Use the LSAY data of previous assignments to do multilevel
latent variable modeling. LSAY data are hierarchical with students
nested within schools. Ignore for this
assignment that students move between schools, particularly in higher grades,
and assume for simplicity that school membership is captured by the variable
schcode = 7th grade school code
Do two-level analysis, choosing between factor analysis, structural equation modeling, latent class analysis, growth modeling, and growth mixture modeling. Choose between sets of variables such as math attitudes, math achievement, and the push variables:
parapsh =
parent academic push (7th grade)
parcpsh =
parent college push (7th grade)
parmpsh =
parent math push (7th grade)
peerapsh =
peer academic push (7th grade)
peermpsh = peer math push (7th grade)
Compare the results to those obtained when the clustering of students within schools is ignored and simple random sampling is assumed. Interpret the two-level solution.