Applied Multiple Imputation

Advantages, pitfalls, new developments and applications in R

Springer R code for chapters 5 and 6 R package countimp R package imputerobust

Book “Applied multiple imputation”

Chapters

1. Introduction and Basic Concepts
2. Missing Data Mechanism and Ignorability
3. Missing Data Methods
4. Multiple Imputation: Theory
5. Multiple Imputation: Application
6. Multiple Imputation: New Developments

This book explores missing data techniques and provides a detailed and easy-to-read introduction to multiple imputation, covering the theoretical aspects of the topic and offering hands-on help with the implementation. It discusses the pros and cons of various techniques and concepts, including multiple imputation quality diagnostics, an important topic for practitioners. It also presents current research and new, practically relevant developments in the field, and demonstrates the use of recent multiple imputation techniques designed for situations where distributional assumptions of the classical multiple imputation solutions are violated. In addition, the book features numerous practical tutorials for widely used R software packages to generate multiple imputations (norm, pan and mice). The provided R code and data sets allow readers to reproduce all the examples and enhance their understanding of the procedures. This book is intended for social and health scientists and other quantitative researchers who analyze incompletely observed data sets, as well as master’s and PhD students with a sound basic knowledge of statistics.

R package countimp - Multiple imputation of incomplete count data

Description

Count data require special analysis and imputation techniques. countimp provides addon-functions to the popular multiple imputation package mice, and extends mice’s functionality. Imputation models include: - the standard Poisson model,
- the quasi-poisson, and negative binomial model for overdispersed count data, - zero-inflation and hurdle models for count data with an excess of zero counts - multilevel versions of these models.

Download / Installation

• countimp version 2 is available from github under the GPL-3 licence.
• The latest version can be installed from GitHub using function install_git from package devtools:

# install.packages('remotes')
remotes::install_github('kkleinke/countimp')

• countimp version 2 is described in detail in chapter 6 of our book.
• The user’s guide with some applications and Monte Carlo Simulations is available here.

Our missing data research

Featured free and open access articles

• How to and how not to impute incomplete count data - In: Proceedings from the 9th European Congress of Methodology

• Quantile regression-based multiple imputation of missing values – An evaluation and application to corporal punishment data

• Estimation of partially observed non-linear terms in a multilevel model: An evaluation of the robustness of ad hoc and state-of-the-art missing data methods

• Multiple imputation by predictive mean matching when sample size is small

• Multiple imputation of zero-inflated and overdispersed multilevel count data

• Multiple imputation of incomplete ordinary and overdispersed count data

All MI-related publications

Kleinke, K. (2017). Multiple imputation under violated distributional assumptions – a systematic evaluation of the assumed robustness of predictive mean matching. Journal of Educational and Behavioral Statistics, 42(4), 371–404. https://doi.org/10.3102/1076998616687084

Kleinke, K. (2018). Multiple imputation by predictive mean matching when sample size is small. Methodology, 14(1), 3–15. https://doi.org/10.1027/1614-2241/a000141

Kleinke, K. (2021). Estimation of partially observed non-linear terms in a multilevel model: An evaluation of the robustness of ad hoc and state-of-the-art missing data methods. Psychological Test and Assessment Modeling, 63(3), 432–455. https://www.psychologie-aktuell.com/fileadmin/Redaktion/Journale/ptam-2021-3/PTAM__3-2021_6_kor.pdf

Kleinke, K., Fritsch, M., Stemmler, M., & Lösel, F. (2024). Multiple imputation of longitudinal data – a comparison of robust imputation methods regarding sample size requirements, with an application to corporal punishment data. In M. Stemmler, W. Wiedermann, & F. Huang (Eds.), Dependent data in social sciecnes research – forms, issues and methods of analysis (second edition) (pp. 565–588). Springer Nature. https://doi.org/10.1007/978-3-031-56318-8_23

Kleinke, K., Fritsch, M., Stemmler, M., Reinecke, J., & Lösel, F. (2021). Quantile regression-based multiple imputation of missing values – an evaluation and application to corporal punishment data. Methodology, 17(3), 205–230. https://doi.org/https://doi.org/10.5964/meth.2317

Kleinke, K., Jong, R. de, Spiess, M., & Reinecke, J. (2011). Multiple imputation of incomplete ordinary and overdispersed count data [Technical Report]. University of Bielefeld, Faculty of Sociology. https://kkleinke.de/static/pdf/2011_technical.pdf

Kleinke, K., & Reinecke, J. (2013a). countimp 1.0 – A multiple imputation package for incomplete count data (Technical Report 01-2013). University of Bielefeld, Faculty of Sociology. https://doi.org/10.13140/RG.2.1.3889.3286

Kleinke, K., & Reinecke, J. (2013b). Multiple imputation of incomplete zero-inflated count data. Statistica Neerlandica, 67(3), 311–336. https://doi.org/10.1111/stan.12009

Kleinke, K., & Reinecke, J. (2014). Multiple imputation of zero-inflated and overdispersed multilevel count data [Technical Report]. University of Bielefeld, Faculty of Sociology & Centre for Statistics. https://kkleinke.de/static/pdf/2014_technical.pdf

Kleinke, K., & Reinecke, J. (2015a). Multiple imputation of multilevel count data. In U. Engel, B. Jann, P. Lynn, A. Scherpenzeel, & P. Sturgis (Eds.), Improving Survey Methods: Lessons from Recent Research (pp. 381–396). Routledge, Taylor & Francis.

Kleinke, K., & Reinecke, J. (2015b). Multiple imputation of overdispersed multilevel count data. In U. Engel (Ed.), Survey Measurements. Techniques, Data Quality and Sources of Error (pp. 209–226). Campus/The University of Chicago Press.

Kleinke, K., & Reinecke, J. (2019). Countimp version 2 – A multiple imputation package for incomplete count data [Technical Report]. University of Siegen, Department of Education Studies; Psychology. https://kkleinke.de/countimp

Kleinke, K., & Reinecke, J. (2022). How to and how not to impute incomplete count data. In A. Hernández & I. Tomás (Eds.), Proceedings from the 9th European Congress of Methodology (pp. 86–92). Universitat de València. https://doi.org/10.7203/PUV-OA-438-5

Kleinke, K., & Reinecke, J. (2024). Multiple imputation of incomplete panel data based on a piecewise growth curve model – an evaluation and application to juvenile delinquency data. In M. Stemmler, W. Wiedermann, & F. Huang (Eds.), Dependent data in social sciecnes research – forms, issues and methods of analysis (second edition) (pp. 589–615). Springer Nature. https://doi.org/10.1007/978-3-031-56318-8_24

Kleinke, K., Reinecke, J., Salfrán, D., & Spiess, M. (2020). Applied multiple imputation. Advantages, pitfalls, new developments and applications in R. Springer Nature.

Kleinke, K., Reinecke, J., & Weins, C. (2021). The development of delinquency during adolescence: A comparison of missing data techniques revisited. Quality & Quantity, 55(3), 877–895. https://doi.org/10.1007/s11135-020-01030-5

Kleinke, K., Stemmler, M., Reinecke, J., & Lösel, F. (2011). Efficient ways to impute incomplete panel data. Advances in Statistical Analysis, 95, 351–373. https://doi.org/10.1007/s10182-011-0179-9

Spiess, M., Kleinke, K., & Reinecke, J. (2021). Proper multiple imputation of clustered or panel data. In P. Lynn (Ed.), Advances in longitudinal survey methodology (pp. 424–446). John Wiley & Sons.