|
|
Latent Class Growth Model |
|
Message/Author |
|
|
I am working on a latent class growth model to estimate aging trajectories using self reported health status (ordinal), functional limitations (count), and mortality (binary). My analysis is as follows: Analysis: TYPE = MIXTURE complex; PROCESSORS=12 ; START = 500 50; I got the following warning in my output: 306 perturbed starting value run(s) did not converge or were rejected in the third stage. Final stage loglikelihood values at local maxima, seeds, and initial stage start numbers: -118989.819 unperturbed 0 -119089.084 945065 255 -119089.084 345070 114 -119177.591 723775 97 -119358.628 217744 326 -119448.669 685657 69 -120243.330 829540 324 -120250.540 173191 422 -120871.402 437181 135 -120972.982 311214 64 -121937.917 568405 233 39 perturbed starting value run(s) did not converge or were rejected in the third stage. Does this mean that I have convergence issues? Is there anything I can do to improve the model fit or should I try a 3 class instead of a 4 class model? |
|
|
Q1: Yes Q2: Yes. |
|
Back to top |
|
|