Bayesian Shape Invariant Model for Latent Growth Curve with Time-Invariant Covariates

In the attention-deficit hyperactivity disorder (ADHD) study, children are prescribed different stimulant medications. The height measurements recorded longitudinally along with medication time. Differences among patients captured by parameters suggested Superimposition Translation and Rotation (SITAR) model using three subject-specific to estimate their deviation from mean growth curve. this paper, we generalize SITAR in a Bayesian way time-invariant covariates. allows us predict latent factors. Since suffer common disease, they usually exhibit similar pattern, it is natural build nonlinear that shaped invariant. semi-parametric, where population time curve modeled cubic spline. original shape invariant model, motivated epidemiological research on evolution of pubertal heights over time, fits underlying function for age estimates deviations terms size, tempo, velocity maximum likelihood. usefulness illustrated attention deficit study. Further, demonstrated effect medications gender.