There have been two surveys of high schools connected with the NLSY97. The first survey was conducted in 1996, and a follow-up survey was conducted in 2000. The surveys asked principals and their staff to provide information about the school's curriculum, facilities, staff, and student body.
The 1996 School Survey, which preceded Round 1 of the NLSY97, actually was a census of all the schools with a 12th grade within the 147 Primary Sampling Units (PSUs) from which the original NLSY97 sample was selected. A PSU is a metropolitan area or, in non-metropolitan areas, a single county or group of counties. The listing of eligible schools in the PSUs was obtained from a commercial database.
The 1996 School Survey used both a long form and a short form of the questionnaire. All the high schools in the 1996 sample originally were mailed the long form. In an attempt to boost the response rate for the survey, a short form of the questionnaire was created during the fielding period to make it easier for reluctant school administrators to respond. The short form included the questions from the long form that were regarded as the most important. The survey was conducted primarily by mail, but some schools provided information by telephone after numerous unsuccessful attempts at gaining cooperation through the mail. Of the 5,253 schools that responded to the 1996 survey, 3,065 responded using the short form and 2,188 responded using the long form.
The 2000 School Survey included all the schools in the original 1996 sample, plus vocational high schools in those original PSUs (vocational schools were not included in the 1996 survey), plus high schools that NLSY97 respondents had attended after moving to a residence outside of the PSU where they had lived at the time the NLSY97 sample was selected. Many of the schools in the 1996 and 2000 School Survey samples had NLSY97 respondents attending them, but some schools did not have any NLSY97 respondents attending them. The 2000 School Survey was conducted by mail and included many of the same questions as the 1996 survey, although the 2000 questionnaire was somewhat streamlined to make it easier to respond. There was no short form of the 2000 questionnaire.
The two School Surveys were intended to be longitudinal, although, by design, the 2000 sample included some schools that the 1996 sampled did not include. In addition, some schools that were in the samples for both years may have responded in only one year but not the other.
The School Surveys also were designed to enable researchers to match the information obtained from youths in the NLSY97 sample with information obtained from their high schools.
Shown below are the response rates for the NLSY97 School Surveys in 1996 and 2000, as well as information on the number of schools that participated in both years:
Number of schools that responded = 5,253
Number of schools eligible to participate = 7,342
Response rate = 71.5%
Number of schools that responded = 6,393
Number of schools eligible to participate = 9,013
Response rate = 70.9%
Sample retention from 1996 to 2000
Number of schools that responded in both years = 3,900
Number of schools that participated in 1996 = 5,253
Longitudinal retention rate = 74.2%
Access to data from the NLSY97 School Surveys is only available for approved researchers at the BLS National Office in Washington, D.C. For information on how to apply for access, see Researcher Access to Confidential Data Files at the Bureau of Labor Statistics. [http://www.bls.gov/bls/blsresda.htm]
2000 School Survey Questionnaire [PDF, 25 pages, 64K)
Last Modified Date: March 18, 2003