NELS:88 is a longitudinal study of a national probability sample of eighth graders. The base year student population excluded students with severe mental disabilities, students whose command of the English language was insufficient to understand survey materials, and students with physical or emotional problems that would limit their participation.
The survey used a two-stage stratified, clustered sample design. The first stage, selection of schools, was accomplished by a complex design involving two sister pools of schools. The second stage included selection of about 24 to 26 students per school. At the second stage, 93 percent of 26,435 selected students agreed to participate. Hispanic and Asian students were oversampled.
Data were collected via questionnaires from 24,599 students from 1,057 public and private schools from all 50 states and the District of Columbia in the base year. Eighth graders participated in group sessions at their schools where they completed student questionnaires and cognitive tests. School administrator data were collected from the senior school administrator (usually the principal or headmaster). For base year teacher data, each school was randomly assigned two of four subject areas of interest (English, math, science, social studies) and teachers were chosen who could provide data for each student respondent in these two subjects. Parent data were obtained through the mail.
For the first (1990) follow-up, all students were surveyed in schools containing ten or more eligible NELS:88 respondents. Only a sub-sample of students was surveyed in schools with fewer than ten students. Because 90 percent of students changed schools between eighth and tenth grade, it was necessary to sub-sample schools in this way. The 1990 sample size was more than 19,000 students, and the 1992 sample size is about the same.
The sample was freshened in 1990 and 1992 to give 1990 tenth graders and 1992 twelfth graders who were not in the eighth grade in 1988 some chance of selection into the NELS:88 follow-up. Such students included primarily those who had skipped or repeated a grade between 1988 and the follow-up year, and those who had moved to the U.S. after 1988. This freshening was conducted so that the first and second follow-up samples were representative of U.S. tenth graders in 1990 and U.S. twelfth graders in 1992.