GSS data are archived about six months after data collection is completed.

​​​​​​Yes. It is essential that users of the GSS consult the codebook to learn the exact wording of questions in the survey. The brief variable labels distributed with the data (e.g. “Abortion if woman wants for any reason” for variable ABANY) give only a general guide to variable content. Users must also learn the possible responses for each question and the codes assigned to them. The codebook also provides full documentation of other methodological features of the GSS that users must understand.

​​​​The on-line codebook on the GSS website covers all variables in the cumulative GSS. It also includes a series of important appendices that describe the study design, sampling methods, and coding protocols used in the GSS.

Much, but not all, codebook information is available electronically via the Survey Documentation and Analysis website at the University of California, Berkeley.

​​​The vast majority of GSS data is obtained in face-to-face interviews. Computer-assisted personal interviewing (CAPI) began in the 2002 GSS. Under some conditions when it has proved difficult to arrange an in-person interview with a sampled respondent, GSS interviews may be conducted by telephone.

​​​​Survey questions in the “GSS replicating core” are regularly administered as part of each GSS. Core items include background information about respondents (for example: age [AGE], sex [SEX], education [EDUC, DEGREE], region of residence [REGION]) and measures of attitudes (such as views about gender roles [e.g. FEHOME], confidence in institutions [e.g. CONFINAN] or gun control [GUNLAW]) or behaviors (such attendance at religious services [ATTEND] or voting in the most recent US Presidential election [VOTE68 through VOTE12]). Each GSS since 1985 also includes an International Social Survey Program (ISSP) module. Many items in ISSP modules are repeated from earlier modules on the subject covered. Items are very occasionally added to or removed from the GSS core. An especially notable number of items was removed from the core after 1994. To see how regularly any particular GSS item is measured, consult codebook Appendix U: Variable Use by Year/Ballot. Also see the memo on the Replicating Core.

​​​GSS interviews last 90 minutes on average, which is very long by the standards of survey interviews. Even so, the number of questions that can be included in any GSS is limited. Aside from the replicating core items which track change in selected social indicators, the GSS regularly includes special “modules”: sets of questions that add innovative content to the GSS and expand the range of topics it covers by going into depth on a subject within a single GSS.

NORC is the US member of the International Social Survey Program (ISSP). Each year, ISSP develops a cross-national module that focuses on a subject matter area such as the role of government, social inequality, national identity, or religion. ISSP modules consist of about 60 questions on that subject, which are answered by GSS respondents as well as respondents to national surveys in other countries. Many items in ISSP modules are repeated when that subject is again the focus of an annual ISSP module. Click here for a complete listing of ISSP modules.

Each ISSP module is obtainable in two ways. If you are only interested in the U.S. data, they are incorporated into any datafile associated with the year that the module was fielded. The full cross-national datasets can be obtained through GESIS' ZACAT and can be downloaded by registered users. Please note that there is often a two year lag on the release of ISSP data, to allow for harmonization of data from many countries.

"Topical” modules in the GSS focus on one subject area for US respondents only. Most items in topical modules appear on the GSS only once. There have been many GSS topical modules, including major modules on social networks ("Topical” modules 1985, 2004), intergroup relations and multiculturalism (1990, 1994, 2000), religion (1988, 1998), work and work organizations (1991, 2002), medical care (1996, 1998) and the Internet (since 2000). Occasionally a topical module will include items that have appeared in a previous GSS. To see the years in which any particular GSS item was measured, consult Appendix U: Variable Use by Year/Ballot.

​​​No. There are several reasons that questions may be asked of only some GSS respondents. First, some questions are not applicable to some respondents. For example, the question on marital happiness (HAPMAR) is only asked of currently married respondents. Second, since 1988 many items in the replicating core of the GSS have been measured for a random two-thirds of each sample. This enables the GSS to monitor more trends than would be possible if each core item was asked of each respondent. There are three overlapping “ballots”, so that the bivariate association of every core item with every other core item can be measured each year. For more detail about the ballot design, see Appendix Q: Rotation and Double Sample Designs. To see which items are on which ballot in any given year, see Appendix U: Variable Use by Year/Ballot.

Third, since 1994 the GSS has been administered to two samples in even-numbered years, rather than to a single sample each year. Questions in ISSP modules are usually asked of respondents in only one of these two samples. Questions in some topical modules are asked of respondents in only one of the two samples, but other topical modules cover both samples. In 2002 and 2004, many replicating core items were asked of respondents in only one of the two samples. For more detail about the two-sample design, see Appendix Q: Rotation and Double Sample Designs. To see which items are in which sample in any given year, see Appendix U: Variable Use by Year/Ballot.

​​​From 1972 until 1993, the GSS was administered almost annually. The target sample size for th​e annual surveys was 1500; actual sample sizes ranged between 1372 (1990) and 1613 (1972). Additionally, there were oversamples of black respondents in 1982 (oversample of 354) and 1987 (oversample of 353). There were no GSSs in 1979, 1981, or 1992.

