Abstract
This study demonstrates a better method of regression analysis than Ordinary Least Squares (OLS) method under certain conditions. Ridge Regression, as it is called is useful in situations where there are strong intercorrelations among regressor variables – a condition called multicollinearity. When OLS regression is used to model the relationship between the response variable and the regressor variables the model may exhibit some undesirable properties. Using ergonomics data exhibiting multicollinearity the use of the Ridge technique is demonstrated. An OLS model for the same data is also presented and compared with the Ridge models. The Ridge model was superior to the latter. It exhibited more realistic properties and predicted more accurately. It is therefore proposed as a valuable tool to the Human Factors/Ergonomics researcher in the development of regression models with highly intercorrelated regressor variables.
Get full access to this article
View all access options for this article.