A biosocial analysis of the sources of missing data in criminological research

•Missingness is a pervasive methodological issue.•Missing data methods rely on identifying a comprehensive list of covariates.•Genes and environments have a significant influence on patterns of missingness.•Missing data methods should incorporate measures of genetic influence.

PurposeFailing to deal with missing data patterns effectively may result in biased parameter estimates and ultimately may produce inaccurate results and conclusions. The vast majority of criminological research has addressed this issue with listwise deletion (LD) and multiple imputation (MI) techniques. Identifying the specific covariates that directly contribute to patterns of missingness is highly important in deciding which technique to use. One of the more surprising omissions from the identified list of covariates is the potential role of genetic influences in the development of missingness.MethodsThe current study addresses this gap in the literature by estimating genetic (A), shared environmental (C), and the nonshared environmental (E) influences on missingness across measures of delinquency and self-control within a longitudinal sample of twin and sibling pairs.ResultsThe results indicated that genetic influences explain a significant portion of the variance in missing values related to both delinquency and self-control.ConclusionsCurrent methodological techniques aimed at addressing missing data should be amended to take genetic influences into account. Such modifications and the implications of the findings for future research are discussed.