sist NEWSLETTER vol. 1, no. 3 PAPERS PRESENTED AT THE SECOND lASSIST NORTH AMERICAN MEETING, MAY 11 - 12,1977 [Editor's Note: The following are abstracts of papers presented at the Second IASSIST North American meeting. The full proceedings of the Conference will be available in the near future. All participants of the Conference will re- ceive a full copy of the Proceedings ; lASSIST members will receive a copy at a reduced rate; all other individuals should contact Sharon Henry of the Data Clearing House for the cost to non-IASSIST members.] lASSIST - Progress Towards Solving Problems in Data Archiving Carolyn Geda Inter-university Consortium for Political and Social Research The paper reviews the establishment and development of lASSIST including the rationale for such an Association. Problems encountered during the forma- tion of the International Organization are presented with interim proposals for dealing with them. The governing body, regional structure, and functions of the action groups are discussed. A brief review of Regional activities is given and potential directions for the Association are projected. Canadian Secretariat Report Rachel Des Rosiers Data Clearing House for the Social Sciences The report summarizes the membership campaign for lASSIST in Canada. It outlines the results as well as expected plans to increase the membership in Canada, e.g., by contacting related associations in the field and publicising lASSIST information in professional journals and at conferences. It also dis- cusses its relationship with the American secretariat, especially in regard to Action Groups' projects and plans. Finally, it invites comments and sugges- tions from the Canadians present concerning its intended role and functions. iBi^sist NEWSLETTER vol.1, no. 3 lASSIST Secretariat Report: United States Judith Rowe Princeton University The report of the United States Secretariat covers the current status of mem- bership enrollment as well as proposed membership recruitment activity. Specifi- cally, it addresses the goal of recruiting for Action Group activity all of the professional staff of each data archive and data library. It includes a general summary of both secretariat and Action Group activity, as well as a report on the first North American Action Group Conference in Cocoa Beach. Other topics include an initial report on the plans for next year's conference and some activities and programs planned by other organizations which would be of interest to lASSIST mem- bers. Among the latter are the annual conferences of Special Libraries Association (SLA), Association of Public Data Users (APDU) and the American Association for In- formation Science (ASIS), as well as the Inter-university Consortium for Political and Social Research (ICPSR) workshop for data librarians. Data Archive Registry-Survey of Past Effort, Suggestions for the Future Lisa Lasko Canadian Consortium for Social Research The mandate of the lASSIST Action Group for Data Archive Registry is described. The major portion of the paper is devoted to a brief review, description, and eval- uation of some of the more important existing directories. These are: Social Science Data Archives in the United States , published by the Council of Social Sci- ence Data Archives in 1967; A Directory of Information Resources in the United States: Social Sciences , 2nd edition, published by the Library of Congress in 1973; the second edition of the Encyclopedia of Information Systems and Services , edited by Anthony Kruzas and published in 1974; and the Directory of Data Bases in the Social and Behavioral Sciences , edited by Vivian Sessions and published in 1974. Recent developments in data archive registries, such as the UNESCO sponsored Directory of Data Services and the Directory of Data Centres , to be published by the Data Clearing House for the Social Science in Canada are also discussed. Es- sential elements and user requirements of social science data archive directories are dealt with in reference to these directories. Finally, a summary of the problems facing the Data Archive Registry Group, both abroad and in Canada, is given, along with several proposed options include a discussion of the viability of a Data Archive Registry Group in Canada; unique contributions to be made in this area, one of which might be a compilation of a list of subject headings for social science data archives; and, the production of a directory of data archive personnel. 20. laiSsist NEWSLETTER vol. 1, no. 3 Problems in Handling Process-Produced Data John DeVries Social Science Data Archive Carleton University The paper is an attempt to outline the problems one encounters in the handling of process-produced data, starting with their definition through the stages of ob- taining information about them and acquiring them, to the phases of transforming and analysing them. In addition to enumerating and discussing the often unique problems with which the user of process-produced data is faced, I will try to men- tion various projects which have already begun, and indicate future plans by the lASSIST Action Groups on process-produced data. Definitions . In the last few years, various definitions have been suggested. Al- though there is considerable overlap between them, the remaining differences make it important that we come to some agreement about the proper "domain" for an Action Group on process-produced data, and thus to a clear definition which unambiguously specifies this domain. I will discuss the definitions of which I am aware, and present some arguments in support of the definition which was developed at the Cocoa Beach conference. Information and acquisition . Generally, we will find that process-produced data have been generated by governments and other public organizations. Problems in this phase are thus to a large degree subsumed under the general heading: "The relationship between the research community and the government-" Specific issues I will discuss include: privacy versus the right to information, i.e., the govern- ment as a research resource; and, obtaining information from governmental agencies. Documentation . In many non-trivial ways, process-produced data differ from surveys. Some of the unique aspects regarding documentation are: definitions of the universe to which the data pertain, description and estimates of the errors of coverage and content, and requirements for a "codebook". Data transformation . Once a data set has been acquired by a researcher or an ar- chive, there are various difficulties one faces in transforming and analyzing the data. Many data sets are "ragged" rather than rectangular, thus requiring either intricate transformations or specialized statistical packages. Aggregated data require a high level of statistical sophistication on the part of the analyst. Finally, one will frequently wish to link data from various data sets. In doing so, one often runs into problems of comparability: stimuli may not be standard- ized across data sets; for identical or comparable stimuli, response categories are frequently not identical across data sets; different data sets may not re- late to exactly identical universes; finally, where spatial and/or temporal de- limiters are involved, they are often not identical across data sets. )i^slst NEWSLETTER vol. 1, no. 3 Data Acquisition for Archives Pierre Lacasse University of Sherbrooke The purpose of this paper is to raise a number of issues about data acqui- sition in order to narrow down the subject to the most important avenues for further discussion. With so diversified stock of social data