Data Analysis Of Multiple Repeated Measures Of Children's Behavior

Longitudinal data are repeated measurements over time of each individual, a combination of cross section data and time scries data, reflecting the longitudinal study. Longitudinal data are repeated measurements.so the same individual's measurements at different times must be related. arid the measurements of different individuals are independent.When the response variable is considering the same individual's one indicator, you can use longitudinal data models to describe,which are mature and widely used. however, if considering multiple indicators,the data are more complex, and the researchers not only concerned with how a single response variable changes over time, but also taking into account the relationship between variables and variables. In this paper, dealing longitudinal data with random effects models.First introduced random effects models Y = Xβ+Zb+e, parameter estimation of random effects models, and parameter estimation of random effects model estimates are complicated, with fixed parameterβis estimated, variance parameterθis estimated, and random effect b is estimated. The data sets discussed in this paper are more than a hundred children's multiple indicators's repeated measurement data. The children come from two kindergartens in Changchun City. This paper firstly considers a single indicator using random effect model, in order to study relationship of indicators, so useing Pairwise method to establish random effects model with multiple indicators.