Xiao Lo posted on Wednesday, November 09, 2016 - 3:03 pm
I'm modeling a LCGA for a binary variable in 3 time points (0, 1). SEX is used as a predictor (0, 1). The Mplus output looks very confusing: (1) I don't know how to understand the plot (estimated probabilities for categorical latent variable as a function of SEX), as the y axis is probability but what does it mean if the x axis is SEX (from 0 - 1)? (2) In the output, C1 on SEX and C2 on SEX have the same estimates (+1.6 and -1.6). Is it normal? (3) There is only the Odds Ratio of C1 on SEX but no OR of C2 on SEX can be found. I also do not know how to explain this OR.
I'm new to LCGA, hope you can guide me to understand these issues. Thanks.