MLN/MLWIN MACROS

Below you can find a number of macros for use in the multilevel computer program MLn/MLwiN.
When clicking, you will see the macros; you can then save them using the "Save" command in our web browser.

• At the webpage for the second edition of the Snijders & Bosker textbook, you can find a number of macros for model checking, and other materials (data and setups for other multilevel modeling software). Some background explanation for these macros is also given in Snijders and Berkhof (2007).

• The SPSS include file PreML.inc, written by Jurjen Iedema, for getting data with variable names from SPSS into MLwiN (the versions from January 2006 have two extra features: missing values can be retained and recoded; decimal commas can be converted automatically into decimal points; the new version available from December 6, 2006, corrects an error which occurred for large numbers of variables.).

• In addition to the macros at the website of Snijders & Bosker (see above), several older macros for checking assumptions in MLwiN are available:
  HET1.OBE for testing level-one heteroscedasticity,
  RES1.OBE for calculating level-one OLS residuals,
  DINFL.OBE for calculating level-two influence diagnostics similar to Cook's distance and standardized multivariate level-two residuals.
  These are the macros used for Chapter 9 of the first (1999) edition of Snijders and Bosker (2012), now superseded by Chapter 10 of the second (2012) edition.

• The file srmmacros.zip contains several MLwiN macros, adaptations by Marijtje van Duijn of the original MLn macro SRM.OBE which was a companion to the paper
  The Social Relations Model for Family Data: A Multilevel Approach by Tom A.B. Snijders and David A. Kenny, Journal of Personal Relationships, 6, 1999, 471-486).
  Some of the models discussed in this paper are also treated, more briefly, in Snijders and Bosker (1999), Section 11.3.3.
  This was dropped, perhaps unfortunately, in the second edition. The macros fit the Social Relations Model (SRM) in various specifications, with or without covariates, for one or multiple groups.