Editor's note:

FYeI invites the University Faculty Senate to author a column in each issue.

The remarkable success of the Internet and especially the World Wide Web with its graphical browsers have provided an opportunity for easier access to social science data. Software to permit point-and-click access to data and online data analysis is just now becoming available at several sites around the country. The City University is at the cutting edge in this effort, and QServe, a new program developed in the sociology department at Queens College by Jesse Reichler and Dean Savage, is now available online at Hunter and Queens.

This point-and-click browser allows users on any level either to create data extracts which can be downloaded for local analysis, or to choose variables and types of analysis (tables, scatterplots, bar charts, etc.) for online display and analysis. Users can then save their charts and tables (along with a page of comments) in a gallery on the server, where they can be browsed by other users, or downloaded and stored or printed locally.

No programming skills are necessary. All a user needs is a knowledge of Web browsers and the basic logic of data analysis. To select variables or types of display, users simply point and click their way through various choices. There is also a feature which allows respondents (e.g., all the members of a class) to complete an online questionnaire, the results of which are immediately available for analysis.

The initial datasets put up at Queens College include the General Social Surveys, 1972-1993. The GSS is an annual representative survey of 1,500 U.S. adults that is one of the most widely-used datasets in the social sciences. Some questions are the same from year to year, while others are new or asked only occasionally, which provides for a rich database of more than 2,000 variables. Users can also search the codebook online, as well as an accompanying bibliography of some 3,000 abstracts of articles which have used the GSS as a basis for research (the URL is http://www.soc.qc.edu/qserve).

At Hunter College the initial data set will be the 5% Public Use Microsample (PUMS) of the 1990 census for the New York City Metropolitan Area. Data on households and persons from the long form of the 1990 census for some 800,000 respondents will be available (the URL is http://moe.hunter.cuny.edu/pums). QServe has been written for use on all types of Web servers, no matter which operating system is used. After an initial testing period (including some benchmarking on slower speed processors), it will be made freely available for use at other Web sites. Locally installed versions can be set up to run with any data set. Plans for enhancement include a "travel" button, by which the user will be able to locate data on other servers running QServe at CUNY (or beyond).

One of the more interesting possibilities involves classroom use: data analysis teams in classes at two separate campuses could conduct surveys or analyze data, and then post their results on the Web for collaborative review and evaluation.

Some initial support for this project came from CUNY's Office of Instructional Technology. The Queens team is applying to the National Science Foundation and to FIPSE for additional funding which, if successful, will allow the authors to continue to develop and enhance this useful program. Stay tuned for further developments!