ABSTRACT
In Chapter 5 we studied how to conduct various ANOVA inferences for functional data under the assumption that different groups have a common covariance function. In this chapter, this assumption is dropped as in practice, this assumption may not always be satisfied. From this point of view, the ANOVA models studied in Chapter 5 can be referred to as homogeneous ANOVA models while the ANOVA models considered in this chapter are called heteroscedastic ANOVA models. When there are only two groups of functions involved, the associated problem may be referred to as a two-sample BehrensFisher problem for functional data.
