The Effects of Methods of Imputation for Missing Values on the Validity and Reliability of Scales

Abstract

The main aim of this study is the comparative examination of the factor structures, corrected item-total correlations, and Cronbach-alpha internal consistency coefficients obtained by different methods used in imputation for missing values in conditions of not having missing values, and having missing values of different rates in terms of testing the construct validity of a scale. The research group of the study, which is of a basic research, consists of 200 teacher candidates who attended the Department of Elementary Education at Ankara University, Faculty of Educational Sciences during the 2008-2009 Academic Year's spring term. The data were gathered by the Fatalism Scale (Sekercioglu, 2008), and exploratory factor analysis based on principal component analysis method was used. The findings showed that the "single factor" structure of the scale, whose construct validity was examined in the context of the study, was also found as "single factor" when it was obtained by the original data set having no missing values in situations of imputation for missing values with different methods whereas it also caused decrease in explained variance for imputation for missing values. A similar decrease was also seen in eigenvalues and Cronbach-alpha internal consistency coefficients.