New Tools, New Talents, New Titles: Mainstreaming Data Services1

Revised notes from a presentation to the Canadian Association of Public Data Users, University of Calgary, June 11, 1994

Alan H. MacDonald, Director of Information Services, The University of Calgary (Presenter)2

with Sharon Neary, Data Librarian, The University of Calgary3

ABSTRACT: It is time for data to join the mainstream of academic services, and to have access to the financial and human resources of the mainstream. This article describes the creation of a Campus Data Service at the University of Calgary, through the joint efforts of the Library and Academic Computing Services.

RÉSUMÉ: Il est temps que le secteur des données rejoigne le courant principal des services universitaires et puisse accéder à ses ressources financiéres et humaines. Le présent article raconte comment les efforts et la collaboration de la bibliothèque et des services d'informatique de l'Université de Calgary ont abouti à la création du Campus Data Service [Service de données sur le campus].

Thank you. I really am quite pleased to be here. Alben Barkley, Harry S. Truman's Vice-President once said, "The best audience is one that is intelligent, well educated--and a little drunk." Two out of three isn't bad!

On behalf of President Fraser and all of my colleagues in the Administration and in Information Services, it is a real pleasure to bring you greetings and to bid you welcome. I hope you enjoy your stay at the University, in Calgary and here under the Big Sky.

These are wild and woolly times in Alberta, times of particular challenge for those of us in the academy. I hope you will be impressed as am I by the quality of academic effort and professional service that you will see in the University sector of this province.

Truth in advertising requires that I declare at the outset that no one in this room knows less about data than I.

Also, in this multicultural audience, I confess to being a librarian-- thirty years on June 30--even if I have possibly served the profession by leaving it for the fetid cells of administration.

What I hope to do today is consider, from an administrative perspective, the background of data in Canadian institutions. More particularly I am interested in the role of institutional agencies such as libraries and computing services as opposed to those of individual scholars. I shall attempt to address the big picture with a bit of local colour.

I will also comment on the changes in service and administration that must be made if we are to intelligently accommodate a new class of information and a new role in analysis within a campus. It is time for data to join the mainstream of academic services and to have access to the financial and human resources of the mainstream.

Let me start at the beginning of the journey that has brought me to this conclusion.

In careers as in life, I suspect that when one is a minority one learns there are certain pleasures of independence in being the outsider--one is unencumbered by the crowd and the crowd's thinking. Given the choice, the role of the border collie is more fun than that of the sheep.

As one of the few at the University of Toronto Library School who had no desire to work in Toronto (or even Ontario) and as one of the few in the gender-challenged minority--it was fairly easy to find interest in the areas that were not along the well-travelled road.

One of these areas was the publications of government, an area which still produces paroxysms of panic in many of my colleagues and which I found fascinating probably because I had acquired some understanding of government and of certain aspects of the political process in my native Nova Scotia.

I believe that the ability of the publications of government to produce the response of catatonia in librarians had nothing to do with form or content but the perverse application of arcane cataloguing rules that produced unfathomable hierarchical author entries which seemed designed to assure that no librarian or patron would ever be able to break through to the wealth of information included in the publications of government.

Librarians like governments seem to have a perverse instinct for disinformation, public protestations to the contrary notwithstanding.

Very early in my career as a very young government publications librarian in the basement of the Macdonald Library at Dalhousie, I undertook the herculean task of organizing the shambles that was then called the DBS--Dominion Bureau of Statistics--deposit collection. All DBS means now is direct broadcast by satellite.

As an organizational challenge it was marvellous but I was of the lies, big lies and statistics school of observers, not users, of data. I had little comprehension of the content. Then I met Dr. Roy George, a professor of economics who was involved in the work of APEC--the Atlantic Provinces Economic Council.

Roy was on the other end of the scale--infectiously enthusiastic about data, he would regularly back his car up to the back door of the library to haul away another group of publications for the support of his analytical work.

When we talked about his work, in addition to the marvelous enthusiasm, there was palpable frustration about the lack of data resources and about the enormous time delays in publication (some things never change). He chafed at the restrictions created by unwanted homogenization and pre-processing of the published data. He would rail against the lack of ready access to raw data and even the lack of analytical tools. The tool of choice was a mechanical calculator. Computers were expensive behemoths and SAS was just an airline. This made an impression on me that came home to roost in 1989, twenty-five years later.

Statistics Canada, whether forced by Mulroney ideology or its own enthusiasm, was recruited to embark on an insidious program of double dipping--selling back to us the prepaid fruits of our own labours. This is like the commercialization of scholarly publishing that so frustrates librarians today. It meets the definition of colonialism--selling back the labours of the people which have already been paid for.

At that time I had been Director of Libraries at the University of Calgary for about ten years and had just become responsible as well for computing and media.

The suggestion came to CARL--the Canadian Association of Research Libraries--that we form a consortium to acquire academic access to the Canadian Census data which was about to become a commodity that would sell for an out-of-reach quarter million dollars.

I should comment that CARL was not only a excessively large committee (twenty-nine members) but all its members were directors of their respective libraries.

As Australian Senator Gareth Evens once suggested, a committee is often "a cul de sac into which good ideas are lured and quietly strangled." As a committee of directors, CARL had twenty-nine chairs, so getting a decision was somewhat difficult.

It would be safe to say that I had had little or no experience with data since my tutorials from Roy George twenty-five years before, in spite of my reputation as a manipulator of numbers with the mythical reputation for institutional alchemy of turning debits into credits.

This delightful ignorance was shared by the vast majority of my directorial colleagues who knew little of quantitative methods and felt no void in their existence for its absence.

Here I should express my admiration for Laine Ruus, an occasional blunt instrument who, in alliance with several fifth columnists like myself, was able to somehow convince, cajole and bludgeon most of the directors of CARL libraries to participate in the CARL Census data consortium. I think Roy George would have been proud of us.

Within this institution, digital data files were not a new information resource either for researchers or for their institution's libraries. A substantial collection of data files already existed on campus.

At the institutional level, officially the University had been a member of ICPSR--the Inter-university Consortium for Political and Social Research--for many years and faculty had acquired an assortment of their data files. In fact this was buried in the budget of the Faculty of Social Sciences and only attracted interest when annual budget exercises called for new sacrifices. For the most part, the data files graced private offices, undocumented and unshared. Even when I agreed that the Library would pick up half the cost, little changed.

The ICPSR, like a number of other aberrant acquisitions such as CIHM microfiche, multi-disciplinary videos and major memberships, had to be protected from the whining of disciplinary specialists by being buried in the Director's fund where fortunately there was no scrutiny.

Spatial data, both cartographic and LANDSAT imagery, had also been acquired for teaching and research and was used independently in various labs around campus. Institutional records for either type of data did not exist in the catalogues or listings of either the Library or Academic Computing Services.

In 1990 the Library participated in the acquisition of data from the 1986 Census of Canada and the existing data files for the Statistics Canada General Social Surveys. These strengthened the collection and also complicated the issue of how to organize and administer these expensive resources.

Informal methods of organization had proved adequate, not to mention cheap, up to this point. Demand was controlled by the necessity of accessing virtually all available data files via the academic mainframe. This requirement clustered users and potential users around a small number of early adopters among the faculty, staff and services, who referred new users and who provided oral history and documentation for the collection.

Three factors changed the balance in this informal but adequate information loop:

Pre-eminent among these influences was the phenomenal increase in the computing capacity available to faculty and students through the widespread conversion to UNIX and the proliferation of a network of workstations and microcomputers capable of greater and more independent processing of data files. Micro-based versions of statistical packages such as SPSS had also become abundant and popular thanks to site license arrangements that made the cost palatable to individuals.

The second important factor was the rapid increase in the scope of the data collection. Canadian publishing patterns for data had expanded in concert with the availability of computing power at the desktop level.

The purchase of the 1986 Census data aroused growing interest among researchers. A new, collaborative service was obviously needed since no single unit had the requisite combination of technical and information expertise to meet the demand.

A final and unexpected ally in the process was the increasing financial constraint on basic services of the University.

In data, as in many other areas, the shrinking budget forced all departments to consider a more cooperative approach to resources and resource rationalization. Until the early nineties, the comparative availability of funds made it acceptable for money to be spent on specialized services and collections which were independently enjoyed and profited the research of relatively few faculty and students. This could no longer be tolerated.

Yes, the data files languishing in faculty offices should have been properly documented and available to the campus community, but who was to be given the task? There were many 'shoulds' for our Library and its staff who, as is often the case in academic services, appeared to feel the sting of cutbacks before the teaching units.

