
An official website of the United States government
Most National Longitudinal Surveys (NLS) cohorts were selected to represent a particular segment of the United States population at the outset of the survey:
An additional cohort, the NLSY79 Child and Young Adult (NLSY79-CYA) cohort, was sponsored by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD). It consists of all children who were born to women in the NLSY79 sample.
Additional information on the NLS cohorts can be found on the NLS cohorts webpage.
To facilitate statistically reliable analyses of certain subpopulations, some of the cohorts include oversamples, and the program provides weights to help users estimate population characteristics. Once selected, members of a cohort remain in the sample throughout their lives, and no additional or replenishment members are added.
The size of NLS cohorts and presence of oversamples among them is summarized in the following table:
Cohort | Sample size | Oversamples |
---|---|---|
Older Men |
5,020 | Black men sampled at twice the rate of White men |
Young Men |
5,225 | Black men sampled at twice the rate of White men |
Mature Women |
5,083 | Black women sampled at twice the rate of White women |
Young Women |
5,159 | Black women sampled at twice the rate of White women |
NLSY791 |
9,964 | Supplemental Black and Hispanic/Latino sample |
NLSY97 |
8,984 | Supplemental Black and Hispanic/Latino sample |
NLSY79-CYA |
11,545 | - |
Note: Dash (-) indicates blank cell. 1 The initial NLSY79 sample also included a supplemental sample designed to oversample economically disadvantaged non-Black/non-Hispanic youth in the same age group, and a military sample designed to represent the population aged 17 to 21 serving in the military as of September 30, 1978. Including these groups, the total sample size in 1979 was 12,686. However, these samples were later dropped. |
Although the details vary among them, all NLS samples (except the NLSY79-CYA) are selected using multistage designs. Starting from an address-based household register, the country is divided into approximately 150–250 sampling areas; within each sampling area, smaller areas are chosen. Finally, within these smaller areas, households are randomly selected to be contacted for a screening survey. Households identified during screening as having at least one member in the cohort are selected to be in the sample. This process ensures that the sample will be representative of the target population while conserving collection costs. For the first several years of the cohort, it results in a geographically clustered design that facilitates in-person collection; as the survey follows its members as they relocate, the amount of clustering diminishes over time.
More detailed information about the sampling and screening processes for the NLS cohorts can be found at the following links:
As the ability to retain respondents over the course of a longitudinal survey is a key component of ensuring the value of the data for analysis, NLS makes every effort to locate, contact, and gain cooperation from its sample members in every round. Two measures of its success are the retention rate, defined as the percentage of the initial sample responding in each round, and the response rate, defined as the percentage of remaining (nondeceased) sample members responding in each round. In 2020–21, during the 29th round, the NLSY79 maintained a retention rate of 65.6 percent and a response rate of 74.4 percent. In 2021–22, during its 20th round, the NLSY97 maintained a retention rate of 74.7 percent and a response rate of 76.9 percent. More details about retention among the NLS cohorts can be found at the following links: