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…

Recent investigations of the extent to which students with disabilities are allowed to participate in major national data collections used in measurement-driven education reform suggest that 40 to 50 percent of students with disabilities are typically excluded from major assessments, although they are included to a greater degree in assessments…