Experimental Design and Analysis

Syllabus

Spring Semester, 2009

Reference Text

Statistical Principles in Experimental Design (2nd Ed.) by B. J. Winer, McGraw Hill, New York, 1971.

Reading Material- Pages in Winer 

Review of Elementary Statistics (pp.4-160) 

Basic Terminology in Statistical Sampling and Estimation (pp.4-14)

Biased vs. unbiased

Expected value of a statistic

Sampling distribution

Standard Error of a Statisitic

Central limit theorem

Level of significance

Type I and type II errors

Power of a Statistical Test

Testing Hypothesis about Means (pp14-26)

Sigma assumed known

Sigma unknown

Interval estimation

Testing Hypotheses about the Differences Between Two Means (pp.26-54)

Outline of steps in an experiment

Testing for homeogeneity of variance

Test differences of means of random independent samples

Test differences of means of random dependent samples

Analysis of Variance (ANOVA) (pp149-160) 

Logic behind the ANOVA

Completely randomized designs (CR-k) 
            definitions

symbolization

formula derivation

The Design of Experiments (pp.309-319, 431-445,452-465) 

Criteria for a Good Experimental Design (p150)

Completely Randomized Factorial Designs (CRF-pqr...) 

Derivation of formulas for SS and df

Comparative advanatages and disadvantagesof the CR-k and CRF-pqr...designs.

Randomized Block Designs, RB-k (pp.261-267). 

General purpose

Derivation of computational formulas

Structural models

Estimated means of squares

Advantages and disadvantages of design

Randomized Block Factorial Designs, RBF-pqr...(p514-18) 

Purpose

Assumptions

Derivation of computational formulas

 Expected values of mean squares

Advantages and disadvantages

Split-Plot Factorial Designs SPF-pq.rs..(pp514-571

General Purpose (pp514-518)

Case II (pp559-571)

structural model

computational formulas

Generalization to all other designs

Advantages and disadvantages of model

Hierachal Designs, CRH-p(q)(r)...(pp-359-366;464-468)

Purpose and use

Examples

Computational; proceedures

Structural Models

Generalization to all cases
 

Partial Hierarchal Designs, CROH-p(q)r (pp.464-468)

Examples

Computational proceedures

Structural Model

Estimated mean Square

Multiple Comparisons Following ANOVA 

F-tests (pp30, 145,545, and 563-567) 

Tukey's Test (p.198) 

Scheffe's Test 

Test for Trends (pp.177-182) 

Additional Designs

Latin Squares (pp685-700)

Greco Latin Squares (pp 709-711)

Covariance Design, ANCOVA (755-778)

Terminology

General Purpose

Computational proceedures

                        Distinguishing designs

 

Lecture Topics

1. Definitions and Concepts – January 15

2. Testing Hypothesis – January 20

3. 1-way ANOVA – January 22

4. 2-way ANOVA Terminology – January 27

5. 2-way ANOVA Procedures – February 03

6. Multifactor ANOVA's – February 10

7. Expected Mean Squares – February 12

8. Randomized Block Des. – February 17

9.Split-Plot Designs – February 19

10. SP- Procedures – February 24

11. SP-Procedures – February 26

12. Hierarchal Designs – March 10

13. Nested Factors – March 12

14. Individual Comparisons – March 17

15. Individual Comparisons – March 19

16. Individual Comparisons – March 24

17. Individual Comparisons – March 26

18. Tests for Trends – April 7

19. CRF & Added Controls – April 9

20. Latin Square Design – April 14

21. Greco Latin Square – April 16

22. ANCOVA Terminology & Procedures – April 23

Grading

Course grades will be assigned according to the following levels of achievement

Course Grade Percentage Points

    Course Grade                    Percentage Points
        A                                       92.6-100
        A-                                      90-92.5
        B+                                     87.6-89.9
        B                                       82.6-87.5
        B-                                      80.0-82.5
        C+                                     77.6-79.9
        C                                       72.6-77.5
        C-                                     70.0-72.5
        D+                                    67.6-69.9
        D                                       62.6-67.5
        D-                                     60.0-62.5
        F                                       59.9 and Below

Grades will be based on written in-and-out-of-class assignments and class participation. In-class written assignments will be due at the end of the class period. Other written assignments, except for the final analysis, will be due one week after they are assigned. All errors in written assignments should be corrected so that you have a complete set of notes.

 Academic honesty Policy
 The policy on academic honesty as stated in the Student Handbook will be followed. According to the Student Handbook,  Academic dishonesty refers to "any unacknowledged use or misappropriation of material, as in the cases of cheating or plagiarism."

Additional Information

        1. Office Information
                a. Office hours: 10:00-12:00, (M&W): other times by appointment
                b. Office Location: Fisher Science Bldg. Room 10
                c. Office Phone: 4388
      
      

 

 

Susquehanna University assumes no responsibility for the content of this personal web page. Please read the disclaimer.