STAT 516 (Statistical Methods II)

Summer II 2009

Syllabus

Syllabus (Word document) or Syllabus (pdf format)

Instructor

David Hitchcock, assistant professor of statistics

Office Hours -- Summer II 2009

Mon-Tues-Wed-Thu 9:40-10:15 a.m. and 1:00-1:30 p.m.
Please feel free to make appointments to see me at other times.

209A LeConte College
Phone: 777-5346
E-mail: hitchcock@stat.sc.edu

Course Meeting Times

Mon-Tue-Wed-Thu 10:30AM-12:45PM, LeConte College, Room 210A

Purpose: To complete a basic two course sequence (in conjunction with STAT 515 or 509) in statistical techniques available to the general practitioner for analyzing experimental data. To introduce students in many different disciplines to multiple regression and analysis of variance for basic experimental designs. To provide students with the knowledge to implement and interpret these standard linear models.

Current Textbook: Statistical Methods, Second Edition, by R.J. Freund and W. J. Wilson, Academic Press, 2002.

Course Notes

Homework Assignments

DatePossible Quiz Problems for this Date's Quiz Homework Assigned this Date
Tuesday, July 7No quiz on this date Page 326: Concept Questions 3, 4, 5
Pages 327-328: Exercises 1(a,b), 2(a), 5(a), 6(a).
Wednesday, July 8No quiz on this date Page 326: Concept Questions 12, 15, 18
Pages 327-328: Exercises 1(d), 3(a,c,d).
Thursday, July 9 Page 326: Concept Questions 4, 5, 12, 15, 18
Pages 327-328: Exercises 1(b,d), 2(a), 3(a,d), 5(a), 6(a).
 Page 326: Concept Questions 2, 11, 13, 14, 17, 19, 20
Pages 327-328: Exercises 1(c), 2(b),
3(b,c,e)[for part(e), interpret r^2],
5(b)[calculate a 95% prediction interval]
Monday, July 13 Page 326: Concept Questions 2, 11, 13, 14, 17, 19, 20
Pages 327-328: Exercises 1(c), 2(b),
3(b,e)[for part(e), interpret r^2],
5(b)[calculate a 95% prediction interval]
 Pages 399-400: Concept Question 4
Pages 400-408: Exercises 6(a)[except computing VIFs], 8(a)
Tuesday, July 14 Pages 399-400: Concept Question 4
Pages 400-408: Exercises 6(a)[except computing VIFs], 8(a)
 Pages 399-400: Concept Questions 1,
3(a,b,c)[see formula near p. 363 bottom,
see p.627-628 for F critical value], 9
Pages 400-408: 5(a), 5(b)[consider log transformations
of weight, dbh, height. Why?],
6(a)[compute VIFs], 6(b), 8(b)
Wednesday, July 15 Pages 399-400: Concept Questions 1,
3(a,b,c)[see formula near p. 363 bottom,
see p.627-628 for F critical value], 9
Pages 400-408: 5(a), 5(b)[consider log transformations
of weight, dbh, height. Why?],
6(a)[compute VIFs], 6(b), 8(a, b)
Pages 399-400: Concept Question 5
Pages 400-408: Exercises 8(a,c)[part(c):
also check for influence points]
Thursday, July 16No quiz on this date Pages 279-280: Concept Question 2
Monday, July 20 Pages 279-280: Concept Question 2
Pages 399-400: Concept Question 5
Pages 400-408: Exercises 8(a,c)[part(c):
also check for influence points]
 Pages 279-280: Concept Questions 7, 9, 10, 11, 15
Pages 280-286: Exercises 1(a), 2(a), 4(a), 15(a)
[Do the ANOVA F-test for each of these]
Tuesday, July 21 Pages 279-280: Concept Questions 7, 9, 10, 11, 15
Pages 280-286: Exercises 1(a), 2(a), 4(a), 15(a)
[Do the ANOVA F-test for each of these]
 Pages 279-280: Concept Questions 4, 6, 12, 13
Pages 280-286: Exercises 1(b,c)[do Levene test, residual plot, Q-Q plot],
2(b)[do Tukey procedure], 2(c), 15(b)[do Tukey procedure]
Wednesday, July 22 Pages 279-280: Concept Questions 4, 6, 12, 13
Pages 280-286: Exercises 1(b,c)[do Levene test, residual plot, Q-Q plot],
2(b)[do Tukey procedure], 2(c), 15(b)[do Tukey procedure]
Pages 455-460: Exercises 2, 3, 9
[For these exercises, just use R to compute
complete ANOVA table as we did in class]
Thursday, July 23 Pages 455-460: Exercises 2, 3, 9
[For these exercises, just use R to compute
complete ANOVA table as we did in class]
Pages 455-460: Exercise 2, 3, 9
[For all 3 exercises, do F-tests for interaction
and (IF APPROPRIATE) for main effects of each factor]
