Imputing Missing values in Relational Event History data: A Framework for Social Network Research

Abstract

[""Missing data can have significant effects on reliability of results and lead to incorrect conclusions. This study examines the effectiveness of multiple imputation in relational event history data. The study compares the estimates of a relational event model of the complete data with a 100 simulations where missing data was generated using the assumption that the data is missing completely at random (MCAR). It was found that, overall, the imputation method gave accurate estimates. However, the significance of the estimations changed from being not significant to significant. This change in significance should be taken into consideration when interpreting results after imputation."]