Longitudinal data exploration of upper respiratory tract infection disease in Bandung city

Longitudinal data have a complex structure. Unlike cross-sectional data, there are more associated features with longitudinal data. Associated features to longitudinal data including the heterogeneity between subjects, the serial correlation and the characteristics of changes over time. Therefore exploring longitudinal data is more complicated than that of cross-sectional data. Various methods were developed by many researchers to explore longitudinal data. In general, there are at least three things that could be described from longitudinal data: mean structure, variance structure, and correlation structure. Visual methods and subject-specific profile were studied for such datasets to detect these structures. Methods were applied to data sets from upper respiratory tract infection data in Bandung City. From the result, we can describe the growth patterns at both group and individual level and then selecting suitable statistical models which contain within and between groups variations.