Respondent attrition is a common problem in national longitudinal panel surveys. To make full use of the data, weights are provided to account for attrition. Weight adjustments are based on sampling design information and data from the base year; information from subsequent waves is typically not utilized. Alternative methods to address bias from…

Quantitative research in vocational education and training (VET) is routinely affected by missing or incomplete information. However, the handling of missing data in published VET research is often sub-optimal, leading to a real risk of generating results that can range from being slightly biased to being plain wrong. Given that the growing…