The application of mixed model in dependent data analysis

Aiming at the difficulty of dependent data analysis in data measurement field and the application limitation of traditional variance analysis for the processing of correlation data, the principle of mixed model and the corresponding design for the unequal variance data sets were discussed. The fixed effects and random effects were combined with the mixed model which realized the analysis of dependent data. Simulation experiment testified the validity of the model. However, the optimized design for data reconstruction still need further research.