The data archive was considered by many to be, at best, a marginal collection. The data enthusiasts on Library staff were eager to establish expanded service for this important information resource. Many other staff felt, however, that greater priority was commanded by aspects of Library service involving greater numbers of clientele: implementation of CD-ROM bibliographic databases, the management of journal cancellations and the evolution of basic reference service to name a few.

A certain data anxiety likewise reduced staff enthusiasm. For many involved in the administration of traditional information collections, quantitative analysis loomed like the revels of far-off pagan tribes: interesting to speculate about, but safest viewed from afar. If one paid appropriate tribute, as one had with government publications and maps and music, maybe they would just stay away.

The lack of appropriate infrastructure for sharing and promoting the files likewise contributed to their unavailability and underutilization. It naturally did not take long before administrators such as the Dean who was paying for half of ICPSR and the Director of Information Services/Director of Libraries who was paying for the other half as well as for other things began to question the use of resources to support such "expensive luxuries". The problem was clear: improve documentation and access to the data files or lose the resource altogether. Use it or lose it!

Within Information Services we were beginning the tentative connection of librarians with computing and media personnel and the inventive and the creative were beginning to see the possibilities. Data was one of the very first beneficiaries of this process.

Formal discussions on the creation of a data service had began at a time when the impact of these factors was sufficient to form a confluence. In 1990 we commissioned a consultant to collaborate with Library staff and others in preparing a report to advise us in the Administration--the Area Heads--the "Air Heads"--upon the future of the collection.

Considering the amount of institutional resources already committed to the collection, it was not surprising that the report recommended the creation of a Data Service cooperatively administered by Academic Computing Services and the Library.

Now, a positive recommendation may be one thing, reality can be quite another. As we all know reorganizing a university is harder than reorganizing a cemetery. The report provoked little response since the unit heads were simply unwilling to further commit already overextended staff.

At issue, however, was the credibility of any information service which did not include access to Canadian and other sources of statistical and spatial data. The demand for the data collection was undeniable and slowly increasing.

To relieve the impasse, the data conspirators proposed and implemented an interdepartmental pilot to investigate what the Data Service might look like--if only someone would adopt it.

The project focused on the 1986 Canadian Census data. In addition to their high priority for research on campus, the Census data files presented the technical challenge of being available in both tape and CD-ROM formats which researchers could select based on the complexity of their research. Through the pilot project, basic interdepartmental procedures were developed for referral service, as well as for accessing the tapes on the (then) mainframe which was to stay in service until early 1993.

The pilot project established that the data collection was both significant and labour intensive. For adequate support, both infrastructure and more formal procedures had to be developed. A further report was submitted outlining the requirements identified through the pilot project. The familiar administrative refrain was that the needs were legitimate and even admirable but were not matched by resources.

This pilot project did, however, achieve the important first steps towards shared responsibility and cooperation between two separate services of the University--the Library and Academic Computing Services. These units had not entered into any type of cooperative enterprise before since it simply had not been needed. Computing support of the Library up to that point had come entirely from the separate Department of Administrative Systems.

The only direct benefit to those involved was an expansion in workload, and the entire nature of the collaboration was unfamiliar territory. Was the work involved for the two participants coextensive? Was the partnership symbiotic? Would gains be experienced mutually and equally? How could progress be requested? How could another department's political sphere be respected and yet exploited?

Inertia was resolved by an event which caused a number of faculty and administrators to feel the hot sting of academic embarrassment. First, SSHRC introduced its requirements for deposit of data and we had no local place to put it.

Then, overcome by fiscal restraint the Faculty of Social Sciences indicated that it would not renew the fiscal arrangement for ICPSR. The benefits were too few for the cost. The lack of return was partially due the absence of a campus wide program to document, promote and provide access to the ICPSR files already acquired and those available though the ICPSR Archive under our membership.

Advancement through word of mouth was just not going to cut it in the new fiscal environment. Informing researchers that so far as anyone knew a required data file was available in Professor Brown's office, second drawer, left-hand filing cabinet was not sufficient resource sharing for the cost conscious.

However, ICPSR was a very sacred symbol for many and an important information resource for the rest. Abandoning our membership in the Club was a sign of how far the situation had deteriorated and of the drastic steps which had to be taken. The data conspirators struck and used this crisis as an opportunity to advance.

In order to maintain participation in the ICPSR, a membership consortium of ten academic libraries was formed. As so often happens in the history of institutions, all the internal lobbying for program improvements meant nothing compared to the motivation created by new external enterprises.

