This thesis has three chapters on the effect of social ties under different topics in China. The first chapter tries to measure the peer effect in user-generated contents. By using the novel data from the review website Dianping and the framework of Bayesian MCMC to structuralize the process of network formation, the chapter confirms the importance of understanding the connection between customers, which brings a significant difference in the scale of effects. The second chapter focuses on the relationship between migration and the educational expectation for the child of migrants. By using the oral fluency of Mandarin as the instrument variable, the chapter identifies a strong relationship between migrated fathers. However, from marginal treatment effects, it shows that the topic still requires further explorations. The third chapter introduces the sequential analysis into unique data for Chinese civilians and military leaders. Various measures have been evaluated.