NLS user since 2001
The NLS data has enabled learning about the nature of measurement error in schooling by exploiting a clever survey design that asked individuals their own schooling level as well as the schooling attainment of the individual’s sibling. The result is having two reports for schooling that can be employed to statistically test a number of different models of measurement error (Flores-Lagunes and Light, Contributions to Economic Analysis & Policy, 2006).
Another interesting feature of the NLS data has allowed learning about the nature of the labor market returns to attaining educational degrees. One feature in NLS is the availability of both schooling attainment (years of schooling) and educational degrees attained by sample individuals, which allows estimating the returns to credentials above and beyond the years spent in school. Based on that NLS feature, we learned that the years of schooling is negatively correlated with ability for degree earners—the more able graduate the fastest. Conversely, the years of schooling is positively correlated with ability for non-degree earners (dropouts) because the most able benefit from increased years spent in school (Flores-Lagunes and Light, Journal of Human Resources, 2010).
The NLS data have several noteworthy features that allowed me to refine the measurement of important variables and attain special insights in my research. Examples of these features are given in the answer to the previous question: availability of self- and sibling-reported schooling attainment, and the availability of schooling attainment and educational credentials earned. Other noteworthy features that are extremely useful but easy to overlook are the longitudinal nature of the survey and the availability of a measure of actual labor market experience. These are just examples of the may interesting features of the data produced by the NLS!