[If interaction exists, find and interpret interaction plot]
Pages 455-460: Exercise 2:
[Using R: Use Tukey procedure to determine which pairs of levels
(among M,P,R) of factor A have significantly different mean response]
Pages 455-460: Exercise 3:
[Using R: Does mean response for level 3 of C differ significantly
from mean response at other levels, when factor A is at level "6"?]
Pages 455-460: Exercise 9:
[Using R: Use Tukey procedure (alpha=0.05) to find which factor level
COMBINATIONS significantly differ in terms of mean volume]
Monday, July 27 Pages 455-460: Exercise 2, 3, 9
[For all 3 exercises, do F-tests for interaction
and (IF APPROPRIATE) for main effects of each factor]
[If interaction exists, find and interpret interaction plot]
Pages 455-460: Exercise 2:
[Using R: Use Tukey procedure to determine which pairs of levels
(among M,P,R) of factor A have significantly different mean response]
Pages 455-460: Exercise 3:
[Using R: Does mean response for level 3 of C differ significantly
from mean response at other levels, when factor A is at level "6"?]
Pages 455-460: Exercise 9:
[Using R: Use Tukey procedure (alpha=0.05) to find which factor level
COMBINATIONS significantly differ in terms of mean volume]
 Extra Problem: For the *altered* Table 9.22 data set on the
course web page, try to perform the ANOVA F-tests.
What is the problem? Do an alternative analysis to
get F-test results.
Tuesday, July 28No quiz on this date --
Wednesday, July 29 Extra Problem: For the *altered* Table 9.22 data set on the
course web page, try to perform the ANOVA F-tests.
What is the problem? Do an alternative analysis to
get F-test results.
 Pages 498-507: Exercise 1:[Test for a signif. difference in
mean oxygen consumption across the 3 treatments. Also test for
significant variation among blocks (i.e., replications or locations).]
Pages 498-507: Exercise 1:[Using a contrast, test whether the control
yields significantly lower mean oxygen than then other regimes.]
Pages 498-507: Exercise 8:[Test for significant differences
in mean mileage across additives.]
Pages 498-507: Exercise 8:[Use Tukey's procedure to determine which
pairs of additives significantly differ in terms of mean mileage.]
Thursday, July 30 Pages 498-507: Exercise 1:[Test for a signif. difference in
mean oxygen consumption across the 3 treatments. Also test for
significant variation among blocks (i.e., replications or locations).]
Pages 498-507: Exercise 1:[Using a contrast, test whether the control
yields significantly lower mean oxygen than then other regimes.]
Pages 498-507: Exercise 8:[Test for significant differences
in mean mileage across additives.]
Pages 498-507: Exercise 8:[Use Tukey's procedure to determine which
pairs of additives significantly differ in terms of mean mileage.]
 Pages 498-507: Exercise 15(b)
Monday, August 3 Pages 498-507: Exercise 15(b) Pages 547-556: Exercise 9(a) [Conduct F-tests for interaction
and for main effects using Type III SS]
Pages 547-556: Exercise 9(b)
Tuesday, August 4 Pages 547-556: Exercise 9(a) [Conduct F-tests for interaction
and for main effects using Type III SS]
Pages 547-556: Exercise 9(b)
Pages 547-556: Exercise 1 [Obtain estimated ANCOVA model;
Test for significant effect of weaning-weight on 9-week-weight;
Test for significant effect of weaning time on 9-week-weight;
Interpret estimated coefficient of weaning weight;
Test the equal-slopes assumption]
Pages 547-556: Exercise 5(c) [Obtain estimated logistic regression
equation; estimate and interpret odds ratio]
Wednesday, August 5 Pages 547-556: Exercise 1 [Obtain estimated ANCOVA model;
Test for significant effect of weaning-weight on 9-week-weight;
Test for significant effect of weaning time on 9-week-weight;
Interpret estimated coefficient of weaning weight;
Test the equal-slopes assumption]
Pages 547-556: Exercise 5(c) [Obtain estimated logistic regression
equation; estimate and interpret odds ratio]
 --