being created, we need to ask "who will keep what, and on what basis"? To clarify this question we will particularly discuss three major inputs to the subject: (1) The col- lectors of data: Who are they? What types of data are they collecting? What data are we interested in? (2) The archives: What are they? What are their purposes in holding data files; and for what types of secondary users? What mandate do they have and what is their scope of interest? (3) The users for whom data is held: Who are they? Where do they stand? On what issues? What are their needs? How will they use the data held by the archives? After discussing these questions, we will take a look at what has been done in the data acquisition action group. In particular we will discuss the questionnaire prepared by the European section of lASSIST on "Archives data acquisition policies and problems." Consider the group's mandate, which requires proposing ways of linking collectors and archives, with the stress on defining the group objectives and conditions which must be met in order to close the gap between collectors and archives. Finally, we will try to suggest some possible courses of action in the short and long term. Organization of Data Archives Laine Ruus University of British Columbia The mandate of the lASSIST Data Archive Development Action Group is two-fold: the creation of a "procedures manual consolidating current archival organization- al, administrative, and personnel structures, procedures, and pol icies. . .to aid developing archives", and the organization of training workshops and seminars to aid in personnel training and professional development. The efforts of the Data Archive Development Action Group have been entirely concentrated on the develop- ment of A guide to providing social science data services . While response to past workshops has indicated the need for this guide, plans are for the moment in abeyance. Problems in data archive development occur at three levels. Intra-ar- chival problems are primarily administrative: planning and policy making, staff, acquisition, data management, technical and user services. Intra-institutional problems, impinging on the former, occur in the areas of planning, inter-depart- mental cooperation, support services, and funding. Problems at the inter-archi- val level result primarily from lack of adequate standards and conventions and from underdeveloped formal channels of information dissemination. l^$sist NEWSLETTER vol. 1, no. 3 standards for Data Documentation Dave Salley Statistics Canada In recent years, a number of initiatives related to the concept of data docu- mentation have been undertaken within the Canadian Federal Government. The motives for these initiatives have centered around the need to inform users on data avail- ability, the need to monitor duplication of data collection, the need to control questionnaire content and the need to assess the costs incurred in the collection, compilation and dissemination of federal information. While these motives are not exhaustive, they do represent a significant class of problems for consideration by the lASSIST Action Group. The shortcomings of most efforts to date have centered on the particular or special purpose nature of the systems. In many cases, a particular subject area or class of users restricted the approach severely. In short, little considera- tion has been given to the development of standard components of a Data Documen- tation system with a view to meeting general requirements. The reason for this problem may well be that information managers in the Federal Government have not yet developed a real appreciation of the need for standard documentation proce- dures, both as a tool for managers and planners of information activities and to assist "end data users". Thus, most initiatives have been quite ad hoc and nar- row in scope. However, one should not underestimate the useful technical ap- proaches and methodologies developed to date. As a departure point for the Action Group, it is suggested that Terms of Reference be developed which contain a strong "Standards" ingredient. Further- more, an intensive needs and benefits study should be undertaken with respect to the whole data documentation issue. Certainly, the group should bring together current work in a systematic fashion, but the development of standard approaches in the scope, technique, uses and purposes of data documentation systems will have far more useful results in the long run. Cataloguing and Classification of Machine-Readable Data Files; A Preliminary Report on the US lASSIST Classification Cataloguing Project Sue Dodd University of North Carolina The primary emphasis of the US Classification Action Group of lASSIST has been on establishing standards and on the study of library information systems as they may apply to social science data files. Some of the recent developments within the library system which the Classification Group is examining include: 1) the development of rules and guidelines for cataloguing machine-readable data files (MRDF); 2) the development toward the acceptance of the MARC (Machine-Read- able Catalog) record format as a universal standard for the automated biblio- graphic record; 3) the development of networks and on-line information systems which allow for multiple input and immediate retrieval of information; 4) the development of thesauri and controlled vocabularies for social science terms; and, 5) the development towards future considerations of a national union list of available MRDF and their location. i»i$sist NEWSLETTER vol. 1, no. 3 Given the current work and interest in cataloguing MRDF, the first task of the lASSIST Classification Group v/as to participate in an organized project de- signed to test the feasibility of cataloguing social science data files according to the American Library Association (ALA)'s designated Subcommittee to recommend rules for cataloguing MRDF. To facilitate the test, which was conducted by mail, a "Working Manual for Cataloguing Machine-Readable Data Files" was compiled based on an interpretation of the ALA Subcommittee's recommendations. The task required that the partici- pants select six data files, either numerical, text or program files, and proceed to catalogue these data using the information and guidance in the manual; apply subject descriptors on the content of the files; and complete an evaluation form. The outline of the project, the actual test, and the preliminary results will be discussed in this paper. Data Organization and Management Applications in Data Archiving Greg Morrison Social Science Data Archive Carleton University The formal mandate of the Data Organization and Management Action Group of IASSIST is used as the starting point for consideration of various kinds of problems with which the Group might concern itself -- problems which most data archivists will confront sooner or later. These include: the transfer of data- sets; the cleaning, editing, transforming, merging and sub-setting of data; the organization of data files for the above activities and for statistical analysis; the technical aspects of the documentation of individual datasets by codebooks; and, the technical aspects of the documentation of collections of datasets by cat- alogues or inventories. While an obvious organizing focus of the Group is computer software, it will try to define appropriate procedures in these areas, and to encourage the exchange of information about programs and procedures. The activities of the Group to date are reviewed, and some suggestions are made for the future.