Difficulties in fitting scientific data

Abstract

The major potential difficulties in obtaining parameters of a model from scientific observations are discussed in the context of the theory of best approximation.