IMPORTANT: Changes coming in gimme 8.3

[Katie Gates]

Hey all, 

We've been hearing about sub-optimal solutions where individual-level models either (a) do not meet 'good fit' according to fit indices or (b) have individual-level paths that aren't significant. 

These issues are inter-related... when we prune non-signfiicant paths, the models don't meet good fit, and sometimes there aren't any paths that can be added that resolve the situation and converge. 

We're improving gimme by:

• having better handling of these situations so that exhaustive searches are done to obtain good fit 
• having better identifications of when models do not converge normally (i.e., "last known convergence" or "no additional significant paths" for when the model fit is not good). 

If you're currently doing work that will use individual-level path information, I suggest you wait until the new version is out. I'll email when completed. 

Best,

Katie