Multi-level Analysis is used for analyzing nested data. It is for studying the effect of multi-level predicted variables that influence dependent variables. It can check the cross-level effect by taking the effect size in the lower level to be the dependent variable for the next level. There are four sub-models of multi-level analysis: 1) one way random effect ANOVA model or null model, 2) means as outcome regression model, 3) random-coefficient regression model and 4) intercepts and slopes as outcome model. It is more valid than the regression analysis because the maximum likelihood is used for estimating the coefficient, which is different from using the least square method in regression analysis. In addition, the result shows that multi-level analysis can explain variance of the dependent variable more extensively.