Statistical Methods, MT2011

This webpage gives the material for the presentations in this course by Professor Tom Snijders, Tuesdays 10-11 am in weeks 4, 6, 8 of Michaelmas Term.

For the part taught in Hilary Term, go to Statistical Methods_HT2012_Snijders.htm.

Principles

• Week 1: Simulation-based inference - prof. Ripley.

• Week 4: Model checking

  Lecture notes: Model Checking (For printing it may be good to choose "Page scaling - Multiple pages per sheet - 2 pages - Auto-rotate pages).

  
  Script: LM_ResidualChecking.r

• Week 5: Model selection - prof. Ripley.

• Week 6: Robustness

  Lecture notes: Robustness
  Script: LM_Robustness.r for a do-it-yourself simulation study.
  A totally accidental exercise question, unrelated to residuals or robustness: The Metro of November 4 had this alarming news item. Is the rise in road deaths statistically significant? (You should be able to do this in your head - not even a back of the envelope is required.)

• Week 7: Splines etc - prof. Ripley.

• Week 8: Missing data

  Lecture notes: Missing Data
  Scripts: misdat.r and missingdat.r
  Data set used with the second script: child_beh.txt.
  Literature:
  • Joseph L. Schafer and John W. Graham Missing Data: Our View of the State of the Art. Psychological Methods 7 (2002), 147-177.
    (The parts in this paper about methods under MNAR, about weighting, and about software are not examinable.)