• Simon Grund
  • Oliver Lüdtke
  • Alexander Robitzsch
Multilevel research is often faced with missing data. Over the past years, powerful methods such as multiple imputation (MI) and maximum likelihood estimation (ML) have become available for the treatment of incomplete data. In this chapter, we provide a general introduction to the problem of missing data, and we discuss the theory and application of these methods as well as their individual strengths and weaknesses. We offer guidance on how ML and MI may be used for an effective treatment of missing values in multilevel research and what role the multilevel structure may play in the treatment of incomplete data. Finally, we provide results from a computer simulation study as well as an empirical example that illustrates the use of these methods in multilevel analyses.
Original languageEnglish
Title of host publicationHandbook for multilevel theory, measurement, and analysis
EditorsStephen E. Humphrey, James M. LeBreton
PublisherAmerican Psychological Association
Publication date2017
StateAccepted/In press - 2017

ID: 801462