Abstract:
|
The aim was to provide information and data application of key models utilized in longitudinal data analysis such as the generalized linear mixed model (LMM) and the generalized estimating equations (GEE) model. Two data applications were used, which highlighted the stories of these data. The first data application uses a National Youth Survey (from the United States of America), in which 80 participants were tested annually from 11 to 15 years old, to assess their tolerance of deviant behavior. The results from the LMM indicated that age was significantly associated with tolerance of deviant behavior (? ± s.e = 0.13 ± 0.04, p=0.00). There was also a significant association between exposure to bad behavior and tolerance of deviant behavior (? ± s.e = 0.61± 0.22, p=0.02). The second data application examined a randomized clinical trial of patients with epilepsy. The GEE models indicated that age, treatment and baseline counts of epileptic seizures had a significant association with the reduction in the number of epileptic seizures. Visualizations and adjustments for correlations were conducted on the data. The R software package was utilized for the statistical analyses.
|