STAT 512

Design and Analysis of Experiments

Spring

11:15-12:05

Course Personnel:

Professor: Naomi Altman naomi@stat.psu.edu ; 312 Thomas 865-3791

Web Page: http://www.stat.psu.edu/~naomi/stat512/

- office hours, homework and exam solutions, practice exams, exam times, homework hints and corrections, etc. will be posted here

Texts:

Required: Applied Linear Statistical Models, 4th Edition Neter, Kutner, Nachtsheim, and Wasserman (NKNW) 1996 Irwin, Inc. or 5th Edition Kutner, Nachtsheim, Neter and Li 2005, McGraw-Hill/Irwin.

There will also be handouts for material not covered by the textbook.

Prerequisites: STAT 511 or permission of the professor.

STAT 512 is intended primarily for first-year students in the Department of Statistics, but qualified students from other departments are welcome to attend. We will assume familiarity with elementary applied statistics, elementary probability theory, basic calculus and some matrix algebra, particularly the material in Chapter 5 of the text.

Auditors

Auditors are encouraged to participate fully in lectures but may not turn in material for grading.

Computing

We will use the SAS Statistical Analysis System. A brief SAS tutorial will be part of the first homework.

Homework

Weekly assignments are designed to help you assimilate and practice the techniques covered in class and to prepare you for the material to come. Some small assignments may be given in any class. These are to prepare you for class and are due as you come into class. Longer assignments are due at the start of Friday’s lecture but may be turned in as late as Friday at 2:30. (They can be placed in my mailbox.)

Homework, data and sample SAS commands will be posted to the Web. Homework should be typed. SAS output should be downloaded to a word processor and edited appropriately. Equations and other mathematical notation can be added by hand as needed.

Students are encouraged to discuss homework, but the work you hand in must be your own. Copying will be penalized.

Late assignments will be allowed only with the permission from Prof. Altman, documented by a note or e-mail. Permission will be granted only in cases of medical or family emergencies, or conflicting university commitments. You are welcome to check with Prof. Altman if you feel that you need an extension.

Exams

Dates (subject to change):

MidTerm Exam 1 week of Feb. 18 (evening)

MidTerm Exam 2 week of Mar. 24 (evening)

Final Exam week of May 5

Grade Breakdown

homework 25%

mid-term 1 20%

mid-term 2 20%

final exam 35%

 

Minimum total for letter grade

 

     60       65       70       75      81     86      90

      C        C+     B-        B       B+    A-      A

 

All Penn State and Eberly College of Science policies regarding academic integrity apply to this course. See http://www.science.psu.edu/academic/Integrity/index.html for details.

 

 

Statistics 512

Design and Analysis of Experiments

Syllabus

 

Topic

Text 4th Edition

Text 5th Edition

Design of Experiments and Observational Studies

26**

Kuehl Chap. 1

15

Kuehl Chap. 1

Planning for One-sample and 2-sample Tests

    

Planning for Linear Regression Analysis

    

Regression with Indicator Variables

11

8.3-8.7

One-way Fixed Treatment Designs

16.6 – 16.11, 18.1 - 18.6

16.1-16.6,16.8

Equivalent Forms of Coding Categories

11.6 + computer manuals

16.1,16.7

Analysis of Treatment Effects

17

17

Analysis of Covariance

25

22

Two-Factor Fixed Treatment Designs

19, 20, 21, 23

19,25

Multifactor Fixed Treatment Designs

23

24

Confounded Designs

31

29

Response Surface Designs

32

30

Random and Mixed Effects

24

25

Design of Randomization

26

15.3,15.4

Randomized Complete Block Design

27

21

Nested Design

28

26

Split Plot Design

29.6

27.6

Incomplete Blocks

Kuehl Chap. 9

28.1, 28.2

Latin Square Design

30

28.3-28.7

Repeated Measures and Longitudinal Studies

29

28.9

Sequential Experiments

    

Other Advanced Topics (time permitting)

    

**Chapter 15 of the 5th edition is preferable.

Some Very Important Vocabulary

Research program

Population of interest

Study population

Experiment

Comparative Experiment

Controlled Experiment

Observational Experiment

Prospective Study

Retrospective Study

Treatments

Factors

Levels

Experimental units

Experimental error

Response variable(s)

Covariate

Control treatment

Treatment Design

Experiment(al unit) Design

Replication

Efficiency

Power

Block