4

I'm fairly new to this site. I'm currently doing some research work that involves the use of non-linear mixed effect model on longitudinal data (R). I'm presently using Pinheiro and Bates' book, Mixed-Effects Models in S and S-PLUS, with the nlme package. By non-linear, I mean non-linear functions of independent variables with normal responses.

I'm still a beginner with statistics and I'm finding the theoretical treatment of the subject in the above book quite daunting to understand. I was wondering if there are other "easier to digest" resources (including books) out there to try and get to grips with the theory. I understand how to use it in R but I still find that there are some fundamental concepts and gaps in my knowledge.

gung - Reinstate Monica
  • 132,789
  • 81
  • 357
  • 650
John_dydx
  • 593
  • 1
  • 5
  • 19
  • 1
    By "the Jose and Douglas book", do you mean Pinheiro and Bates' *Mixed-Effects Models in S and S-PLUS*? – Glen_b Dec 02 '13 at 17:59
  • You may want to check out this CV thread: [advanced statistics books recommendation](http://stats.stackexchange.com/questions/33197/). In addition, it may help you to read through some of the threads on CV that discuss / explain the topics that you are currently having trouble w/. Lastly, feel free to ask questions here concerning those topics that you don't understand that are not currently covered here. – gung - Reinstate Monica Dec 02 '13 at 18:09
  • @Glen_b, yes you're quite right. I meant the Pinheiro and Bates' Mixed-Effects Models in S and S-plus. – John_dydx Dec 02 '13 at 19:08
  • @gung, thanks for the link. I'll check out the books. – John_dydx Dec 02 '13 at 19:10
  • You're welcome, @John. Can you clarify what you mean by "non-linear"? Eg, do you mean things like GLMM & GEE (to deal w/ non-normal response variables), or do you mean non-linear functions of your predictor variables w/ response variables that may even be normal (there's a good example [here](http://stats.stackexchange.com/questions/63978/do-statisticians-assume-one-cant-over-water-a-plant-or-am-i-just-using-the-wro/64039#64039))? – gung - Reinstate Monica Dec 02 '13 at 19:15
  • @gung, I meant non-linear functions of independent variables with normal responses. – John_dydx Dec 02 '13 at 19:23

1 Answers1

1

Regarding mixed-effects models, in my opinion, the best applied book is:

For more on fitting them with different software,

is a handy supplement. (I've never read Pinheiro & Bates, which presumably covers this territory R-wise in better detail.)

I'll let someone else comment on the ideas underlying non-linear regression, but note that there is this book:

It's not about the underlying theory; it's just trying to get you quickly and easily to the point where you can fit them in R. I don't have much to say about it. I perused it once and it was easy to follow, but I don't actually use nonlinear regression.

gung - Reinstate Monica
  • 132,789
  • 81
  • 357
  • 650