several recent papers, and appears highly useful in practice. When a parametric transformation family such as the power transformations is used, then the transformation can be estimated by maximum likelihood. The MLE, however, is very sensitive to outliers. This article proposes diagnostics to indicate cases influential for the transformation or regression parameters. Also proposed it a robust bounded-influence estimator similar to the Kraskeer-Welsch regression estimator. Keywords: Heteroscedasticity; Maximum likelihood. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 1986
- Accession Number
- ADA177531
Entities
People
- David Ruppert
- Raymond J. Carroll
Organizations
- University of North Carolina at Chapel Hill