General Social Survey, 2021
DOI
10.17605/OSF.IO/YGTZDCitation
Davern, M., Bautista, R., Freese, J., Morgan, S. L., & Smith, T. W. (2022, March 7). General Social Survey, 2021.Summary
The General Social Surveys (GSS) have been conducted by the National Opinion Research Center (NORC) annually since 1972, except for the years 1979, 1981, and 1992 (a supplement was added in 1992), and biennially beginning in 1994. The GSS are designed to be part of a program of social indicator research, replicating questionnaire items and wording in order to facilitate time-trend studies. This data file has all cases and variables asked on the 2021 GSS.The 2021 cross-sectional General Social Survey has been updated to Release Version 3 as of July 2023. This Release includes the addition of respondent spouse/partner religious identities, socioeconomic statuses, and work information (including Occupation and Industry coding), and additional information about respondents' religious background.
To download syntax files for the GSS that reproduce well-known religious group recodes, including RELTRAD, please visit the ARDA's Syntax Repository.
The ARDA has added six additional variables to the original data set to enhance the users' experience on our site.
Data File
Cases: 4032Variables: 735
Weight Variable: FORMWT, OVERSAMP, WTSSPS, WTSSNRPS
Data Collection
Dec. 1, 2020 - May 3, 2021Original Survey (Instrument)
General Social SurveyFunded By
National Science FoundationCollection Procedures
No in-person contact with selected households and respondents was done; outreach was conducted via mail and phone using commercially available phone number matches for addresses in the sample or inbound phone contact. GSS staff redesigned the mail-based outreach to respondents to introduce the GSS to fresh address-based sample and encourage them to participate, either by web or over the phone. Throughout the data collection period, selected households were sent postcards, invitation packets, and reminder letters using a combination of USPS and FedEx and urging them to complete the 2021 GSS survey. The sample was released in three batches, with evaluation of respondent recruitment protocol for each batch allowing the fine-tuning of protocol for subsequent batches.Within households, respondents for the 2021 GSS Cross-section were selected by the last birthday method. In previous rounds of the GSS, NORC interviewers have used a Kish grid to select respondents within households, but this method was not possible without in-person interviewers.
A professional phone interviewer team conducted phone outreach to complete screeners, answer inbound phone calls from sampled households, identify respondents, and complete interviews by phone and/or prompt them to complete the web survey.
Sampling Procedures
The GSS 2021 sample was released in three batches, each with different start dates. The first sample batch consisted of 5,891 addresses from the NORC 2010 NORC National Sampling Frame, which we describe below, and 4,200 unclustered addresses from a version of the United States Postal Service (USPS) Computerized Delivery Sequence File (CDS), for a total of 10,091 addresses. Batches two and three were based on the CDS and composed of 10,000 and 7,500 unclustered addresses from the CDS, respectively. Batch 1 was released to the field on Dec. 1, 2020. Batches 2 and 3 were released to the field on Jan. 21, 2021, and Febr. 24, 2021, respectively; all three batches remained open for data collection until the end of the field period on May 3, 2021.The three sample batches supported the main mode of data collection based on a web-instrument and supplemented with phone for a limited number of cases who required it. The fact that the data collection method did not rely on interviewer-administration modes represented an opportunity to use a stratified un-clustered address sample. This is in contrast with previous years, which were primarily face to face, with a stratified clustered sample.
As part of batch 1, the NORC National Sampling Frame provides the basis for an equal-probability multi-stage cluster sample of housing units for the entire United States. Every decade, NORC reselects a National Sampling Frame based on the Decennial Census. The 2021 GSS was the fourth to use the 2010 NORC National Sampling Frame.
First-stage units on the National Sampling Frame are called National Frame Areas (NFAs), each of which is composed of a USPS Metropolitan Statistical Area (MSA) of one or more counties outside of MSAs. The number of first-stage selections in the 2010 NORC National Sampling Frame is 126 NFAs and the GSS sample uses only 76 NFAs. Fifty-nine of the GSS NFAs are non-certainty NFAs in the GSS each represents 1 percent of the U.S. population, and the 17 of the GSS NFAs are certainty NFAs containing 41 percent of the U.S. population. The second-stage selection yielded a sample of segments that are census tracts in urban areas (where post office address lists are usually available) and a sample of segments that are block groups (smaller than census tracts) in rural areas where physical listing is often required.
The first-stage units (NFAs) and second-stage units (either census tracks or block groups) were selected with probabilities proportional to size, where the measure of size is determined by Census housing unit totals. The third stage units (i.e., housing units) were selected with an equal probability of selection method (EPSEM), which means that each housing unit has the same probability of selection. Using the EPSEM approach resulted in roughly the same number of housing units selected per second-stage sampling unit. The number of second-stage selections in the 2010 NORC National Frame is 1,514, but GSS uses only 400, which is four per 1 percent of the U.S. population.
Most of the addresses in the NORC National Sampling Frame come from the USPS CDS, but other than the 5,891
addresses in the first batch from the NORC National Sampling Frame, all other addresses sampled come from an
EPSEM sample of all addresses from the CDS, which are all city-style addresses serviced by USPS.