iassist quarterly - 5

Educating the
data user:

the data archivist and
bibliographic
instruction

by Bliss B. Siman'

Associate Professor, Data Archivist

Baruch College, City University of New York

Information specialists have long been aware

that graduates usually leave academia with onh

a rudimentary knowledge of information

strategies and resources. The expansion of

information access opportunities has not changed

this situation radically. In the area of

information about public data sources, the gap

between what students know and what they

could know, is tremendous. At Baruch, this

problem is particularly important because there

are few areas more dependent on information

access and utilization, particularly in

machine-readable formats, than business, and

few academic disciplines, therefore, in which

this "information gap" has greater significance.

Since the Librar\' at Baruch College provides

'Presented at the International Association for

Social Science Information Service and
Technology (IASSIST) Conference held in

Washington. D.C., May 26-29, 1988

access to information in almost all cunently

available formats: print, microform, audiodisc,

microcomputer diskette, CD-ROM, and

machine-readable data files, its instructional

program has tried to include all of these

information technologies in its workshops and
courses. Sometimes this has come about quite

by accident

When the library began to collect

machine-readable data files (MRDF), the

activit)- was assigned to me based on my
expressed interest Since it was not a full-time

assignment, I continued teaching in the

bibliographic instruction program. In retrospect

it was a fortuitous combination of

responsibilities. Working with data users, I

became aware of the gaps in their information

skills, the ver>- skills I was teaching in my
sections of the "Information Research in

Business" cotirse. Most data users were capable

of using SAS or SPSS.x to analyze their data,

and many could write elegant COBOL
programs, or download data into LOTUS, but

almost none had training in how to search for

and identify quality- data. Few graduate

students or faculty members were sufficiently

aware of the vast potential of public data for

their research or teaching.

Surprisingly, many sophisticated faculty

researchers continued to use datasets first

introduced to them by their Ph.D. mentors

simply because they were unaware of

alternatives. Few were familiar with the varied

storage media for data and how these cotild be

eflectively combined. For example, a facultv'

member working on a project using the Census

of Population and Housing on magnetic tape

might be totally unaware that portions of the

work cotild be better accomplished using the

same data in print Clearly there were many
exceptions to this bleak picttire, but data users

were not able to use the wide variety of sources

available due to lack of information skills.

Finding the means to overcome this deficiency

became an important objective of a combined

Summer 1988



6 - iassist quarterly

Data Archives/Bibliographic Instruction Program.

At the same time that the Data Library was

being developed, other sections of the

instructional program were beginning to include

numeric data as part of the information process

being taughL Online computerized retrieval

services began to provide access to numeric

data, and instruction in print sources of data

was increasing in sophistication. Naturally, with

all these programs being developed in the same

division, there was a good deal of beneficial

"cross-fertilization" in the planning and

implementation of the instructional programs as

well as the data service. Recognition of the

competitive importance of numeric, quantitative

information in business reinforced our desire to

equip Baruch graduates, many of whom are the

first generation of their families to go to

college, with the data access skills they lacked.

With respect to information literacy, business

requirements were growing and we wanted to

be sure that our students were prepared. Our
method of attacking this problem also reflects a

basic philosophy that information resources in

general are underutilized due to the public's

lack of training and education in research

resources and research skills.

Consistent with this philosophy, the Data

Resources Service immediately organized a

seminar series to introduce users to numeric

data sources. These seminars, which are still

ofTered on a regular basis, were organized by

subject field and publicized to faculty and

graduate students. Unlike many similar

seminars given in computer centers and data

archives, these presentations included not only

discussions of important datafiles, but also

information on the same data in print, or the

major print reference works in the same field.

Often the data being discussed was also

available through online vendors or on

microcomputer diskette, and the criteria for

using the various media were presented with

practice problems to illustrate important points.

The objective of the seminars was not to

provide a "shopping list" of datafiles, but rather

to equip the audience with the skills to find

and evaluate the data needed for a particular

project Most importantly, no seminar failed to

include information on important sources,

whether data archives, government agencies or

private vendors, of new data in the field, both

general and specialized resources.

Models for locating data in a new field were

discussed. These strategies paralleled those

taught in the "Information Research in

Business" course, in which students are

equipped with the ability to identify sources in

new fields as a basic part of modem
information retrieval skills. The reception this

information received underscored the need the

modem data user has to identify quality data

when beginning research in an unfamiliar area.

