A Comparison of Imputation Techniques for Handling Missing Data

Abstract

Researchers are commonly faced with the problem of missing data. This article presents theoretical and empirical information for the selection and application of approaches for handling missing data on a single variable. An actual data set of 492 cases with no missing values was used to create a simulated yet realistic data set with missing at random (MAR) data. The authors compare and contrast five approaches (listwise deletion, mean substitution, simple regression, regression with an error term, and the expectation maximization [EM] algorithm) for dealing with missing data, and compare the effects of each method on descriptive statistics and correlation coefficients for the imputed data (n = 96) and the entire sample (n = 492) when imputed data are inculded. All methods had limitations, although our findings suggest that mean substitution was the least effective and that regression with an error term and the EM algorithm produced estimates closest to those of the original variables.

Peer feedback content and sender's competence level in academic writing revision tasks: Are they critical for feedback perceptions and efficiency?

“…We used the EM (Expectation Maximisation) procedure for imputation (Musil, Warner, Yobas, & Jones, 2002). Subsequently, we checked whether this imputation influenced the outcomes of all planned analyses.…”

Timely Intervention – Before It’s Too Hard

“…In all areas of quantitative and qualitative research there is the likelihood of missing values in data collection (Acock, 2005;De Leeuw, 2001;Dodeen, 2003;Molenberghs, 2007;Musil, Warner, Yobas, & Jones, 2002;Raaijmakers, 1999;Wainer, Bradlow, & Wang, 2007). There are various reasons why data are missing.…”

“…The major ways of data collection include "the self-administered questionnaire, the face-to-face interview, and the telephone interview" (De Leeuw, 2001, p. 152 (Acock, 2005;Huisman, 2000;Musil, et al, 2002).…”

Multiple Imputation of Missing Data with Genetic Algorithm based Techniques

“…Musil et al [10] provided empirical comparative analysis on list wise deletion, mean substitution, simple regression, regression with an error term and the EM algorithm. Junninen et al [11] experimented on univariate linear, spline and nearest-neighbor interpolation algorithm, multivariate regularized expectation-maximization algorithm, nearest-neighbor, self-organizing map, multilayer perceptron (MLP) as well as hybrid methods where combining the best features of univariate and multivariate methods are combined in air quality datasets.…”