Statistics 850
Midterm 2 - answer key
2011-04-25 (p. 1)
1. In the generalized linear mixed models t to the employ data the family argument is given as
binomial. What is the link function, the inverse link function and the variance function associated
with
Statistics 850
Midterm 2
2011-04-25 (p. 1)
This is a closed book exam. You may use ve pages of notes and a calculator only. Put your
answers in the blue books provided. Each question is worth 10%. Your best 10 grades only are used.
Choose the questions yo
Preliminary exploration of the WRAP data
Douglas Bates
February 22, 2011
1
Building data frames, preliminary examination.
The WRAP data will be used in the rst midterm for Statistics 850, Spring 2011. You have had
the opportunity to examine the data for t
Nonlinear regression models
Douglas Bates
Spring 2011
Contents
1
Nonlinear regression
Nonlinear regression models
A linear or generalized linear regression model is usually an empirical model in the form
of the model is derived from the data.
Sometimes
Statistics 850, Spring 2011
Homework #4
Due: 2011-04-04 (p. 1)
The data used in these exercises are from the SabreR Coursebook (http:/sabre.lancs.ac.
uk/sabreR_coursebook5.pdf). Although the data sets themselves are available as a zip archive
(http:/sabre
Statistics 850, Spring 2011
Homework #2
Due: 2011-02-18 (p. 1)
Install the MEMSS package that contains data sets from the book Mixed-eects Models in S and
S-PLUS that I wrote with Jos Pinheiro.
e
1. Check the structure of the Rail data and the documentati
Mixed models in R using the lme4 package
Part 5: Generalized linear mixed models
Douglas Bates
Spring 2011
Contents
1 Denition
1
2 Links
2
3 Example
7
4 Model building
9
5 Conclusions
14
6 Summary
15
1
Generalized Linear Mixed Models
Generalized Linear Mi
Statistics 850
Midterm 1 - suggested answers
2011-02-25 (p. 1)
1. Explain the dierence between a xed-eects term for a factor and a random-eects term. Explain
why race is modeled as a xed eect.
A xed-eects term for a factor provides estimates of coecients
Representations of linear mixed models
Douglas Bates
Spring 2011
Contents
1 Denition
1
2 Obtaining parameter estimates
3
3 Model specications
4
1
Representations of linear mixed models
Representations of linear mixed models
Linear mixed models or multile
Examples used in software comparisons at NIST
Douglas Bates
Statistics 850
University of Wisconsin - Madison
<[email protected]>
Spring 2011
Douglas Bates (Stat. 850)
NISTnls
Spring 2011
1 / 67
Background
Researchers at the National Institutes of Standards a