Since 1994 the GSS has been administered to two samples in even-numbered years, each with a target sample size of 1500, rather than to a single 1500-person sample each year. Total sample sizes for these biennial GSSs ranged between 2765 (2002) and 2992 (1994). In 2006, a third sample was added and the sample size was 4510. There have been no oversamples during this period.

The GSS augemnted its replicating cross-sectional design with a design that uses rotating panels starting in 2006, 2008, and 2010. Under the new three-wave, rotating panel design in each round of the GSS, a new panel is started, the first reinterview from the previous GSS is conducted, and the second and final reinterview from the GSS panel started four years ago is conducted. In 2008 there were two components: a new 2008 cross-section with 2,023 cases and the first re-interviews (panel) with 1,536 respondents from the 2006 GSS. The 2,023 cases in the cross-section have been previously released as a part of the 1972-2008 cumulative data. This new release includes those 1,536 re-interviewed panel cases along with the 2,023 cases. In 2010 the rolling panel design was fully implemented for the first time. The rotating panel design was discontinued after 2014.

For more information on the GSS sampling design, and completion/response rates, see Appendix A: Sampling Design & Weighting.

​​​A short answer is yes and use WTSSALL. If you use the 1982 and 1987 data, there is one more step you need to do (please see below). Please read on for a long answer.

Users must decide for themselves whether and how to use weights when analyzing the GSS. The paragraphs that follow outline some reasons that users might wish to weight GSS data, and describe some weights made available as part of the GSS cumulative file.

From 1975 to the 2002 GSS used full-probability sampling of households designed to give each household an equal probability of being included in the GSS. Hence, for household-level variables the GSS is self-weighting.

Only one adult per household is interviewed, however, so persons living in large households have lower probabilities of selection. For person-level variables, weighting statistical results in proportion to the number of persons aged 18 or over in the household (variable ADULTS) can compensate for this.

The 1982 and 1987 GSSs included oversamples of black respondents. To adjust statistical results for this oversampling, one may either exclude cases in the black oversamples (codes 4, 5, and 7 on variable SAMPLE) or weight statistical results using weights in variable OVERSAMP.

Beginning in 2004, the GSS began to use a two-stage sub-sampling design for nonresponse. Cases from which no response has been obtained after the initial stage of the field period are subsampled, and resources are focused on gaining cooperation from this subset. Responses from persons in the subsample must subsequently be weighted up in order to represent all of those who had not responded by the time the subsample was drawn. Analysis of data from the 2004 and later GSSs should use weights WTSSALL, WTSS, or WTSSNR.

From 2006 until 2014, the GSS was divided into a nationally representative cross section and a 3-wave repeating panel. When using data from a single year with both sample types combined, two weights are avaialble, WTCOMB and WTCOMBNR. WTCOMBNR accounts for the subsamplling implemented in 2004. Analysis of any single year datafile between 2006 and 2014 should use WTCOMB or WTCOMBNR.

When working with full panel datafiles (e.g. the 2008-2010-2012 merged panel), four different weights are available.  WTPAN12 and WTPAN12NR cover the first two waves of the sample, with WTPAN12NR accounting for the subsampling procedure. When working with all three waves, WTPAN123 and WTPAN123NR should be used.

For additional detail about weights in the GSS, see Appendix A: Sampling Design & Weighting.

​​​Until 2000, the GSS measured race mostly by interviewer observation (variable RACE), using categories of white, black, and other. If in doubt about how to code a respondent’s race, interviewers asked the respondent “what race do you consider yourself?”

Beginning in 2002, the GSS measured race following the procedures used in the decennial Census, asking all respondents for a racial self-identification and recording up to three mentions. These data are in variables RACECEN1, RACECEN2, and RACECEN3 (The same questions also were asked as part of the “Multi-Ethnic United States” topical module administered to one of the two samples in the 2000 GSS).

For 2002 and later GSSs, a value of RACE has been imputed based on RACCEN1 and other information.

​​​Proposals to add questions to the GSS are considered by the PIs and the GSS Board of Overseers. Proposals to add questions ideally should be initiated about two years in advance of the administration of a GSS. For those who wish to propose items supported by the basic GSS grant from NSF see “Call for Proposals to Add Questions to GSS​.” For those who have funds to support paid supplements to the GSS see "Guidelines for Paid Supplements to the GSS​." Investigators seriously interested in proposing items for inclusion in a future GSS should contact the Director of the GSS, Dr. Tom W. Smith (smitht@norc.uchicago.edu).​

​​​NORC was (with Australia, Germany and Great Britain) one of four founding members of the International Social Survey Program (ISSP). The ISSP mounts a program of collaborative international comparative survey research. It develops an annual international module, administered in national surveys that take place in participating countries; all participating countries also collect data on a set of core demographic items. The ISSP has included 58 countries.

