lASSIST Newsletter, Vol. 2, No. 4 (Fall 1978)

TOWARD CREATING THE PROFESSIONAL DATA LIBRARIAN

Al ice Robbin

University of Wisconsin - Madison

Abstract

This article describes the need for training the professional data
librarian, archivist, and information scientist in a framework of social
science research and applications and library and information sciences.
The University of Wisconsin-Madison Intersession 1978 course was designed
to meet this need. The course is described.

evaluation involves the collection.
Since President Lyndon Johnson dissemination, and secondary analy-

called for a "War on Poverty" in sis of statistical machine readable
his State of the Union Message of data files (MRDF) . Some of these
January 1964, we have seen an expo- files are produced in the course of
nential growth in the production of individual research projects; oth-
statistical information in order to ers, by organizations in the course
allocate resources at the national, of their operations; and still oth-
state, and local levels of govern- ers, by ongoing data collection
ment; to plan, audit and evaluate efforts funded by a consortium of
the distribution of resources; and data users,
to ensure adequate planning for
human needs and an equitable deliv- While a substantial portion of
ery of benefits. Almost all fed- these data are probably not useful
eral legislation has included for reanalysis, a great body of
requirements to collect, analyze, data continues to be useful for
and report the findings of data research related to public policy
gathering. Recent trends in fed- analysis, planning, and evaluation,
eral reporting requirements suggest While some of the data have been
an even greater increase in the transferred to national archives
rate of production of statistical and data centers whose major func-
data during the next decade as the tions are to preserve, describe,
need increases for more information and disseminate these data, many of
by policy planners and analysts in these data files which are poten-
both the public and private sec- tially rich sources of information
tors. remain outside the public domain.

Most of the data have been col- The problems of access to infor-
lected as part of the administra- mation about MRDF, the quality of
five record keeping process of data, and the need for good docu-
governments, but a large portion of mentation describing MRDF have been
the data gathering has been funded issues discussed by secondary ana-
as part of the research, policy, lysts and data archive staffs for a

and evaluation activities of the number of years. One aspect of the
Federal government. In the social MRDF problem, however, has not been
sciences, an increasing amount of sufficiently addressed, and that
scientific research activity as has been the insufficient and
well as policy planning and

- 95 -



lASSIST Newsletter, Vol. 2, No. 4 (Fall 1978)

inadequate concerted national
efforts to facilitate access to
MRDF through the development of
training programs for professional
librarians and information scien-
t ists

.

When IASSIST was established in

1976, it was with the recognition
that members of data archives and
libraries needed a vehicle to com-
municate information about organiz-
ing, managing, and disseminating
machine readable data files. In
the IASSIST Constitution, the
'Objectives' include the establ-
ishement of training courses for
data center personnel (NEWSLETTER,
1(1), 1976). This objective not
only represents the recognition
that data center personnel need
assistance, but that established
data services have the potential
for providing training programs to
assist others in understanding the
nature of MRDF and the special
problems of organization, manage-
ment, and dissemination associated
with this medium. The Inter-univ-
ersity Constortium for Political
and Social Research responded to
this objective by holding two Work-
shops on Data Library Management in
the Summers of 1976 and 1977, as
adjuncts to their regular summer
program (Rowe, 1977). The Work-
shops were taught by Carolyn Geda.
Alice Robbin, and Judith Rowe.
Participants included trained data
center personnel and professional
librarians who wanted to become
more informed about MRDF and inte-
grating them in a library collec-
tion.

These Workshops were a source of
satisfaction to their instructors
and a good deal of information was
communicated and exchanged. But it
became increasingly obvious that a

Workshop was not the best structure
in which to communicate a concep-
tual framework for organizing and
managing MRDF. There were few

incentives to utilize the computing
and data processing facilities, and
within the time constraints there
was little possiblity of dealing
with major social research and
applications concepts needed to
understand statistical MRDF, data
base management, and organizational
behavior. In addition, if the
ideas were to reach an audience who
could most directly benefit from
this learning exper ience--the pro-
fessional librarian—the course had
to be taught in a university
library school environment, and
integrated in the library school
and social science departments'
cur r icula.

Why two different departments
and indeed ones which rarely commu-
nicate with one another? Although
social scientists continue to
demonstrate negative attitudes
toward the library profession, it

is the library which has the exper-
tise in our society to organize and
disseminate information. Libraries
are a natural environment for MRDF,
because it can be treated as an
additional informational resource,
albeit in a different storage med-
ium. Library schools train profes-
sionals to handle a variety of
informational resources. Courses
in library automation, systems
design, information storage and
retrieval, and on-line biblio-
graphic data bases are becoming
integral offerings of library
schools throughout the country.
Thus, a ' library school is the
natural setting for introducing the
concept of numeric or statistical
MRDF. But while library schools
routinely address problems of tex-
tual data in machine readable form,

they have not addressed the prob-
lems of numeric MRDF and future
professional 1 ibrar ians and infor-
mation scientists are ill-equipped
to serve the quantitative bent of
today's social scientist.

