|
|
| Message/Author |
|
|
|
|
Hello,
I am trying to get a better understanding of the PLOT3 functionality following GMM.
From what I can tell, the percentages reported in the legend of plots created using PLOT3 match the percentages reported in the "Final class counts and proportions for the latent classes based on estimated posterior probabilities". However, when the factor scores from the model are saved using SAVEDATA=CPROB, the percentage of the sample assigned to each class match those reported in "Final class counts and proportions for the latent classes based on their most likely latent class membership".
My questions:
Is the way that PLOT3 creates the graph inherently based on the posterior probabilities, such that if it were plotted based on most likely latent class membership instead, the graph would look different?
Would it therefore be inaccurate to display the graph produced using PLOT3, but replacing the legend with one displaying the percentages based on most likely latent class membership rather than posterior probabilities?
Many thanks! |
|
|
|
|
Q1: No, the graph would look the same. The percentages are only so you easily see which class is which.
Q2: I would use the percentages based on the estimated model as shown in the output. |
|
|
| Back to top |
|
|
