|
|
Markov model with measurement error |
|
Message/Author |
|
Chris Weber posted on Tuesday, April 15, 2014 - 7:12 pm
|
|
|
I am trying to estimate a markov model using a single three category variable measured at 6 time points. get the following error for every indicator for every response pattern: * THRESHOLD 2 OF CLASS INDICATOR PI2 FOR PATTERN 1 1 1 1 1 1 ... As well as an error regarding the standard errors and a problem of non-identification.I constrained the indicator thresholds to be the same across time points. Here are the relevant parts of the model: %OVERALL% [c1#1] (a1); [c1#2] (a2); ..... [c6#1] (a1); [c6#2] (a2); c6#1 on c5#1 (b11); !Homogenous transitions c6#2 on c5#1 (b21); c6#1 on c5#2 (b12); c6#2 on c5#2 (b22); .... c2#1 on c1#1 (b11); c2#2 on c1#1 (b21); c2#1 on c1#2 (b12); c2#2 on c1#2 (b22); model c1: %c1#1% [pi1$1*1](a); [pi1$2*2](b); %c1#2% [pi1$1*-2](c); [pi1$2*2](d); !Ind %c1#3% [pi1$1*-2](e); [pi1$2*-1](f); !Rep model c2: %c2#1% [pi1$1*1](a); [pi1$2*2](b); .... Any thoughts on what I might be doing incorrectly? Thanks, Chris |
|
|
Don't see anything - please send output and license number to Support. |
|
Back to top |
|
|