Chapter 3: Modeling a complex conceptual theory of population change in the shiras moose: history and recasting as a structural equation model.
Bruce H. Pugesek
NOTE: To download files: right click on file and select "save link
target as"
Models for chapter3. Note: the user will need to modify lines requiring paths to data, covariance matrices, etc.
- comp300.eqs - simulation model for data creation. Note: change the seed number to
create a new randomly generated data set
- com001.dat - raw data generated from EQS simulation comp300.eqs.
- com001.cov - covariance matrix for LISREL runs created with PRE3.job
- com002.dat - raw data generated from EQS simulation comp300.eqs
- com002.cov - covariance matrix for LISREL runs created with PRE3.job
- pre3.job - PRELIS job to convert data matrices to covariance matrices.
Note: disregard warnings, variables are meant to be continuous.
- COO1D.job - Model in figure 3.6, uses com001.cov
- G001e1.job - Model in figure 3.5, uses com001.cov, true model except gamma from Hunt to Mortal not free
- G001e2.job - Model in figure 3.7, uses com001.cov, true model except frees gamma Snow to Disease showing improved fit based on spurious modification indice.
Note: the wrong P value was given in the book. The correct
P value is 0.58
- G002e2.job - Model in figure 3.8, uses com002.cov, model freeing Snow to Disease
to show lack of confirmation
- mmall1.ls8 - LISREL8 program for
analysis of the measurement model. Note: modify CM line for location of
covariance matrix.
|
