Imputation of missing item responses with the Mokken scaling
imputemok varlist [,prefix(string) max(#)]
Description
imputemok imputes missing item responses with the Mokken scaling as defined in Huisman and Molenaar (2001). This module runs only with dichotomous items.
The following algorithm is used:
First, the items are ordered according to the percentage of positive responses (in a decreasing order).
For each individual, if a positive response follows a missing response, it is imputed to 1.
Else if a negative response precedes a missing response, it is imputed to 0.
Else we count the number of positive and negative responses preceding a missing response and if the number of negative response is larger or equal than the number of positive responses, the missing value is imputed to 0.
Else we count the number of positive and negative responses following a missing response and if the number of positive response is larger or equal than the number of negative responses, the missing value is imputed to 1.
Else, the missing value is imputed by drawing a random number based on the observed proportion of positive responses to the item.
Options
prefix defines the prefix to use to name the imputted variables (this prefix is followed by the name of the initial variable). By default, this prefix is "imp".
max allows imputing missing values only for individuals with a maximal number of missing values defined with this option.
Example
. imputemok itemA*
. imputemok itemA*,prefix(new)
Reference
Huisman M. and Molenaar I. W., Imputation of missing scale data with item response models. In A. Boomsma, M.A.J. van Duijn, & T.A.B. Snijders (Eds.), Essays on item response theory (pp. 221-244). New York: Springer-Verlag, 2001.
Author
Jean-Benoit Hardouin, PhD, assistant professor EA 4275 "Biostatistics, Clinical Research and Subjective Measures in Health Sciences" University of Nantes - Faculty of Pharmaceutical Sciences 1, rue Gaston Veil - BP 53508 44035 Nantes Cedex 1 - FRANCE Email: jean-benoit.hardouin@univ-nantes.fr