Participating in the Consortium meant integrated acquisition procedures and the reporting of collections to a union file of ICPSR holdings. A proper commitment clearly had to be made by the University of Calgary and so, shotgun in hand, the marriage took place and the Campus Data Service was formed--graced by a dowry of paltry proportions.

Formal establishment of the Data Service program had achieved a tacit admission that the collection and service were significant. Nevertheless, this was little guarantee of support for already over-committed staff and there was residual hostility among equally burdened services that this new mouth to feed would be at their expense.

Levels of service and the resources to support them were predicated on demand in our newly re-engineered and stricter University. Demand for even an important collection would be offset by inaccessibility. The strategy had to be to increase awareness of the collection among potential users and then hope that interest would inspire improvements.

If you build it, they will come!

It was easy to publicize the Data Service and collection, and this was quickly done among strategic academic departments, individual faculty members, library subject specialists and administrators.

The hard part was winning resources to develop the infrastructure-- when "they" come, they're certainly going to want a place to sit. Since the Campus Data Service was exclusively an internal service, a treasure hunt was begun among the university-wide maze of grants and funds for worthy academic projects.

An interesting ally and a benefactor was found in the Teaching Development Office, a new unit administering a five year grant of 3/4 of a million from the Royal Bank to improve the quality of teaching across the curriculum.

The Data Service was a natural for such a program. What greater opportunity to teach sophisticated research techniques than instructing students in the use of data files ? The highly user-defined analysis skills that data files allow were not only of obvious academic merit, they were much in demand by employers.

The Teaching Development Office provided a modest amount of funding for a Campus Data Service Project for teaching based on one specialized aspect of the data collection. It would focus on the use of spatial data files for teaching Geographic Information Systems. GIS-related activities happened to be an academic strength of the University of Calgary, one of the few Canadian universities with an established geomatics program. This was a worthy enough goal, but the true benefit of the project was development of the infrastructure we argued was an absolute requirement for the project.

Using the money and the leverage from the Teaching Development project, an enhanced interface was developed for the SCEPTRE software used at the time for tape mounts and file extractions. This interface, affectionately called DRAT, allowed all users of the AIX workstations providing basic computing for the campus to list files available from the tape archive and extract copies of anything they needed. DRAT processed extraction requests within minutes for single files, and within hours for multiple files.

This was revolutionary compared with the days previously needed for staff-arranged tape mounts and extractions. DRAT automatically provided sufficient disk space for processing even the largest data files. It also tailored accompanying SAS and SPSS command files to make the processing more efficient. DRAT represented a quantum improvement in service for the University data community.

In terms of interdepartmental cooperation, the Teaching Development Office project was critical. It had allowed the two previously autonomous member units of the Campus Data Service to work together as a coalition of forces to achieve mutually shared benefits. Despite the fact that the actual money involved in the grant was small, just a few thousand dollars, it paid for part of the programming costs.

The strategic nature of the project helped to protect a staff position from the menace of abolition. Some of the award was also spent on the acquisition of specialized spatial data files beyond the scope of the regular data budget. The alliance had proved advantageous indeed.

More subtly, the Teaching Development award final provided a certain deserved prestige for the units involved in the Campus Data Service. It was unprecedented for two non-teaching departments to receive such an award.

Improvements realized by the Teaching Development Project demonstrated to faculty that the Data Service was a credible ally in furthering their goals for developing effective teaching opportunities. The Data Service was subsequently asked to participate in a number of inter-faculty study groups which had formed to examine academic issues across the curriculum.

The composite view provided through these interdenominational groups identified two main priorities among faculty involved in teaching quantitative techniques. The first was a collection providing data files appropriate for a broad range of applications and levels of analysis. The second need was for a flexibility of service for instructors needing to accommodate large numbers of students and their correspondingly diverse skill levels.

Based on these needs, the Campus Data Service was able to negotiate new acquisitions more suitable for generalized learning styles. One of these was PCensus Canada, a politically productive joint purchase. PCensus was bought with the financial assistance of the Faculty of Management, a key war chest on campus.

Faculty assistance was also critical in the negotiation of a major acquisition of cartographic files. Participating in a GIS/LIS study group allowed the Data Service to begin discussions with the Land-Related Information Services Project of the Alberta government for the assisted purchase of Digital Elevation Models covering the Calgary Region and other strategic parts of Southern Alberta.