96



lASSIST Newsletter, Vol. 2, No. 4 (Fall 1978)

On the other hand, most social data library specialists, and
science disciplines offer at least computer specialists at the Univer-
one course and at a growing number sity of Wisconsin-Madison. Because
of institutions, major course it would be offered during Inter-
offerings consist increasingly of a session 1978, it could be viewed as
number of areas related to quanti- a potential course for in-service
tative social research. Social training for library and archive
science departments train students professinals who wished to become
in methodology, statistics, survey familiar with this informational
research, modeling and simulation, resource,
data handling, and the like, sub-
jects which provide the basis for The course was cross-listed by
understanding the construction and the School of Library Science and
analysis of MRDF. In fact, person- Department of Economics, and for a

nel of most of the data services in variety of bureaucratic reasons was
North America and Western Europe entitled, "Micro Data Collection
have been (and continue to be) peo- Methods in Economics." Funding for
pie who trained in one of the Intersession 1978 was made possible
social sciences. They have not with the generous support of the
been professional librarians and UW-Madison, which encourages its
have been slow to recognize that faculty and staff to use Interses-
the tasks they perform or the prob- sion as an opportunity to develop
lems they encounter in organizing, new courses in response to per-
managing, and retrieving MRDF (and ceived scientific and social
information about MRDF) are tasks changes within and outside the
which have been traditionally per- University. The course was
formed by reference librarians who designed to meet the needs or

work with other media (see, Carmi- interests of social scientists,
chael , 1978; and Robbin, 1978). users, and generators of numeric

machine readable data, and profes-
A course which integrates the sionals engaged in information ser-

theoretical foundations of library vices, whose present or future res-
and information science and social ponsibil i ties might include
science research is therefore one managing large numeric data files
which potentially speaks to the or providing data services or

formation of a professional infor- information about numeric MRDF to

mation scientist and manager of users. The objective of the course
numeric or statistical MRDF. With was to provide the student with the

this in mind, in September 1977, I underlying principles of access to

recommended that the Data and Com- and management of MRDF in a library
putation Center, of which the Data and archive setting. Students were
and Program Library Service is a introduced to social science
part, design a course which would research and applications, data
respond to a perceived need to collection techniques, computing
train professionals (both in the and data processing, statistical
social sciences and library and analysis and file handling, policy
information sciences) to deal with issues and problems regarding data
the explosion of information in libraries, and bibliographic docu-
machine readable form. It would be mentation and control of numeric
an interdisciplinary, graduate MRDF. Problem-solving was an

level course, one semester in integral part of the course, and

length, which would draw upon the included exercises in statistical
expertise of social scientists, analysis and building a biblio-

- 97 -



lASSIST Newsletter, Vol. 2, No. 4 (Fall 1978)

graphic data base of numeric MRDF

.

reviews, and archive and library
In addition, there was a heavy dose problems and their relationship to
of daily readings on which the lee- MRDF).
tures and discussions were based
and a final paper (in lieu of an Primary r esponsibl i ty for the
examination). course was in the hands of Martin

David, Department of Economics and
Lectures included (1) introduc- Director of the Data and Computa-

tion to social science research tion Center; Alice Bobbin, Head of
methods and applications, introduc- the Data and Program Library Ser-
tion to statistical processing and vice; and Al Schubert, Head of the
file handling, orientation to data Program Consulting Service. Mem-
library and data processing and bers of the Madison Academic Com-
computing facilities; (2) introduc- puting Center (MACC) contributed
tion to systems analysis: data their expertise during the course,
library as an informat ionmanagement as did experts in data base manage-
system, complex data bases, data ment from the Department of Lands-
base management systems (DBMS)

,

cape Architecture and the Center
networking; (3) selected policy for Demography and Ecology. Le,c-

issues concerning MRDF and the data tures on social science research
library and archive; (4) biblio- and data collection were given by
graphic documentation and control an economist, sociologists, and a

of MRDF; (5) planning a data survey methodolog ist

.

library and information service for
MRDF; (6) special issues of concern Fourteen registered students and
for MRDF and data librarians: auditors began and completed the
copyright, strategies for file course. One half of the students
preservation and handling. Practi- were professional archivists (From
cal exercises included (1) statis- the Wisconsin State Archives) and
tical analysis of a MRDF specially library students and the other half
prepared for the course, using the were graduate students (primarily
SPSS package; (2) use of a data from developing countries) in eco-
base management system; (3) use of nomics, sociology, political sci-
networks; (4) data library proce- ence, business, and history. One
dures; and (5) bibliographic docu- of the students is a professional
mentation and control (building a data librarian from the University
bibliographic data base of informa- of California at Los Angeles,
tion for MRDF). The Independent
Project (final paper) assignment Most of the students had never
was a choice of (1) designing a worked with a computer before, but
research problem, carrying out 1 im- within a few days were keypunching
ited statistical analysis on the control cards and submitting sta-
data file prepared for the course, tistical runs. Although the course
and reviewing the findings in a was highly concentrated and very
short paper; (2) writing a short demanding (students attended lec-
analysis of the problems of creat- tures from 8-11 a.m., Monday
ing a bibliographic data base for through Friday, and worked on
MRDF; and (3) designing an indepen- assignments with the help of a

dent project with the approval of teaching assistant from noon, often
the instructors (most chose to do until 10 p.m.), enthusiasm and com-
this and selected a wide range of mitment never flagged. It was an
topics on confidentiality and pri- exciting time for instructors and
vacy, content analysis, book students. A detailed evaluation

