Diary

Internal Reports

Completed Papers

Papers in Progress

Conference Presentations

Site Visit and Quarterly Reports

Main DLI Page

Web Client- DeLIver

Social Science Research- Social Science Team

Quarterly Report

(work from September 1994 through February, 1995)

The Sociology Research is being performed by faculty and students in the Graduate School of Library and Information Science (GSLIS), with support from the Departments of Sociology and of Economics. Its primary focus is Testbed Evaluation, i.e. studying the actual usage of the evolving Testbed and the needs of its users for document manipulation.

A variety of research methods is being applied to the evaluation. Focus groups of selected ranges of users are interviewed to determine their information needs and desires. The first set of these interviews was performed during this quarter and its results have already affected the prototype search interface. (A summary has been posted to our DLI Web pages under Working Papers.) For example, there was a strong desire to selectively search the individual special parts of documents, such as figure captions and table contents. Now that the prototype interface for the Testbed is complete (see above), a sequence of Usability testing can take place in the next quarter, where a range of users will be brought into the Grainger Usability Laboratory and observed while using the prototype on sample problems. Towards the end of this year, when a substantial number of users is using the initial production system, serious planning for the first user survey will begin. Surveys are useful for evaluating hundreds or thousands, as opposed to the tens or hundreds for the interviews. For the economics trials, where a page charge for documents is made, there is a nearly arranged collaboration with Carnegie-Mellon on the NetBill project, including a Mosaic interface to NetBill to be built by NCSA.

As a research investigation, we are also planning system instrumentation via transaction logs with the hopes of eventually measuring user behavior in the range of millions. The design phase for this is nearly complete and the implementation will begin in the next quarter. The Mosaic instrumentation will measure user commands issued and document structure viewed, a fine- grain strategy which may lead to net-ethnography, where cognitive observations about user needs can be deduced from statistical automatically recorded usage data.

co-PI Ann Bishop is the team leader for the sociology evaluation with co-PI Leigh Star as technical lead for the ethnography and research. (Star has a NSF sabbatical fellowship this year from SBE to learn information systems in Schatz’s lab and Bishop has applied for one for the coming year.) Larry DeBrock is the lead on the economics issues. Alaina Kanfer was on loan from NCSA for this quarter to help plan the instrumentation. Ed Lakner, director of the Library Research Center, will be analyzing the surveys. Bill Mischo will be providing technical assistance with the usability testing. One indication of the integration of our project is that both the first generation (testbed) and the second generation (research) system teams have assigned a student from their subprojects to attend the regular sociology meetings.