Missing values are a ubiquitous problem in quantitative data analysis. Data for a unit of study can be missing either completely (unit-nonresponse), e.g., when a respondent refuses to participate in a survey in principle, or only for certain variables (item-nonresponse), e.g., when a respondent refuses to answer only individual questions. The traditional approach to dealing with item-nonresponse is to restrict the study population to all complete cases (listwise deletion) and consists of simply excluding all observations with missing values from the analysis. While this approach is very easy to implement, it is also very wasteful and often leads to biased estimators. Multiple imputation (MI) is a procedure that can better handle missing values and, provided certain conditions are met, results in efficient and unbiased estimators. For this reason, MI has increasingly emerged as the method of choice for dealing with item-nonresponse in recent years. The goal of the course is to introduce participants to the conceptual and statistical foundations of dealing with missing values and MI and to enable them to use MI independently and flexibly for their own analyses. The focus of the course is on the practical application of MI and the many decisions and challenges that need to be considered. The practical implementation is done with the data analysis software Stata. The seminar is aimed at MA students with experience in quantitative data analysis (esp. regression analysis) and good knowledge of Stata. close