Abstract:
In this study we tried to bring out the importance of two methods used for visualizing multicollinearity diagnostics. The first method is called Tableplot (see Friendly and Kwan, 2009). Through this technique, multicollinearity diagnostics ideas are combined with a visualizing approach. The second method is called Biplot (see Gabriel, 1971; Gower and Hand, 1996). The biplot technique is used for demonstrating significant properties of multivariate data structure. Both of these approaches have been examined by one real and three artificial data. Results indicate that Biplot method is preferred to instead of Tableplot for too large condition indexes.
Page(s):
59-78
DOI:
DOI not available
Published:
Journal: Pakistan Journal of Statistics, Volume: 29, Issue: 1, Year: 2013