While planning the SEM Mastering, getting guidance from Jian, and looking for an industrial example, I spoke with Laura Castro-Schilo @LauraCS . She shared some details about 15.1 SEM. Also, a reminder that she presented at 2019 Tucson Discovery. Laura wrote:
Hi everyone,
I’ll put the SEM 15.1 improvements in a wiki page shortly but here they are for now. I tried to limit the 15.1 features I included in the blog post Gail mentioned, so the list below is more complete:
- There’s a new Model Status tab that includes dynamic checks for identification of the model and lists useful descriptive data for the model (note this is in the process of being patented as we’re the only ones who have this feature –as far as we know :smiling_face_with_smiling_eyes:).
- The path diagram includes options for moving, selecting, and modifying elements.
- Model estimates are now illustrated in the path diagram.
- The Fit Indices table has been added to the report which includes a number of new values that characterize the adequacy of the model fit.
- AICc weights were added to the Model Comparison table as an additional tool to aid model selection.
- R^2 values are provided for all endogenous variables in the model so analysts can assess the amount of variance explained in any outcome variable
- There’s a huge performance improvement to the estimation algorithm, particularly when missing data are present.
