STA 2201

Methods of Applied Statistics II

University of Toronto, Winter/Spring 2011

Friday 2:10-5:00 p.m. SS 2110

This page is under construction, and will be updated frequently throughout the course.

Homework assignment 7 is cleaned up and re-posted the evening of April fools' day. This version is clearer in several places, and you should use it rather than the original.

The two last sets of overheads are posted too.

• Plan of the Course
• Homework Assignments
  1. Assignment 1: Your written answers and printouts are due at the beginning of class Friday on Jan 21. You may be asked to present your solution to the class before handing it in.
  2. Assignment 2: This is the final version, I hope! Your written answers and printouts are due at the beginning of class Friday on Jan 28. You may be asked to present your solution to the class before handing it in.
  3. Assignment 3: Your written answers and printouts are due at the beginning of class Friday on Feb 18. You may be asked to present your solution to the class before handing it in.
  4. Assignment 4: Your written answers and printouts are due at the beginning of class Friday on March 11th. You may be asked to present your solution to the class before handing it in.
  5. Assignment 5: Your written answers and printouts are due at the beginning of class Friday on March 18. You may be asked to present your solution to the class before handing it in.
  6. Assignment 6: Your written answers and printouts are due at the beginning of class Friday on April 1st. You may be asked to present your solution to the class before handing it in.
  7. Assignment 7: Your written answers and printouts are due at the beginning of class Friday on April 8th. You may be asked to present your solution to the class before handing it in.
• Handouts and Overheads: You may want to print these and bring to class so you can write notes on them.
  1. A little Large Sample Theory
  2. Convergence of random vectors: I did not have enough copies of this handout in class. The material is also in the large sample slide show, but this one-page version may be more convenient.
  3. Maximum Likelihood
  4. Wald and Score tests
  5. More Likelihood with R
  6. Logistic Regression
  7. Birth WeightLogistic Regression Example
  8. Stepwise logistic regression example
  9. Poisson Regression
  10. Language Development Poisson regression example
  11. Generalized Linear Models: Part One
  12. Generalized Linear Models Part Two: Iteratively re-weighted least squares
  13. Sample size selection with power: Normal linear models
  14. Matrix method: A continuation of the handout above.
  15. Power estimation after the data have been collected
  16. Sample variation and popyulation variation methods: Alternatives for the normal linear model
  17. Power of Large Sample Chi-square Tests: Part 1
  18. Power of Large Sample Chi-square Tests: Part 2 (Illustration with R)
  19. Power by simulation
      • Simulation as Estimation
      • A Toy Example
      • Monte Carlo Sample Size for estimating power
      • Power for regression with random independent variables
      • Using logistic regression to zero in on the required sample size.
      • Regression with measurement error
      • BMI Study: Good and bad analysis
• Readings: Background material, not absolutely required, but possibly useful.
  • Introduction to R: A very nice 100-page PDF document. Highly recommended if you are going to be using R for serious work.
  • Small text on large sample theory: More than you need for this course, but nice to have.
  • Generalized Linear Models: A well known text.
• R Statistical Software
  • Official R website, including how to obtain a free copy of R to put on your home computer.
  • More R Links
• Data files
  • bweight.data: Birth weight data used to illustrate logistic regression
  • logreg.data: Logistic regression data used in Homework 3 and the iteratively re-weighted least squares handout
  • georeg.data: Geometric regression data used in Homework 4
  • Assignment6.data: Five multivariate normal variables, n=50.