Advanced Statistics for Biologists

STT 814 Sections III & IV

Michigan State University

Spring, 2012

 

COURSE: Statistics 814 Advanced Statistics for Biologists

 

INSTRUCTOR: Dr. R.J. Tempelman

Department of Animal Science

1205J Anthony Hall

Tel: 355-8445 (or 5-8445 from a campus phone)

e-mail: tempelma@msu.edu

 

TEACHING ASST: Xiaoqing Zhu (zhuxiaoq@msu.edu)

 

 

OFFICE HOURS: Tempelman: Monday 11:20 a.m.-12:00 p.m.

Thursday 3:20 p.m.- 4:00 p.m.

If at all possible, please call ahead as I am sometimes called away to various meetings.

 

 

LECTURE: 10:20-11:10 Mondays, Wednesdays, and Fridays

Room 155 Communication Arts Building

 

SECTION 3 LAB: 1:00 2:50 Thursdays, 222 South Kedzie

 

SECTION 4 LAB: 3:00 4:50 Thursdays, 222 South Kedzie

 

TEXTS:

 

Required:

         Applied Linear Statistical Models, Fifth Edition (2005) by M.H. Kutner, C.J. Nachtsheim, J. Neter, and W. Li., McGraw-Hill Irwin, New York ISBN 0-07-238688-6

 

Recommended:

         SAS System for Regression, Third Edition (2000) by R.J. Freund and R.C. Littell. SAS Publishing (http://support.sas.com/publishing/bbu/companion_site/57313.html ) ISBN:978-1-58025-725-1.

 

         SAS System for Mixed Models, Second Edition (2006) by R.J. Littell, G.A. Milliken, W.W. Stroup, R.D. Wolfinger, and O. Schabenberger. SAS Publishing (http://support.sas.com/publishing/bbu/companion_site/59882.html ) ISBN: 978-1-59047-500-3.

 

         Analysis of Messy Data, Volume 1. Designed Experiments (2009) by G.A. Milliken and D.E. Johnson. CRC Press, Boca Raton, FL ISBN-13:978-1-58488-334-0.

 

 

GRADING PROCEDURE:

 

Hourly Test (~February 24, 2011) ............20%

Take-home Test (handed out early April, due within 48 hours)...............................25%

(Bi)Weekly Laboratory Assignments....................................30%

Final Examination (2 hours Friday May 4, 2011 @ 10:00 a.m.)..........................25%

 

The first test will be in class. The second test will be a take-home test due by the subsequent class period (i.e. within 48 hours).

 

HOMEWORK GRADING:

 

Xiaoqing Zhu (zhuxiaoq@msu.edu) will serve as the homework grader for the course. If you have questions on pending homeworks, please ask me. If you have questions about the grading of your homework, please first consult the homework key (uploaded to ANGEL after the grades are assigned) before checking with Xiaoqing. If you still have questions after all that, then please consult with me.


OBJECTIVE:

To enhance the student's statistical and software-intensive toolbase for the design and analysis of experimental and observational research data in the biological sciences.

 

COURSE CATALOG DESCRIPTION:

Concepts of reducing experimental error: covariance, complete and incomplete block designs, latin squares, split plots, repeated-measures designs, regression applications, and response surface designs.

 

CLASS TOPIC SCHEDULE:

 

TOPIC

APPROXIMATE NUMBER OF LECTURES

Introduction:

Scientific inquiry and the linear model

1

Regression

Linear regression and residual diagnostics review, Lack-of-fit hypothesis testing

Matrix algebra and modelling

Introduction to multiple linear regression

Modelling response surfaces

Type I /Type III sums of squares

Multiple regression influence/diagnostics

13

Factorial Designs

Completely randomized designs

Two-factor analysis of variance

Multifactor analysis of variance

Regression approach to factorial designs

Analysis of covariance

Power assessments

9

Blocking Designs

Randomized Block Designs

Incomplete Block Designs

7

Hiearchical Designs

Nested designs and Subsampling

Split plot designs

Repeated Measures Analysis

Mixed effects models

7

"Efficient" Designs

Latin Square Designs

Crossover Designs

6

 

 

TENTATIVE LABORATORY SCHEDULE:

 

WEEK

TOPIC

1

Data Management and Editing using SAS and Introduction to Regression Analysis Using SAS PROC REG

2

Regression Analysis and Diagnostics using SAS PROC REG

3

Matrix Algebra using SAS PROC IML

4

Multiple Regression Analysis and Diagnostics using SAS PROC REG

5

Response Surface Designs and Analysis using SAS PROC RSREG

6

Completely randomized designs using SAS PROC GLM

7

Factorial Design Analysis using SAS PROC GLM/GLIMMIX (Part I)

8

Factorial Design Analysis using SAS PROC GLM GLIMMIX (Part II)

9

Analysis of Covariance using SAS PROC GLM/GLIMMIX - Power Analysis for Experimental Designs

10

Analysis of Randomized Block Designs using SAS PROC MIXED/GLIMMIX

11

Power Analyses

12

Analysis of Split Plot and Repeated Measures Designs using SAS PROC MIXED/GLIMMIX

13

Analysis of Replicated Latin Squares and Crossover designs using SAS PROC MIXED/GLIMMIX

 

The statistical software SAS will form a major component of the data analysis training in this course. Example SAS programs and datasets required for labs will periodically be available on the course MSU ANGEL webpage (www.angel.msu.edu)