Further Statistical Methods -HT05

Practical - week 3

MIM and graphical models.

  1. Start MIM
  2. Download this file from a study of housing types in Denmark and look at the file. The data is also available in MASS and described there. Four variables. h: type of housing (4 different types); c: contact with neighbours (high/low), i: influence on decisions (low, medium, high); and s: satisfaction (low, medium, high).
  3. Read the file into MIM
  4. Specify the saturated model and display its graph
  5. Which variables are ordinal?
  6. Specify these as ordinal in MIM (use e.g. ordinal i on command line)
  7. Test all conditional independences of any pair of variables, given the others. Both by using a standard LR test and the appropriate test which takes into account ordinality of some of the variables. Note that it may matter which order the variables are mentioned. The variable mentioned second is the "response". Check also whether the asymptotic results seem to conform with Monte-Carlo p-values. Decompose the tests according to the values of the conditional variables (use eg. testdelete hi Z).
  8. Use also a stepwise procedure (stepwise W takes ordinality into account)
  9. Argue that, in light of the above, the graphical model with generator chs,cis, i.e. a single conditional independence of hi given cs, is a reasonable model.
  10. Download also this file containing data from a study of attitudes towards abortions in the USA. The individuals in the study are cross-classified according to four criteria. r: race, (white or non-white), s: sex (male or female), o: opinion legalising abortion (for, against, undecided), and a: age (6 age groups).
  11. Read also the file into MIM (use "New" from the menu to clear everything away).
  12. Specify the saturated model and display its graph.
  13. Specify the variable age as ordinal in MIM. Note that this may or may not be justifiable.
  14. Repeat the analysis for these data and comment on the results. Note here that the ordinality of age and type of test used can be crucial.

 

Back to course overview.

Last updated: Monday, 30 January 2006 11:25.  Steffen L. Lauritzen