Sorry, we do not have enough data to show an influence graph for this author.
- Full text PDF available (0)
Multicollinearity is a serious problem in regression analysis. High correlation among predictor variables can lead to unstable estimates of the regression coefficients. Principal components offers one approach to reducing the sampling variance of the coefficient estimates. However, this method produces biased estimates and the bias cannot be measured… (More)