It was indeed a sign of the changing character of the times that these files were now considered essential for teaching and a resource that should be commonly shared through the Campus Data Service. The influence brought to bear by the faculty convinced Land Information Alberta to provide the data for academic use only at a fraction of the market price.

The development or more accurately the talking into existence of the Campus Data Service was inherently symbiotic exercise in mutual self- interest. The whole is greater than the sum of the co-operating parts.

The data collection and service clearly reflects the needs and priorities of the faculty, and the conformity of our goals have enabled us to negotiate from a position of strength.

From my position in the rarefied atmosphere above the struggle of the data service for sustainable existence, there are two other lessons.

One is that real use, not talk, is what attracts scarce resources these days. Competence does mean something. Co-operation does mean something. Cutting the crap is one of the critical skills of the information age.

The most important lesson is the convergence of talents. A university is a place of many technologies, ranging from those of Herr Gutenberg to those of Billy Gates. Over the years each has become a centre of activity, speaking its own language, competing with and ignoring the other. Even more compelling than the economic circumstances that have forced conjoint activities are the pressures that are being brought to bear on the fundamental characteristics of the priesthoods of the various technologies.

A convergence of skills can be observed resulting from the changes in technology and empowerment of individuals to apply the technologies within their realm of competence.

This convergence around networked computing and resources such as data will affect librarians and other library workers and all of the many other specialities that were segregated from each other in the past, either because of their technologies, or their traditions, or their formation. Whether we work in computing, media, telecommunications; curatorial activities in libraries, archives and museums; in many aspects of publishing, and the information provision roles of all groups, our role will change.

Because of the new tools and the new power they give our clients, traditional career planning is now an oxymoron. We must evolve old talents and find new ones. We will abandon old titles for new.

Librarians must abandon the passivity of the "come to me at my place" philosophy of service. Data service is resource and labour intensive and involves ongoing and complex, time consuming interactions with clients.

Computer specialists must recognize that care for the needs of a community is different than tending to the pursuits of individuals. Grasping the content of the data is important to the servicing of the needs of users.

Faculty and students must learn that use of data, yes even data in the raw, is becoming an essential part of academic preparation, even at the level of the undergraduate.

Our organizations expect us to sort these questions out fairly quickly so that we may efficiently exploit the opportunities of new technologies and the challenges of new resource reductions. This obligation to coalesce is occurring at a time when the philosophical and professional underpinnings of information service groups are also undergoing significant and stressful transformation.

The significant challenge is not the managing of the technological change. The challenge is to effectively manage change in service cultures and in the managerial cultures to bring together various styles, attitudes, motivations and perspectives--to create a new blended culture of reliable informatics service without producing service collapse, client rebellion or professional civil war. It's time to take most of the job designs and classification systems in technical areas back to the drawing board.

We who labour in information services and who believe in the idea of library, now stand at the edge of a new era. We must go to where the hornets live. We must move from the rhythms and pleasures of cross-country directly to the fears and excitements of heli-skiing. We are not alone, but we do not have the choice not to act. If we fail to act wisely, the utility of the academy to the new century will be seriously undermined.

American humorist Joe E. Lewis is reported to have once said, "You can lead a horse to water...but if you can get it to roll over and float on its back--then you've got something!"

One of the characteristics of the Calgary ethos is the firm belief that anything is possible--impediments are merely challenges to be met. Thus it is not surprising that at least some at the University of Calgary believe that at least some of our horses can learn synchronized swimming-- and they have.

Thank you for inviting me.

Through your tradition of co-operation, your powerful lobbying and your adroit trafficking in data, CAPDU demonstrates that it is a truly, albeit unruly, asset to information services and to scholarship in Canada. You are the way of the future.

Thank you to my colleague Sharon Neary for her professionalism and her passion for data.

My sincere appreciation and admiration to librarians Sharon Neary, Christine Hayward and Helen Clarke and to computing manager Janice Bakal for their role as the four wise women who saw the star of data and followed it faithfully through lands of the unbelieving and even the hostile.

They are among the first who are truly making the professional transformation: they understand the new tools, they have the new talents, they will be among the first to achieve the new professional titles. Through their faith, tenacity and sharp elbows, data at the University of Calgary is on its way to being a full information partner with text, images and sound as essential tools for the scholarly and the national interest.

Thank you.


Letters to the Editor / Lettres au rédacteur en chef