Using the American Community Survey as the Sampling Frame for the National Survey of College Graduates
The American Community Survey
The ACS, conducted by the U.S. Census Bureau, has replaced the decennial long form. It is conducted every month, with estimates from the monthly samples aggregated into annual national estimates and aggregated over longer periods for smaller area estimates. The ACS surveys about 3 million households annually. Questions on the ACS are generally identical to the questions that were on the decennial long form. Therefore, the ACS can provide reasonably detailed information about households and individuals each year rather than once a decade from the decennial Census.
The ACS has a new sample each month. It samples about 250,000 addresses each month, or some 3 million each year. Over a decade, the ACS will survey approximately 30 million addresses, compared with the long form's 17 million housing units at one time in the decennial Census. The sampling frame for the ACS is the U.S. Census Bureau's Master Address File, which is updated throughout the decade to keep it current. The samples are distributed throughout the country with no area or other cluster sampling, but there will be some higher sampling fractions in small governmental units.
The ACS is conducted using an initial mail-out, mail-return, self-response questionnaire. Mail nonresponse follow-up is first by computer-assisted telephone interviewing (CATI), followed by computer-assisted personal interviewing (CAPI) of a subsample of the remaining nonrespondents. A key function of the ACS is to produce estimates for various levels of geography (from small areas to the nation). ACS provides estimates annually for areas (and population groups) of 65,000 or more persons. Estimates for smaller areas/groups are provided based on data aggregated over multiple years. Estimates for the smallest areas or population groups are available based on data aggregated over 5 years, with reliability similar to that provided for these groups in recent decennial Censuses.
 The 2008 ACS questionnaire can be found at http://www.census.gov/acs/www/SBasics/SQuest/SQuest1.htm.
 No address will be included in the ACS sample more than once in a 5-year period.
 Beginning in 2006, this information will be made available annually in late summer/early fall for the previous year's sample.
 While small-area estimates are generally thought of as a physical location (often, as in the ACS case, government jurisdictions, such as counties), estimates for small/rare subpopulations (small domain estimates), such as scientists and engineers, are the equivalent statistical problem relative to sample size. NSF needs to use the ACS as the NSCG sample frame not because it wishes to produce small-area estimates but because it needs the ACS sample size for rare populations.