|
|
Merging Transition Probabilities in LTA |
|
Message/Author |
|
|
Given I have three classes A, B, and C at 2 time-points, is there a way to merge B and C to analyze the combined transitions from A to BC or from BC to A and disregard changes within BC? |
|
|
Yes. From the estimated model you obtain the probabilities for the 3 x 3 table corresponding to the 3 classes at the two time points. From those you can add up any combination of classes. |
|
Back to top |
|
|