title: Data from page 162 in EQS program. Input based on Appendix 6.1 page 162. Figure 6.1 page 77, first 3 groups only Output. Table 6.2 (cohort sequential columns), page 79. NOTE. Correlation is standardized value of int with slope. Also see text at top of page 79. Also Table 6.3 (cohort sequential column), computed based on intercept and slope values, e.g. age 12: .310 + 0*.384 = .31 age 13: .310 + 1*.384 = .694 age 14: .310 + 2*.384 = 1.078 age 15: .310 + 3*.384 = 1.462 age 16: .310 + 4*.384 = 1.846 data: file = ch6cs.dat; format = free; type = means covariance; ngroups = 3; nobservations = 248 250 231; variable: names = y1 y2 y3; analysis: type = general mgroup meanstructure; model: ! this is the overall model int by y1@1 y2@1 y3@1; slp by y1@0 y2@1 y3@1; ! overidden for each group [y1-y3@0]; ! Note 1 [int] (mint); ! Note 1 [slp] (mslp); ! Note 1 int with slp (intslp); ! Note 1 int (vint); ! Note 1 slp (vslp); ! Note 1 y1 (e14); ! overidden for each group y2 (e15); ! overidden for each group y3 (e16); ! overidden for each group model g1: ! this is the model for group 1 int by y1@1 y2@1 y3@1; slp by y1@0 ; ! age 12 slp by y2@1 ; ! age 13 slp by y3@2 ; ! age 14 [y1-y3@0]; ! Note 1 [int] (mint); ! Note 1 [slp] (mslp); ! Note 1 int with slp (intslp); ! Note 1 int (vint); ! Note 1 slp (vslp); ! Note 1 y1 (e12); y2 (e13); y3 (e14); model g2: ! this is the model for group 2 int by y1@1 y2@1 y3@1; slp by y1@1; ! age 13 slp by y2@2; ! age 14 slp by y3@3; ! age 15 [y1-y3@0]; ! Note 1 [int] (mint); ! Note 1 [slp] (mslp); ! Note 1 int with slp (intslp); ! Note 1 int (vint); ! Note 1 slp (vslp); ! Note 1 y1 (e13); y2 (e14); y3 (e15); model g3: ! this is the model for group 3 int by y1@1 y2@1 y3@1; slp by y1@2; ! age 14 slp by y2@3; ! age 15 slp by y3@4; ! age 16 [y1-y3@0]; ! Note 1 [int] (mint); ! Note 1 [slp] (mslp); ! Note 1 int with slp (intslp); ! Note 1 int (vint); ! Note 1 slp (vslp); ! Note 1 y1 (e14); y2 (e15); y3 (e16); output: standardized tech1; ! Note 1 ! Note that these constraints are repeated across all ! three groups. We may have been able to just include this ! in the overall model and omitted them from each ! of the group specific models.