Imputation of missing values in survey data
Survey data often includes missing values. An approach to deal with missing values is imputation in order to obtain a complete dataset. However, the process of imputation requires researchers to make various decisions regarding the imputation method to be applied, the number of values to be imputed for each missing value, the selection of predictor variables, the treatment of multivariate nonresponse and the conduct of variance estimation. This survey guideline provides an overview of imputation procedures for missing values. It aims to support the reader with respect to aforementioned decisions when imputing missing values in survey data.
Christian Bruch (2023). Imputation of missing values in survey data. Mannheim, GESIS - Leibniz Institute for the Social Sciences (GESIS - Survey Guidelines). DOI: 10.15465/gesis-sg_en_044
Related Survey Guidelines:
-
Bruch, C. (2022). Applying the rescaling bootstrap under imputation for a multistage sampling design. Computational Statistics, 37, 1461–1494. https://doi.org/10.1007/s00180-021-01164-6
-
Axenfeld, J., Bruch, C., & Wolf, C. (2022). General-purpose imputation of planned missing data in social surveys: Different strategies and their effect on correlations. Statistics Surveys, 16, 182–209. https://doi.org/10.1214/22-SS137
-
Bruch, C. (2019). Applying the rescaling bootstrap under imputation: a simulation study. Journal of Statistical Computation and Simulation, 89, 641-659. https://doi.org/10.1080/00949655.2018.1563898
-
Bruch, C. (2016). Varianzschätzung unter Imputation und bei komplexen Stichprobendesigns. [Doctoral dissertation, Trier University]. https://doi.org/10.25353/ubtr-xxxx-1a14-1c19
-
Münnich, R., Gabler, S., Bruch, C., Burgard, J.-P., Enderle, T., Kolb, J.-P. and Zimmermann, T. (2015). Tabellenauswertungen im Zensus unter Berücksichtigung fehlender Werte. AStA: Wirtschafts- und Sozialstatistisches Archiv, 9(3), 269-304. http://dx.doi.org/10.1007/s11943-015-0175-8