TITLE: Strict Invariance for dichotomous items See Millsap & Tein (2004, p.485) for further details DATA: FILE IS example2c.csv; VARIABLE: NAMES ARE u1-u8 g; USEVARIABLES ARE u1-u8; CATEGORICAL ARE u1-u8; GROUPING IS g (1=male 2=female); ! Theta parameterization is strongly recommended for multiple groups ANALYSIS: PARAMETERIZATION = THETA; ESTIMATOR = WLSMV; ! Put the chi-square values of the model that we wish to compare with this model DIFFTEST=weak2c.dif; ! strict vs. weak MODEL: ! Factor loadings are constrained across groups ! except the marker variables, which are fixed. f1 BY u1@1 u2 (f21) u3 (f31) u4 (f41); f2 BY u5@1 u6 (f62) u7 (f72) u8 (f82); ! Thresholds are constrained across groups [u1$1] (t1); [u2$1] (t2); [u3$1] (t3); [u4$1] (t4); [u5$1] (t5); [u6$1] (t6); [u7$1] (t7); [u8$1] (t8); ! Factor variance/covariance are free across groups f1*; f2*; f1 WITH f2*; ! Factor mean of the first group are fixed to zeros [f1@0]; [f2@0]; ! Unique variances of the all groups are fixed as 1 u1@1; u2@1; u3@1; u4@1; u5@1; u6@1; u7@1; u8@1; MODEL female: ! Factor loadings are constrained across groups ! except the marker variables, which are fixed. f1 BY u1@1 u2 (f21) u3 (f31) u4 (f41); f2 BY u5@1 u6 (f62) u7 (f72) u8 (f82); ! Thresholds are constrained across groups [u1$1] (t1); [u2$1] (t2); [u3$1] (t3); [u4$1] (t4); [u5$1] (t5); [u6$1] (t6); [u7$1] (t7); [u8$1] (t8); ! Factor variance/covariance are free across groups f1*; f2*; f1 WITH f2*; ! Factor mean of the second group are free [f1*]; [f2*]; ! Unique variances are fixed to 1 (equal to the first group) u1@1; u2@1; u3@1; u4@1; u5@1; u6@1; u7@1; u8@1; OUTPUT: TECH1; STDYX; ! Save the chi-square values for difference testing with other nested models SAVEDATA: DIFFTEST=strict2c.dif;