|
Message/Author |
|
Wu wenfeng posted on Saturday, December 25, 2010 - 9:06 pm
|
|
|
Dear Mplus team, in multilevel SEM, is there a simple methoed to find out good starting value? if there is, could you please show me how to do? thank you! |
|
|
In most cases, the default starting values of Mplus are sufficient. If there are convergence problems, Mplus automatically provides an input with starting values of the ending values on the incomplete estimation. |
|
Wu wenfeng posted on Sunday, December 26, 2010 - 4:55 pm
|
|
|
sorry! I'm not fully understand your meaning, do you mean Mplus don't need to set starting value by hand? if there is a convergence problem, the only reason is the model design, isn't it? |
|
|
You do not need to give starting values in Mplus. Convergence problems can come for a variety of reasons. See convergence problems in the index of the user's guide for further information. |
|
Back to top |
|
|