STA 312F 2007

Topics in Statistics: Structural Equation Models

University of Toronto at Mississauga, Fall 2007

Lecture: Monday and Wednesday 11:10-12:00 p.m., Room 142, North Building

Tutorial: Friday 11:10-12:00 p.m., Room 174, North Building

(Quizzes are in tutorial)

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

The last week (Monday and Wednesday) will be review. The exam will have a Formula Sheet. You should download it and use it when you practice the homework problems.

Quiz 11 and the solution are posted below.

The last tutorial (on Dec. 7) will be held in Christine's office, SE2038E. Christine will be available in her office from 9 to 12 on that day.

I will have extra office hours the morning of the exam. Exact time will be announced and posted here.

Scores on Quiz 7 were too low. The quiz will be discarded, and quiz average will be based on one fewer quiz. However, the question was legitimate, and will appear on the final exam.

• Course Outline
• Documentation for SAS proc calis
• Homework Assignments
  1. Assignment 1: Quiz on Friday Sept. 21st in tutorial.
  2. Assignment 2: Quiz on Friday Sept. 28th in tutorial.
  3. Assignment 3: Quiz on Friday Oct. 5th in tutorial.
  4. Assignment 4: Quiz on Friday Oct. 12th in tutorial. For this assignment, it will help to read Pages 4 through 9 in the document on Regression with Measurement Error.
  5. Assignment 5: Quiz on Friday Oct. 19th in tutorial.
  6. Assignment 6: Quiz on Friday Oct. 26th in tutorial.
  7. Assignment 7: Quiz on Friday Nov. 2nd in tutorial.
  8. Assignment 8: Quiz in lecture on Monday Nov. 12th at the beginning of class.
  9. Assignment 9: Quiz on Friday Nov. 16th in tutorial.
  10. Assignment 10: Quiz on Friday Nov. 23d in tutorial.
  11. Assignment 11: Quiz on Friday Nov. 30th in tutorial.
• Quizzes: Solutions are a gift from Christine
  1. Quiz 1
     • Questions
     • Solutions
  2. Quiz 2
     • Questions
     • Solutions
  3. Quiz 3
     • Questions
     • Solutions
  4. Quiz 4
     • Questions
     • Solutions
  5. Quiz 5
     • Questions
     • Solutions
  6. Quiz 6
     • Questions
     • Solutions
  7. Quiz 7
     • Questions
     • Solutions
  8. Quiz 9
     • Questions
     • Solutions
  9. Quiz 10
     • Questions
     • Solutions
  10. Quiz 11
      • Questions
      • Solutions
• Computer handouts: Please print and bring to class. You may want to write notes on them.
  1. Handout 1: Reading data and doing basic descriptive statistics with SAS.
  2. Handout 2: Instrumental variables regression on the Poverty data, using SAS proc calis.
  3. Handout 3: Ordinary Least squares regression on BMI data, ignoring measurement error.
  4. Handout 4: Double measurement latent variables regression on BMI data.
  5. Handout 5: Like Handout 4, except making the proc calis code more compact. The program bmi3.sas is a good model for double measurement regression.
• Readings (Textbook substitutes): More to follow
  • Random matrices: Background material on matrix-valued random variables.
  • The Multivariate Normal distribution.
  • Regression with Measurement Error This is a first draft, but please download and print it anyway. This version does not allow for measurement error in the dependent variables, but it should. We will fix this in lecture and homework, and then I will post a second draft.
  • Regression with Measurement E\rror: Draft Two
  • Large-sample Likelihood Ratio Tests (no example yet)
  • Introduction to unix and SAS
• Data sets and SAS programs from lecture and homework.
• Computing with unix. This is for another course, but it may be useful. Just ignore the parts about cquest. You will be using tuzo.

If you are connecting from a computer not located on campus, you may need to download Adobe Acrobat Reader (free software) in order to see mathematical text, including the assignments.