ABSTRACT
The statistical ordinary least squares (OLS) regression model is the reference thought of when marketers hear the words "new kind of model". Model builders use the regression concept and its prominent characteristics when judiciously evaluating an alternative modeling technique. This chapter presents a method for calculating a quasi-regression coefficient, which provides a frame of reference for evaluating and using coefficient-free models. The quasi-regression coefficient serves as a trusty assumption-free alternative to the regression coefficient, based on an implicit and almost never tested assumption necessary for reliable interpretation. The redoubtable regression coefficient, formally known as the OLS linear regression coefficient, linear-RC(ord), enjoys everyday use in marketing analysis and modeling. The interpretation of the regression coefficient in the multiple regression models essentially remains the same as its meaning in the simple regression model. Machine-learning models without coefficients can assuredly enjoy the quasi-RC method. One of the most popular coefficient-free models is the regression tree, for example, chi-squared automatic interaction detection (CHAID).