- 98 -



lASSIST Newsletter, Vol. 2, No. 4 (Fall 1978)

instrument was completed by the
students. The recommendation was
that the course be integrated in

the University's curriculum and
extended into a full four week pro-
gram during Summer School or the
regular academic year.

The response was so positive
that DACC submitted a request for
refunding for the Intersession 1979
program. (At UW , all Intersession
courses go into competition for
funds.) We have recently learned
that DACC has successfully competed
for funds and the course will once
again be offered and cross-listed
by the Department of Economics and
School of Library Science. Several
changes are anticipated on the
basis of what the instructors
learned last year. First, the
title of the course has been
changed to reflect the actual
course contents. It will be offi-
cially titled, "Management of
Machine Readable Numeric Data for
the Social Sciences." (People were
more than a little mystified last
year to learn that "Economics 615,
Micro Collection Methods in Econom-
ics," would deal with the subjects
I just described.) The import of
the title change and new course
number(s) should not be underesti-
mated: While we have no assurance
that the Divisional Committee of
the College of Letters and Science
will approve a new course offering
(these changes are very hard to

come by), it does suggest the
recognition of the need to respond
to information and technological
changes in our society; universi-
ties must continue to broaden their
course offerings to meet profes-
sionals' needs and to respond to

technological change. Indeed, the
concept of data services, libaries,
and archives has too long remained
the purview of special support
facilities within the university
and the notion that services to

preserve and disseminate

statistical MRDF are a function of
the general or special library
within the university or in both
the government and private sectors
has not been widely accepted.

Second, the course structure was
too demanding, for both the stu-
dents and instructors. The work-
load will be reduced. Rather than
building a bibliographic data base,
time will be spent analyzing the
structure and syntax of its con-
tents. Some of the time allocated
to the data base will now be
devoted to exercises in documenta-
tion and control and records man-
agement of MRDF (known earlier by
its misnomer, "accessioning").
Less time will be devoted to under-
standing the development of statis-
tical software and more time to
understanding problems (through
statistical techniques) of statis-
tical data (assaying the quality of
data). The number of required
readings will be reduced. However,
the basic structure of the course
remains the same. Once again, a

set of instructional materials will
be created, but its size (242 p.)
will be reduced

.

What is evident from the ethusi-
astic response of the students and
instructors is that the course DACC
offered last year is much needed.
It responds to the recognition that
the information explosion must be
managed. Access and retrieval of

the vast quantities of information
in statistical machine readable
form are becoming important issues
in the information and library sci-
ences' professions and social sci-
ence disciplines. The course also
demonstrates that the complexity
and dimension of information in

numeric machine readable form are
such that no one individual has the
expertise to teach future profes-
sionals to organize and manage col-
lections of statistical MRDF;

rather, the approach to teaching

- 99



lASSIST Newsletter, Vol. 2, No. 4 (Fall 1978)

must be a

pi inary o
social re
archiv ist
communica
t ise . I

that Scho
mation Sc
numeric m
the non-1
a place i

n integ
ne, whe
searche
, and c

te thai
n the
ols of
iences
ach ine
ibrar ia

n their

rated
re th
r and
omput
r par
f utur
Libr

will
read

n pro
cur r

, inte
e quant
analys

ing spe
ticular
e i t i

ary and
recogni
able da
fession
icula .

rdisc i-
i tat ive
t, data
cial i St
exper-

s hoped
Inf or-

ze that
ta and
al have

NOTE: For further information on
the University of Wisconsin-Madi-
son's course, "Management of Num-
eric Machine Readable Numeric Data
for the Social Sciences," to be
offered during Intersession 1979
(May 29- June 15, 1978), write the
Data and Computation Center, 4452
Social Science Building, 1180

Observatory Drive, University of
Wisconsin-Madison, 53706.

References

Carmichael, Nancy. "Commentary on
Librarians and Machine Readable
Data Files." Special Librar ies
69(8), 1978, 306-07.

Robbin, Alice. "The Impact of Net-
working on the Social Science
Data Library." lASSIST News -

letter 2(1), 1978, 3-13.

Rowe , Judith. "Training the Pro-
fessional Data Librarian." In

H. D. White (ed.), Drexel
Library Quarterly 13 (1^) , 1977 ,

100-108.

100