We provide an introduction to analysis of complex probability survey samples with a major focus on weighting and calibration methods. We show that the techniques, combined with the empirical likelihood methods, can be used as general inferential tools for missing data analysis and causal inference with observational data. Some recent developments on doubly robust and multiply robust inference and methods for analyzing non-probability survey samples are discussed to illustrate the use of weighting and calibration for these seemingly different but intrinsically connected topics. The workshop is designed for senior undergraduate or graduate students in statistics, biostatistics or related fields and young researchers interested in the topics.