PEDAR: Research Methodology  The Road Lsss Traveled? Students Who Enroll in Multiple Institutions
Beginning Postsecondary Students Longitudinal Study
The 2001 Baccalaureate and Beyond Longitudinal Study
Accuracy of Estimates
Item Response Rates
Data Analaysis System
Statistical Procedures
Differences Between Means
Linear Trends
Multivariate Commonality Analysis
Missing Data and Adjusting for Complex Sampling Design
Interpreting the Results
Executive Summary
Full Report (PDF)
Executive Summary (PDF)
 Beginning Postsecondary Students Longitudinal Study

The Beginning Postsecondary Students (BPS) Longitudinal Study is composed of the students who participated in NPSAS:96. The BPS sample consists of approximately 12,000 students identified in NPSAS:96 who were beginning postsecondary education for the first time in 1995–96. Among the NPSAS:96 respondents the response rate was 85.9 percent.1 The First Follow-up of the BPS cohort (BPS:96/98) was conducted in 1998, approximately 3 years after these students first enrolled. Approximately 10,300 of the students who first began in 1995–96 were located and interviewed in the 1998 follow-up, for an overall weighted response rate of 79.8 percent. This response rate includes those who were nonrespondents in 1996. The Second Follow-up of the BPS cohort (BPS:96/2001) was conducted in 2001, 6 years after students’ college entry. All respondents to the First Follow-up, as well as a subsample of nonrespondents in 1998, were eligible to be interviewed. Over 9,100 students were located and interviewed. The overall weighted response rate was 76.1 percent, with an institutional response rate of 91.1 percent and student response rate of 83.6 percent.2

Nonresponse among cohort members causes bias in survey estimates when the outcomes of respondents and nonrespondents are shown to be different. A bias analysis was conducted on the 2001 survey results to determine if any variables were significantly biased due to nonresponse.3 Considerable information was known from the 1996 and 1998 surveys for nonrespondents to the 2001 interviews, and nonresponse bias could be estimated using variables with this known information. Weight adjustments were applied to the BPS:96/2001 sample to reduce any bias found due to unit nonresponse. After the weight adjustments, some variables were found to reflect zero bias, and for the remaining variables, the bias did not differ significantly from zero. The weight variable used in this report for analysis of the BPS:96/2001 data is B01LWT2.

next section