Diffusion-weighted MRI (dMRI) represents a well-established field for the study of micro-structure in tissue and other porous media. Yet, the pipeline from image acquisition to analysis and application of imaging markers is dynamic and components within it are ever evolving. The field continues to enthuse researchers to extend, refine, and create imaging instruments as well as data processing and analysis methods. It is in this spirit that this work set out to improve – or augment to – the dMRI pipeline in areas associated with modeling, analysis, and validation. Biophysical models represent an important instrument for establishing links between the dMRI signal and biological properties of tissue microstructure. However, due to the number of parameters needed to faithfully describe the signal in terms of the complex tissue micro-architecture, biophysical modeling methods need to rely on assumptions that are often oversimplifying. One component of this work conducted an empirical study of the model in Neurite Orientation Dispersion and Density Imaging (NODDI) regarding assumptions for the intra-cellular parallel intrinsic diffusivity. Understanding individual variability of dMRI-based markers is an important task when studying conditions with high inter-subject heterogeneity such as traumatic brain injury and autism. A second component of this work focused on developing a framework based on the Mahalanobis Distance, as a multivariate approach for individualized evaluation of imaging-based measures. Validation of apparent diffusion coefficients (ADCs) estimated from dMRI measurements requires the use of test media with known diffusion coefficients as well as their temperature dependence. A third component of this work sought to develop a test object (phantom) constructed from liquids whose diffusion constants at multiple temperatures were previously characterized. Models of the diffusion-temperature relationships where investigated for their potential use in estimating the temperature of the phantom from dMRI measurements.