Session 4: Residual Analysis, and Learning from Differences

• Analysis of variance
  • [Testing for differences]
  • Analysis of variance
• Residual analysis
  • Recognizing and dealing with outliers
  • [Logarithmic transformations: NBA salaries]
• Summary: Regression Modeling


• [Chi-square tests]
  • Notes
  • A test of independence

---

You'll receive an email (similar to that sent out with the third sample exam below) containing your exam dataset, a description of why someone collected the data and a discussion of some of the issues of interest to them, and (perhaps) some totally-optional background reading. The email will be sent out on 11/24 (EMP 69) or 12/1 (EMP 70). The sample email below lays out exam-related rules and procedures, which you'll see again in your own email. Both classes will be analyzing the same dataset, and taking the same exam.

---

• Sample pre-exam posting:

  • Optional background readings on Harrah's Entertainment Inc.
  • Caligula's Castle dataset and description
• Exam questions
• Exam answers

  ---

• Sample pre-exam posting:

  • Optional background reading on Microsoft vs. Linux
  • Mantis Technologies dataset and description
    
• Exam questions (as a Microsoft Word document).

• Exam answers

  ---

• Sample pre-exam email:

  • Email sent out with dataset 8 days before exam
  • US: One-on-One dataset and description (with link to optional background reading)
    
• Exam questions (as a Microsoft Word document).

• Exam answers

Note: About half of each exam covers basic material, and the other half consists of questions of a more challenging nature. I grade accordingly, with the "pass" line at around 50%, and the rest of the questions used to separate the high passes from the passes. (So don't panic if you aren't sure how to do a few of the problems.)