I'm a beginner in data analysis who needs to learn (say in a period of 2 to 3 weeks or so) the key ideas and techniques in the linear mixed effect models for longitudinal data. I'll apply them in computational neuroscience, for shape analysis. I'd appreciate if you could please suggest me a good reference, or lecture notes available online which deals with them. I've graduate level mathematics and decent statistics with less programming background, so a mathematically sound book which also gives me the motivation and refer to some programming will be great for me, I suppose.
So far, I finished the chapter 3 and 4 of "Applied Longitudinal Analysis" (by Fitzmaurice, Laird and Ware) to get some introduction to the longitudinal data, and the estimation of their parameters. Chapter 8 of this book discusses linear mixed effect models. But for some reason, I find part of their writing kind of hand-wavy instead of being clear, for example, on page 93, they didn't exactly mention how they'd estimate the covariance matrix of errors/noise. Also, on P. 93, the motivation for the correction term for doing the restricted maximum likelihood isn't clear to me. Because of these reasons, I'm looking for alternate references to explain the mixed effect models in a clear and concise way with sufficient motivation and preferably with sufficient mathematical arguments.