Item Imputation Without Specifying Scale Structure

Imputation of incomplete questionnaire items should preserve the structure among items and the correlations between scales. This paper explores the use of fully conditional specification (FCS) to impute missing data in questionnaire items. FCS is particularly attractive for items because it does not require (1) a specification of the number of factors or classes, (2) a specification of which item belongs to which scale, and (3) assumptions about conditional independence among items. Imputation models can be specified using standard features of the R package MICE 1.16. A limited simulation shows that MICE outperforms two-way imputation with respect to Cronbach’s α and the correlations between scales. We conclude that FCS is a promising alternative for imputing incomplete questionnaire items.

[1]  Russell V. Lenth,et al.  Statistical Analysis With Missing Data (2nd ed.) (Book) , 2004 .

[2]  Xiao-Li Meng,et al.  Multiple-Imputation Inferences with Uncongenial Sources of Input , 1994 .

[3]  Roderick J. A. Little,et al.  Statistical Analysis with Missing Data , 1988 .

[4]  Geert Molenberghs,et al.  Person fit for test speededness: normal curvatures, likelihood ratio tests and empirical Bayes estimates , 2010 .

[5]  S. van Buuren.  Multiple imputation of discrete and continuous data by fully conditional specification , 2007, Statistical methods in medical research.

[6]  Roderick J. A. Little,et al.  Statistical Analysis with Missing Data: Little/Statistical Analysis with Missing Data , 2002 .

[7]  David E. Booth,et al.  Analysis of Incomplete Multivariate Data , 2000, Technometrics.

[8]  Mark Huisman,et al.  Item Nonresponse: Occurence, Causes, and Imputation of Missing Answers to Test Items , 1999 .

[9]  K Sijtsma,et al.  Influence of Imputation and EM Methods on Factor Analysis when Item Nonresponse in Questionnaire Data is Nonignorable , 2000, Multivariate behavioral research.

[10]  Thomas R Belin,et al.  Imputation for incomplete high‐dimensional multivariate normal data using a common factor model , 2004, Statistics in medicine.

[11]  Jeroen K. Vermunt,et al.  Two-way imputation: A Bayesian method for estimating missing scores in tests and questionnaires, and an accurate approximation , 2007, Comput. Stat. Data Anal..

[12]  D. Rubin,et al.  Fully conditional specification in multivariate imputation , 2006 .

[13]  van der Ark,et al.  Multiple Imputation of Item Scores in Test and Questionnaire Data, and Influence on Psychometric Results , 2007, Multivariate behavioral research.

[14]  Ken P Kleinman,et al.  Much Ado About Nothing , 2007, The American statistician.

[15]  S. van Buuren,et al.  Multivariate Imputation by Chained Equations : Mice V1.0 User's manual , 2000 .

[16]  H. Boshuizen,et al.  Multiple imputation of missing blood pressure covariates in survival analysis. , 1999, Statistics in medicine.

[17]  D. Rubin.  INFERENCE AND MISSING DATA , 1975 .

[18]  L. Andries van der Ark,et al.  SPSS Syntax for Missing Value Imputation in Test and Questionnaire Data , 2005 .

[19]  Klaas Sijtsma,et al.  Investigation and Treatment of Missing Item Scores in Test and Questionnaire Data , 2003, Multivariate behavioral research.

[20]  Stef van Buuren,et al.  Imputation of missing categorical data by maximizing internal consistency , 1992 .