A Monte Carlo investigation of scaling as an alternative to regression in the bootstrapping model

This paper examines scaling via conjoint measurement as an alternative to the usual multiple linear regression in the "bootstrapping" model. The research was based on synthetic data from a Monte Carlo simulation. It was found that, for data that cannot be presumed to be at least interval scale measures, scaling is superior to multiple linear regression as a method of approximating past data. In addition, it was found that assuming data to be ordinal measures when the data are actually interval scale measures is a much less severe error than is the error of assuming data to be interval scale measures when the data are actually only ordinal measures.