Validation of Prediction Models in the Presence of Missing Data

• Activity Number: 444
• Type: Contributed
• Date/Time: Tuesday, August 6, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Survey Research Methods Section
• Author(s): Yuanyuan Guo*+ and Dean M. Young
• Companies: Baylor University and Baylor University
• Keywords: complete case analysis ; multiple imputation ; logistic regression ; MCAR

Abstract

Missing data arises in almost all research, part or all of the data are missing for a subject. There are a number of alternative ways of dealing with missing data and we all have to decide how to deal with it from time to time. In this study, we evaluated complete case, single value and multiple imputation methods with prediction measurements of logistic regression model. We found that for missing completely at random (MCAR) mechanism, multiple imputation is better than the other two with small sample size and the single value imputation is doing almost as well as the multiple imputation. The three methods started showing less difference when the sample size was increased.