Week |
Monday |
Tuesday lab |
Weds |
Friday |
1 |
Aug 24
Intro/Types of studies
|
Aug 25
Hands-on randomization test and bootstrap
FYI, I won't use the board again! I now know how to draw/write on
the computer screen.
|
Aug 26
Types of studies
|
Aug 28
Design-based inference for 2 group comparison
|
2 |
Aug 31
Model-based confidence intervals
|
Sept 1
Q/A on HW 1, SAS intro
|
Sept 2
T tests, interpretation of p-value
|
Sept 4
Sample size determination
|
3 |
Sept 3
Labor Day, No lecture
|
Sept 8
Q/A on HW 2 (but no questions were asked), SAS: proc ttest, reading data
|
Sept 9
Assumptions: independence and equal variance, transformations
|
Sept 11
Normality, non-parametric tests
|
4 |
Sept 14
Choosing a test, 1-way ANOVA
|
Sept 15
Cider case study, SAS for diagnostics, Levene's test, Wilcoxon test
|
Sept 16
Fixed, Random effects, Linear contrasts
|
Sept 18
Orthogonal contrasts, multiple comparisons
|
5 |
Sept 21
Philosophy of multiple comparisons, Diagnostics and Consequences
|
Sept 22
Lie detector electrodes case study,
Q/A on HW 4, SAS for contrasts, pairwise comparisons.
|
Sept 23
Pairing and blocking
|
Sept 25
Block designs: Inference, Efficiency and Assumptions
|
6 |
Sept 24
Blocking: interp. of MSE
|
Sept 29
Lie detector electrodes case study (part 2), SAS for RCBD, a bit of
JMP for RCBD, contrasts, and transformations, a bit of Q/A on HW 5
|
Sept 30
Q/A on HW 5, Latin Squares, Linear Regression
|
Oct 2
Linear regression: estimation, inference on parameters and line
|
7 |
Oct 5
Prediction intervals, calibration, ANOVA table, measurement error
|
Oct 6
Q/A for midterm I
|
Oct 7
Measurement error, Assumptions, ANOVA lack of fit test
|
Oct 9
Correlation
Note: The audio went silent about 5 minutes in. The material on p
176 was
repeated after the recording was rebooted.
|
8 |
Oct 12
Correlation example, Intro to multiple regression
|
Oct 13
Ships case study,
Q/A on HW 6, SAS: linear regression
|
Oct 14
Matrices, estimation in multiple regression, the hat matrix
|
Oct 16
Inferences about parameters
|
9 |
Oct 19
Model comparison tests, Cb tests, interpretation of coefficients
|
Oct 20
Netherlands
birth case study, Q/A on HW 7, Geometry of regression, SAS for
multiple regression, including test and contrasts
|
Oct 21
Importance of an X variable, Multicollinearity, Regression outliers
|
Oct 23
Influential points, DFBETA's, DFFIT's, Cook's D, loess concept
|
10 |
Oct 26
LOESS details, unequal variances, Breusch-Pagan test
|
Oct 27
Lack fit for births data, SAS for regression diagnostics, Q/A on HW 8
|
Oct 28
Correlated errors: effects, diagnosis, correction
|
Oct 30
Estimation
in models with AR(1) errors, non-normality, Model construction
|
11 |
Nov 2
Model construction using indicator variables
|
Nov 3
Case study: model construction, Q/A on HW 9.
|
Nov 4
ANCOVA, model selection: intro
|
Nov 6
Trade off between bias and variance, Fit-based model selection,
Mallow's Cp.
|
12 |
Nov 9
Model Selection using AIC and BIC, after model selection
|
Nov 10
Q/A for midterm II, SAS for regression model selection
|
Nov 11
Misusing and using model selection, 2 way factorial terminology
|
Nov 13
2 way factorial: SS as formulae, SS as contrasts among cell means
|
13 |
Nov 16
2 way factorials: SS as model comparison, Interpretation of interactions
|
Nov 17
Case study: ANCOVA and other approaches for baseline data,
Q/A on HW 10,
SAS for 2 way factorials
|
Nov 18
2 way factorials: after the F test, 3 way factorials, factorials in
blocks, unreplicated factorials, fractional factorials.
|
Nov 20
Random effects, broad and narrow sense inference, estimating
variance components
|
14 |
Nov 30
Subsampling, Split plot studies
|
Dec 1
Case studies on mixed models, Q/A on HW 11
SAS for random effects and split plot designs.
|
Dec 2
Split plot studies, repeated experiments
|
Dec 4
Unequal n
|
15 |
Dec 7
Categorical data: inferences on a single proportion, differences in
proportions, intro to odds ratios
|
Dec 8
Smoking/Cancer case study, Q/A on HW 12, SAS for contingency
tables.
|
Dec 9
No lecture because ISU cancelled.
|
Dec 11
Contingency tables, Chi-square statistics, McNemar's test
|