|
|
LTA,distal outcome, and covariates |
|
Message/Author |
|
Dylan John posted on Wednesday, September 13, 2017 - 8:31 am
|
|
|
I am currently running an LTA between two timepoints to look at the association of the transition groups with my distal outcome (y) of suicide attempts (dichotomous). I have two questions that pertain to my covariates (z1,z2,z3) and how to use them. 1. Does this syntax look correct (only showing 1 transition group): model: %overall% x2 on x1; %x1#1.x2#1% [c_wave1#1@25 c_wave1#2@0 c_wave1#3@0]; [c_wave2#1@25 c_wave2#2@0 c_wave2#3@0]; y on z1; y on z2; y on z3; [y$1] (thr11); 2. Some of my covariates are categorical, with k levels. Does this mean that I need to create k-1 dichotomous dummy variables. For example, race: white, black, asian, other. dummy variable 1: 1=black 0=white dummy variable 2: 1=asian 0=white dummy variable 3: 1=other 0=white |
|
|
Q1: The y statements look correct, but I don't know what this means: [c_wave1#1@25 c_wave1#2@0 c_wave1#3@0]; [c_wave2#1@25 c_wave2#2@0 c_wave2#3@0]; Q2: Yes. |
|
Back to top |
|
|