National Education Longitudinal Survey (NELS): Report for Computation of Balanced Repeated Replicate (BRR) Weights for the Third (NELS88:1994) and Fourth (NELS88:2000) Follow-Up Surveys. Working Paper No. 2003-18

The objectives of this project were to construct two sets of balanced repeated replication (BRR) weights for two National Education Longitudinal Study of 1988 (NELS:88) sets of weights: one based on the third follow-up survey done in 1994 (NELS88:94); and the second based on the fourth follow-up survey done in 2000 (NELS88:2000). These weights are described here. "F3QWT" is the first weight. It applies to all members of the third follow-up sample who completed an interview in 1994, regardless of their participation status in previous rounds. When used with the appropriate sample flags, it allows projections to the following populations: (1) Spring 1988 8th graders eligible to complete questionnaires in 1992 and 1994, regardless of 1988 eligibility; (2) Spring 1990 10th graders eligible to complete questionnaires in 1992 and 1994, regardless of 1988 eligibility and 1990 participation; and (3) Spring 1992 12th graders eligible to complete questionnaires in 1992 and 1994, regardless of 1988 and 1990 eligibility and 1992 participation. "F4PNLWT" is the second weight. This panel weight applies to sample members who completed questionnaires in all five rounds of NELS:88 and were in school in the spring of 1988. It can be used to make projections to the population of spring 1988 8th graders. Four appendices present: (1) Weighting Specifications of BRR Weights for Third Follow-Up; (2) Hadamard Matrix Used to Construct BRR Weights; (3) Variables Used for Calibration of Weights; and (4) Weighting Specifications of BRR Weights for Fourth Follow-Up Full Panel Weight.