A researcher is investigating the relationship between the length of criminal trials and the likelihood of conviction. They want to use jurimetrics to analyze this relationship. However, they encounter a dataset with missing information regarding the length of some trials. Which of the following approaches is the MOST appropriate and statistically sound method to address this missing data in the context of jurimetrics research?