Lecture notes
- Review points.
- Factorial and split plot designs.
Data sets:
alkaloid content, and
oats.
- Mean Squares approach for RCBD with and
without subsampling, and for the Latin square design.
Data sets: seed emergence,
millet yield.
- Mean Squares approach and f-test
for CRD with and without subsampling.
Data set: irrigation.
- Testing random and fixed effects with
the likelihood ratio test. Data sets:
pygmy marmoset language and
flexible learning
in bilingual infants.
- Models with random effects: introduction and
estimation of coefficients. Last updated: 3/24.
Data sets: corn and
floweringTime.
- Causal inference and randomization.
- Three-way interactions, overdispersion,
and simulations for inference.
Data set: lupine seedling survival.
- Logistic regression: the logistic curve,
interpretation of coefficients, chi-square (LRT) test, the problem
of perfect separation.
Updated with graphs on 3/5.
Data sets: anesthesia,
runoff,
baby food,
S.aureus.
- Model selection:
significance with f-test and t-tests, correlated predictors.
Pesticide toxicity data set: toxic.txt.
Model selection with different criteria and different search methods.
- Multiple linear regression:
re-parametrizations, two-way interactions, linear and log
transformations, polynomial regressions.
Mussel shell thickness data set:
freemanByers.txt.
Answers to the 5-question quiz on slide 20.
- Multiple linear regression:
interpretation of a linear model.
Birds and bats case study: bats.txt.
- Introduction and quick review of simple
linear regression. fev data set.
Last modified: Wed May 5 13:41:59 CDT 2010