1: \begin{abstract}

2: Motivated by recent work studying massive imaging data in the

3: neuroimaging literature, we propose multivariate varying coefficient

4: models (MVCM) for modeling the relation between multiple functional

5: responses and a set of covariates. We develop several statistical

6: inference procedures for MVCM and systematically study their

7: theoretical properties. We first establish the weak convergence of

8: the local linear estimate of coefficient functions, as well as

9: its asymptotic bias and variance, and then we derive asymptotic bias

10: and mean integrated squared error of smoothed individual functions and

11: their uniform convergence rate. We establish the uniform convergence

12: rate of the estimated covariance function of the individual functions

13: and its associated eigenvalue and eigenfunctions. We propose a global

14: test for linear hypotheses of varying coefficient functions, and

15: derive its asymptotic distribution under the null hypothesis. We also

16: propose a simultaneous confidence band for each individual effect

17: curve. We conduct Monte Carlo simulation to examine the finite-sample

18: performance of the proposed procedures. We apply MVCM to investigate

19: the development of white matter diffusivities along the genu tract of

20: the corpus callosum in a clinical study of neurodevelopment.

21: \end{abstract}