Ridge Regression as a Technique for Analyzing Models with Multicollinearity

This paper focuses on the issue of multicollinearity in family studies research. A technique called ridge regression is presented as a method for analyzing models containing multicollinearity. Multicollinearity is described as an often overlooked, but significant and pervasive problem in the family studies field. The paper examines some problems associated with using ordinary least squares solutions in models containing multicollinearity. In addition, some usual techniques for dealing with multicollinearity are discussed along with some of their advantages and disadvantages. Finally, the ridge regression solution is demonstrated as an alternative technique for identifying multicollinearity problems in a set qf data and as a means fbr producing reliable results in the presence of a multicollinear model. The study uses simulated data and a model of marital satisfaction to demonstrate the ridge regression technique. The results suggest that a ridge regression solution can produce results that are different from the ordinary least squares solution when the predictors are not orthogonal. Each of the different solutions suggests a different interpretation of the data.