Nsfs 383

Understanding the sampling frame of NSFS 383 is essential for any researcher using this dataset. Unlike random-digit dialing surveys, NSFS 383 employs a stratified, two-phase design.

NSFS 383 is a longitudinal panel (following the same individuals every two years). However, high-mobility individuals (those who move abroad or leave the workforce entirely) are lost to follow-up. Weights attempt to correct this, but bias remains.

NSFS 383 has been used in hundreds of peer-reviewed articles. Here are three typical research questions it can answer.

NSFS 383 only includes U.S.-trained doctorates. If you want to study foreign-trained PhDs working in the U.S., you must use the American Community Survey (ACS) instead.

svyby(~salary, ~gender, academic, svyquantile, quantiles = 0.5, ci = TRUE)

Important: Always use the provided survey weights (wtsurvy). Unweighted analysis of NSFS 383 will produce biased estimates because the stratified sampling oversamples small demographic groups.

Given the cryptic nature of NSFS 383, several interpretations have emerged:

The application requires: