note File CH16.obe note MLwiN macro for producing results in Chapter 16 note of Snijders & Bosker, 2nd edition. echo 1 logo 0 logo mlbook_16.log clear note Enter data set with 3773 pupils, 4 missing arithmetic score note and 7 missing language score obey mlbook2_b.obe note basic model definitions. pref 0 post 0 mvar 1 "lang_post" "arit_post" addt "cons" iden 2 "pupilnr" iden 3 "schoolnr" setv 2 setv 3 note IGLS yields the ML estimates batch 1 maxi 100 tole 4 method 1 note Table 16.1 start fixe rand like note Preparations for next model. calc c20 = "IQ_verb" * "ses" calc c21 = "mean_iqv" * "mean_ses" name c20 "IQ_verb * ses" c21 "mean iqv * mean ses" note Table 16.2 addt "IQ_verb" addt "ses" addt "mean_iqv" addt "mean_ses" addt "IQ_verb * ses" addt "mean iqv * mean ses" start fixe rand like note The results differ slightly from those in the book, note mainly the parameter estimates for mean_iqv (in the 2nd decimal). note Substantively this is not important. note Perhaps this is due to different approaches to rounding at some stage note for the calculation and storage of the school means. logo 0