Spring: CSI 991
Section 003
Summer: CSI 991
Section X01
Contacts:
csutton@gmu.edu
jgentle@gmu.edu
The seminar will be conducted in the form a workshop.
Participants are encouraged to put
R
on their own laptops, and to bring them
to the seminars.
R has many different packages. Which are useful for the main (and, perhaps, non-mainstream)
statistical techniques used in analysis of latent variable models or, more
generally, in the social sciences?
The main reference text is Latent Variable Models, Fourth edition, by John C. Loehlin. The book is quite elementary, but we hope the book will point us to some interesting things to discuss.
Clarifications of some issues in Chapter 1.
Factor analysis (one of the earliest instances of the concept of "latent" variables) and its relation to path models.
Forge ahead into Chapter 2.
Jonathan Lisic's R code that uses the
OpenMX package.
Arun's R code for Fig-2.5. This uses the OpenMX package, but it doesn't produce the same results as SAS. We need to figure this out.
Jonathan's R code. Uses OpenMX package for fitting and uses Newton's method and the derived discrepancy function for ML estimation.
Here's an interesting website that uses the sem package to work some of the examples in the text.
Nick's R code that uses the sem package for Loehlin problems Table 2-5 and Table 2-12.
Jonathan's MLE and LR results for SEs.
Dan's R code that uses the sem package for
models in Loehlin pages 97, 98.
Arun's R code for Figure 3.1 and
for Figure 3.7.
Interesting article Jonathan found
Review and preview.
Factor analysis in R.
Factor analysis in R.
More on factor analysis.
Xun Wang's example of latent variables in inflation mesaurement.
More on factor analysis: factanal and SEM-R (Hatice).
More on factor analysis: review of factanal and SEM-R.
More on factor analysis (Hatice).