Missing Values: What is Really Missing with Missing Values

Talk

WHEN June 23, 2021
3-4:30pm WHERE Zoom Webinar WHO Statistical Consulting Laboratory (StatsLab) Open to the Public

In the talk, Christian Heumann will first try to motivate why simply ignoring the problem of missing data is a bad option if a substantial amount of data is missing. Imputation has been proposed a long time ago as a method to treat the problem of missing data adequately. Since single imputation has also severe disadvantages, the standard is (or should be) nowadays multiple imputations. Here, several options can be chosen by the imputer. Finally, Heumann will discuss how to bootstrap and multiple imputations can be combined for inference purposes, e.g. standard errors or confidence intervals.

Email nyuad.statslab@nyu.edu for more details.

Christian Heumann
Ludwig-Maximilians-Universität München
Department of Statistics
Ludwigstraße 33, 80539 Munich
Email: christian.heumann@stat.uni-muenchen.de
http://www.misoda.statistik.uni-muenchen.de/index.html

I’m a professor at the department of statistics of the LMU. My main job is teaching statistics in our Bachelor and Master programs and also as a service in the Bachelor program for the faculty of business administration and economics.

I am (co-)author of a number of (mainly) introductory books to Statistics. My main (statistical) research topics have been likelihood based methods for longitudinal categorical data (PhD thesis), missing data and the combination of imputation and bootstrap methods to inference problems with missing data.
- Linear Models and Generalizations - Least Squares and Alternatives (monograph). C.R. Rao, H. Toutenburg, Shalabh and C. Heumann, Springer, 2008.
- Schomaker M and Heumann C (2018). Bootstrap Inference When Using Multiple Imputation. Statistics in Medicine.
- Faisal MZ, Heumann C (2018). Multiple imputation with sequential penalized regression. Statistical methods in medical research. https://doi.org/10.1177%2F0962280218755574
- Schomaker M, Hogger S, Johnson LF, Hoffmann CJ, Bärnighausen T, Heumann C (2015). Simultaneous Treatment of Missing Data and Measurement Error in HIV Research Using Multiple Overimputation. Epidemiology, 26(5), 628-636.
- Obermeier V, Scheipl F, Heumann C, Wassermann J and Küchenhoff H (2015). Flexible distributed lags for modelling earthquake data. Appled Statistics, 64, Part 3, pp. 1-18.
- Bayerstadler A, Benstetter F, Heumann C and Winter F (2013). A predictive modeling approach to increasing the economic effectiveness of disease management programs, Health Care Management Science.
- Schomaker, M and Heumann, C (2013). Model selection and model averaging after multiple imputation. Computational Statistics and Data Analysis.
- Shalabh, Garg G, and Heumann, C (2012). Performance of Double k-class Estimators for Coefficients in Linear Regression Models with Non Spherical Disturbances under Asymmetric Losses. Journal of Multivariate Analysis, 112, pp. 35-47.
View other publications
- Pubmed: https://www.ncbi.nlm.nih.gov/pubmed/?term=Heumann+C
- Google scholar: https://scholar.google.de/citations?user=H6LdyzoAAAAJ&hl=de

Speakers

Christian Heumann, Professor, Department of Statistics, Ludwig-Maximilians-Universität München

Hosted by

Statistical Consulting Laboratory (StatsLab)

Registration

* Required

Join our events mailing lists

Always be the first to know about what's going on in our community. Sign up for one of our newsletters and receive information on a wide variety of events such as exhibition, lectures, films, art performances, discussions and conferences.