The annual ISSP module is administered as part of the GSS; two international modules have been included in each GSS since 1994, so that all ISSP modules have been fielded in the United States.

The GSS housed the Secretariat of the ISSP between 1997 and 2003. Director of the GSS Tom W. Smith served as the Secretary-General of the ISSP during this period.

​​​The GSS maintains an extensive bibliography of articles, chapters, books, and presentations known to have used GSS data. This currently includes more than 25,000 entries. The GSS wants to know about research you have done using the data. If you have written a report, article, chapter, thesis, dissertation, or book using GSS data, please notify the GSS (gss@norc.org).

This site provides a wealth of information about the GSS, including the on-line codebook, the codebook appendices, and the report series which includes Social Change, Methodological, Cross-National, Topical, and Project Reports.

​If you have any questions, please contact the GSS team at GSS@norc.org​

​​​If you used the datafile and the codebook:

Smith, Tom W., Davern, Michael, Freese, Jeremy, and Morgan, Stephen L., General Social Surveys, 1972-2018 [machine-readable data file] /Principal Investigator, Smith, Tom W.; Co-Principal Investigators, Michael Davern, Jeremy Freese and Stephen L. Morgan; Sponsored by National Science Foundation. --NORC ed.-- Chicago: NORC, 2019.

1 data file (64,814 logical records) + 1 codebook (3,758 pp.). -- (National Data Program for the Social Sciences, no. 25).

If you used only the codebook:

Smith, Tom W., Davern, Michael, Freese, Jeremy, and Stephen L. Morgan. General Social Surveys, 1972-2018: Cumulative Codebook / Principal Investigator, Tom W. Smith; Co-Principal Investigators, Michael Davern, Jeremy Freese and Stephen L. Morgan. -- Chicago: NORC, 2019. 3,758 pp., 28cm. -- (National Data Program for the Social Sciences Series, no. 25).

​​​'Hand Card' refers to a card that is handed to the respondent during the interview. The hand card allows the respondent to read and then select from the response categories for that item.

When a hand card is used, the 'Remarks' section of the variable description page lists which response categories were displayed on the card. When the visual layout of the card is important, the hand card is reproduced in the codebook.

GSS geographic identification code files are made available to researchers under special contract with NORC. The GSS takes its promise of anonymity to its respondents very seriously and this is the basis for the contract process. Under contract, the GSS will provide data on state, primary sampling unit, county, and Census tract, but in no circumstances will individually identifying information (name, address, etc.) be provided. Please read Obtaining GSS Sensitive Data Files in Documentation. If you have any further questions, please contact the GSS team at gss@norc.org.

There are several different ways to search for questions in the GSS. One possibility is to go to the main codebook which includes the wordings of all GSS questions in the cumulative file. A second possibility is to go the questionnaire files and conduct a standard text search using regular search utilities. Be aware that such searches would only cover the particular version and year of the GSS that you were accessing. Also, note that while recent GSS questionnaires generally included GSS mnemonics, this is not the case for most questionnaires.

The GSS Data Explorer search function covers both mnemonics and question text for each variable in the cross-section data. Basic analysis of variables can also be conducted.

The Subject Index in the GSS Codebook is a useful resource to find various information about the GSS. When looking for specific question, be aware that between 2006 and 2014, it only covers questions asked in the cross-sections. Items asked only in the panel reinterviews were are not covered, including two ISSP modules.

There are two different resources available to locate variables in the panel reinterviews. The first is the Panel Codebook, which documents all questions and frequencies asked in the panel reinterviews, similar to the GSS Cumulative Codebook. The second is the Cross-Section and Panel Variable and Module Identification, which can be used to identify both the modules with which any given variable is associated, and which, if any, panel reinterviews the module appeared in.

​​​Not necessarily. In 1983 and again in 1993 there were sample frame experiments that are described in Appendix A. In 1983, 806 cases were from the 1970 frame (SAMPLE=3) and 793 cases from the 1980 frame (SAMPLE=6). In 1993, 817 cases were from the 1980 frame (SAMPLE=6) and 789 case from the 1990 frame (SAMPLE=8).

The response rate for each year can be found here.​​​​

The field period for each year can be found here.

​​​​The target population of the GSS is adults (18+) living in households in the United States. From 1972 to 2004 it was further restricted to those able to do the survey in English. From 2006 to present it has included those able to do the survey in English or Spanish. Those unable to do the survey in either English or Spanish are out-of-scope. Residents of institutions and group quarters are out-of-scope. Those with mental and/or physical conditions that prevent them from doing the survey, but who live in households are part of the target population and in-scope. In the reinterviews those who have died, moved out of the United States, or who no longer live in a household have left the target population and are out-of-scope.

​​How to link the panel data with the merged data.