Hi all,
I am trying to use the GBTM model with the traj package in Stata. It’s the first time I use it and I need some clarification. I am trying to analyze the different trajectories in the population under study.I have three time measurements of a parameter. I create a variable (age5, age14, age24) that identify the ages.
gen AGE5=5
gen AGE14=14
gen AGE24=24
I use this command:
traj, var(variable5 variable14 variable24) indep(AGE5 AGE14 AGE24) model(cnorm) min(min value) max(max value) order(1 3 2)
&
traj, var(Y5_FEV1zscoreGLI Y8_FEV1zscoreGLI Y13_FEV1Z Y23_GLIFEV1zScore) indep(AGE*) model(cnorm) min(-8) max(5) order(2 3 1)
Output:
**BIC= -7955.82 (N=6022) BIC= -7949.65 (N=2155) AIC= -7915.60 ll= -7903.60
**BIC= -7984.96 (N=6022) BIC= -7978.79 (N=2155) AIC= -7944.74 ll= -7932.74
I’ve chosen the model with the low BIC value (the second one) and plot the results.
What I found is that the three trajectory are pretty parallel and I would to estimate the slope.
Anyone could help me?
I am trying to use the GBTM model with the traj package in Stata. It’s the first time I use it and I need some clarification. I am trying to analyze the different trajectories in the population under study.I have three time measurements of a parameter. I create a variable (age5, age14, age24) that identify the ages.
gen AGE5=5
gen AGE14=14
gen AGE24=24
I use this command:
traj, var(variable5 variable14 variable24) indep(AGE5 AGE14 AGE24) model(cnorm) min(min value) max(max value) order(1 3 2)
&
traj, var(Y5_FEV1zscoreGLI Y8_FEV1zscoreGLI Y13_FEV1Z Y23_GLIFEV1zScore) indep(AGE*) model(cnorm) min(-8) max(5) order(2 3 1)
Output:
**BIC= -7955.82 (N=6022) BIC= -7949.65 (N=2155) AIC= -7915.60 ll= -7903.60
**BIC= -7984.96 (N=6022) BIC= -7978.79 (N=2155) AIC= -7944.74 ll= -7932.74
I’ve chosen the model with the low BIC value (the second one) and plot the results.
What I found is that the three trajectory are pretty parallel and I would to estimate the slope.
Anyone could help me?
