GSSDIRS

The General Social Survey Data and Information Retrieval System (GSSDIRS)

The National Opinion Research Center (NORC) and the Inter-university Consortium for Political and Social Research (ICPSR) with support from the National Science Foundation have developed a state-of-the-art Internet service for users of the General Social Survey (GSS). This service is suitable for both the novice and the experienced user, and useful for both research and teaching. This service permits experienced researchers to do things that could not be done before, even by the GSS staff, because the data and documentation is integrated and cross-referenced as never before.

Services include facilities for hypertext viewing and searching of complete survey documentation, an ability to draw customized and documented extracts from data sets, statistical analyses, and File Transfer Protocol (FTP) delivery of full or extracted data sets. Allied to these services is also the capacity to engage in synchronous collaborative research with colleagues at distant sites, electronic conferencing, and the ability to consult substantive experts. The prototype system is known as the GSS Data and Information Retrieval System (GSS DIRS) and contains data from the 1972-1994 GSSs. Data from the International Social Survey Program surveys and other important data sets will later be configured into similar systems.

Netscape and the Web

: GSS DIRS uses the Netscape client to provide service to users of Macintosh, DOS and UNIX desktop environments. Netscape is built around existing network protocols, formats, data sources, and computing environments. With Netscape at one's desktop and an Internet connection, one can move from homepages through hierarchically structured layers of documentation (hypertext). The user can initiate processes such as data extraction and analysis programs. Printing and other desktop services are handled by Netscape, with the user not having to exit the user-friendly Netscape environment for those services.

What will the User be Able to Access?

Through the GSS DIRS Homepage, the user can access information about the GSS through a layered set of documentation, which begin with:

GSS Guide:: A brief overall document designed to explain the survey and the services available, and to direct users to the various services described below.
GSS News:: Contains the latest edition of the project newsletter, GSS News, general interest information, and special announcements.
Ask GSS:: Direct access to either the GSS staff to ask them about GSS or to the ICPSR staff to ask them about the operation of GSS DIRS.
GSS Roundtable:: An open forum for GSS users to discuss matters related to the GSS such as research topics, teaching experiences, and methodological or statistical concerns. GSS staff will participate in the discussions when appropriate.

In addition to these informational and contact services GSS DIRS provides fully documented access to GSS data as follows:


GSS Electronic Codebook:: Provides full question text, response categories, distributions, and special remarks. This and all other components described below are searchable by mnemonic name, topical headings, a Wais index, and scrolling. All components are crosslinked by hypertext.
GSS Electronic Codebook Appendices:: Provides documentation on sampling, interviewing, coding, and other essential documentation.
GSS Trends Tables:: SPSS generated tables for all time series from 1972 to 1994. Items are properly weighted to take into consideration shifts in measurement procedures and sample design.
GSS Bibliography:: The 10th edition of the Annotated Bibliography of Papers using the General Social Surveys contains 2,985 main entries which include a full reference; lists of GSS years analyzed, other data used, and GSS mnemonics utilized; and an abstract. A supplement lists full references for an additional 186 uses. Newly identified uses will be added on a continuous basis.
GSS Reports:: Full-text access to papers in the five GSS Reports series: GSS Methodological Reports, GSS Social Change Reports, GSS Cross-National Reports, GSS Topical Reports, and GSS Project Reports. There are currently over 170 reports and new reports will be added as completed.
GSS Cumulative Data File:: All data from the 1972 - 1994 GSS is accessible. Users can either download all of the data or sub-sets of it (GSS Extracts) or to analyze it directly (GSS Statistical Analysis).
GSS Extracts:: Users can create sub-sets as SPSS or flat files. Users can specify what variables and years should be included.
GSS Statistical Analysis:: Users can conduct simple analysis such as cross-tabulations and frequencies.

Examples of Use

Suppose a researcher wanted to study racial differences in inter-personal trust. After entering the homepage and being introduced to GSS DIRS by GSS Guide, she would go to GSS Electronic Codebook. A search by key word, subject index, or Mnemonic would locate the measures of racial and ethnic identification (RACE and ETHNIC) and several possible trust items (RACTRUST, TRUST, FAIR, HELPFUL). After reviewing the wordings, she could jump over to GSS Bibliography and locate previous GSS uses that used these variables. Clicking on to GSS Trend Tables would show her the trends for these items. If she was concerned about sampling procedures or whether there were any special technical notes about the items, she could go to GSS Electronic Codebook Appendices and GSS Reports. If she found a section or a report that she wished to keep, she could download a copy.

Having decided that these variables were worthy of further examination, she could go to GSS Statistical Analysis and cross-tabulate the measure about trusting people in general (TRUST) by race (RACE) or look at educational difference (EDUC or DEGREE - her choice) on the measure of Blacks trusting Whites (RACTRUST). If these inspections seemed promising, she could use GSS Extract to save all cases from all years for the variables RACE, TRUST, EDUC, and RACTRUST in SPSS format on her hard drive and request a customized mini-codebook for these variables.

Now suppose another researcher wanted to examine President Carter's contention that America was suffering from malaise in the late 1970s. Getting into GSS Electronic Codebook he could search for either 'malaise' as a keyword or subject heading. He would come up empty. He might then try a synonym like 'debility' or a related term like 'dispirited', but he still won't find anything. Rather than giving up (since his colleague had told him that the GSS had data on this topic), we might scroll through the entire codebook to see its whole contents. This would be a slow slog, equivalent to thumbing through 750 pages in the printed codebook. More expeditiously he might do a key word search on 'malaise' in the GSS Bibliography. This would find two papers and these would suggest possible variables on anomia (ANOMIA5-7), general happiness (HAPPY), marital happiness (HAPMAR), and satisfaction (SATCITY, SATFAM, SATFRND). Alternatively he could switch to Ask GSS or GSS Roundtable and ask either the GSS staff or the GSS user community for help. This would turn-up other possible variables such as the confidence in institutions battery or the measure of participation in political groups (MEMPOLIT).

Take the experienced GSS user who might go to GSS News to learn about new data sets, upcoming modules, etc. then go to GSS Statistical Analysis to crosstabulate RACE and HAPPY (already knowing the mnemonics), download the table, and then exit the system.

Summary

Constructing advanced and fully integrated services for the Internet allows maximum access to the GSS at minimum cost. GSS DIRS makes more data and more documentation more readily available than is possible under any other information retrieval and distribution approach. It also enable analysts to do things that even the GSS staff itself cannot now do. In addition, the services are extremely user-friendly with all technical aspects handled by simple menus and options, and capable of serving the needs of both the novice and the quantitative data analysis veteran. Moreover, these services may serve as a model for access to other important data resources and as the foundation for further development of such services throughout the social sciences.

GSS DIRS has been available to the public since May 15, 1996.

Tom W. Smith
NORC, University of Chicago