STAT 516 Supplementary Material

Available Computing Resources: R is available as a free download (from the CRAN home page) and students who want SAS can buy a copy from USC Computer Services.
These packages are also available on the computers in the labs in LeConte College (and a few other buildings).
SAS is available in LC 124 and possibly PSC 109. R is available in LC 124 and LC 303A. Help in using R can be found on the CRAN home page.

Downloading Instructions for R

Computing Tips: Some Review

Computer Code for Class Examples

Statistical TopicExample in SAS Code Example in R Code
Simple Linear Regression
and Correlation
(Chapter 7)		 SAS example: (House data)
Output for SAS example 		R example: (House data)
Output for R example
Multiple Linear Regression
(Chapter 8)		 SAS example (California rain data)
Output for SAS example 		R example (California rain data)
Output for R example
Transformations
(Chapter 7-8)		 SAS example (surgical data)
Output for SAS example 		R example (surgical data)
Output for R example
One-Way Analysis of Variance
(Chapter 6)		 SAS example: (Rice data)
Output for SAS example 		R example: (Rice data)
Output for R example
Two-Way ANOVA & Factorial Experiments
(Chapter 9)		 SAS example (Gas mileage data)
Output for SAS example 		R example: (Gas mileage data)
Output for R example
Randomized Block Design
(Chapter 10)		 SAS example (Wheat data)
Output for SAS example 		R example (Wheat data)
Output for R example
Randomized Block Design with Sampling
(Chapter 10)		 SAS example: (Rubber data)
Output for SAS example 		R example: (Rubber data)
Output for R example
Latin Square Design
(Chapter 10)		 SAS example (Productivity data)
Output for SAS example 		R example: (Productivity data)
Output for R example
Analysis of Unbalanced Data
(Chapter 11)		 SAS example (Table 11.2 data)
Output for SAS example 		R example (Table 11.2 data)
Output for R example
Analysis of Covariance (ANCOVA)
(Chapter 11)		 SAS example (Trigonometry scores data)
Output for SAS example 		R example (Trigonometry scores data)
Output for R example
Logistic Regression
(Chapter 11)		 SAS example (TIF data)
Output for SAS example 		R example (TIF data)
Output for R example

Required Computing Resources: Some problems in this course involve significant computations, and for summer 2009, we will primarily learn to use the software package R to do the needed computations.
You will need access to a computer with R (available as a free download from the CRAN home page). This can easily be downloaded onto your home computer.
Alternatively, you could use the software package SAS (available in the LeConte College computer labs).
Example code in both R and SAS will be provided on the course web page, but we will use R for the in-class examples.
Both R and SAS are available on the computers in the labs in LeConte College. Help in using R can be found on the CRAN home page.

HELPFUL R RESOURCE:


Supplementary material available for download:
Click here for Basics of R: A Primer, by Don Edwards.
Introduction to R, Comprehensive R Archive Network (CRAN).
Available at CRAN home page (click "Manuals" at left side of page; then choose the first manual, "Introduction to R").

HELPFUL SAS RESOURCE:


Data Sets

Review Sheets

Formula Sheets

You may bring to the first exam one standard-sized sheet of paper with anything you want written on it (for example, you should write any formulas you may need).

You may bring to the second exam one standard-sized sheet of paper with anything you want written on it (for example, you should write any formulas you may need).

You may bring to the final exam three standard-sized sheets of paper with anything you want written on it (for example, you should write any formulas you may need).

Exam Solutions

Exams