For new graduates in entr>'-level positions, such

skills can be invaluable. Many of these

seminars included an online demonstration of

some dataset using, where appropriate, SCSS
(the interactive version of SPSS). The object of

these seminars was to present MRDF in the

context of other information sources, as well as

to educate users in the available resources.

Initially, the target audiences were those who
were already data users, individuals who had

used data and were probably aware of the

limitations of their knowledge of public data file

availability. Because they were knowledgeable

about data, they particularly appreciated training

in strategies for finding data sources when the

usual avenues were unproductive. These users

were also receptive to information on selection

of format of data, because they were also not

aware of the many choices that could be made.

Widening the scope of data knowledge and

information retrieval skills among attendees was

the primary objective of the first seminars. This

objective remains cenual, although the seminars

today are often presented to those who don't

know very much about data. Of necessity, lists

of data files are distributed when the seminar is

attended by new data users. Although analytic

Summer 1988



lassist quarterly - 7

issues are sometimes discussed, research

methodology is never the focus. When Baruch

became the coordinator of the ICPSR
membership for the Cit>' University*, these

seminars were extended to the entire University

with equal success.

Building on this basic format, we have

experimented with workshops which include

specific demonstrations of ciata sources that

appear in online format, perhaps in print, and

as machine-readable data files. For example,

the Trinet database, which contains market share

data for companies, is issued online and on

magnetic tape. The Computer Search Services

Librarian and I conducted this seminar to

demonstrate the pros and cons of using nimieric

data online versus on tape. Baruch faculty and

graduate students have become informed users

of online information services, partly because

these services have been free, and partly due to

the excellent assistance that has been available

to online users. But the consequence has been

an expectation among the clients of the online

services that all information, bibliographic and

nimieric. will be foimd neatly set up in database

format, easily retrieved by a packaged quer\-

language. Semmars in which the nuances of

using different formats has been presented, have

increased the sophistication and information

proficiency of faculty and students, especially

the graduate students.

On the other hand, one can be too successful in

encouraging data users to become aware of the

wide variety of data sources available. The

demand for data increases and the multiplicity

of data requested creates budgetary problems.

Since our graduate students are primarily in

business fields, their need for expensive

financial and economic data caimoi always be

met within the Data Resources Service budget

Finding ways in which to meet their needs is a

constant challenge. Recently the number of

very specific, very limited financial daiaseis

being requested (example: 10 years of currency

prices for five specific countries) went far

beyond the capacity of the Baruch computerized

information services. Methods of creating

datasets, rather than purchasing them from

private vendors were explored.

Using an online vendor, I. P. Sharp, we
downloaded small amounts of data to meet

specific needs. In order to make the data more
widely available, they were also uploaded to

magnetic tape on the mainframe and listed in

the data archive holdings. As another

alternative, microcomputer diskette data sets in

Lotus 1-2-3 format were also created. We are

hoping that knowledge of these data will

encourage users to plan their graduate theses,

etc., around data that are available rather than

devising projects dependant on costly new data

sets. However, philosophically, it was important

not just to create these data sets ourselves, but

to educate users to the benefits of this

technique, a technique that could be very useful

to faculty in their current work and to students

in their future business roles. Consequently, the

entire process was demonstrated at a data

seminar. The demonstration included how the

data were identified, accessed online,

downloaded to diskette, uploaded to magnetic

tape and then accessed using SAS. At the same

time, the use of the data at each intermediary

step was discussed, providing greater depth to

the users' understanding of the pros and cons of

each technique.

As the variety of formats increases, we see

increased possibilities for this type of seminar.

For example, seminars on the census, could

present alternative methods of searching for data

files, beginning with bibliographic searches

online, as well as the choices to be made
among formats: print, microfiche, Cendata,

machine-readable data files, diskettes,

CD-ROM, etc. Census data are very much
underutilized at many colleges of the City

University, and we believe such instruction will

assist users in identifying valuable data not

previously considered.

Summer J988



iassist quarterly

Most of this discussion has focused on the kind

of training given sophisticated data users to

improve their information skills. But the

Library Instruction Division and the Data

Resources Service have been equally engaged in

a dialogue concerning the competence that

undergraduate students should master in order

to attain "information literacy" with respect to

public data sources. Actually, this dialogue is

part of a continuing discussion and re-evaluation

of the entire bibliographic instruction program,

necessary in a rapidly changing information

environmenL There is not yet consensus on

what represents an adequate set of information

skills.

Although the Library Instruction Division (LID)

offers a basic course in information research in

business (Library 1016, Information Sources in

Business) the core of materials covered has

changed over time. In fact, although all

instructors teaching the course use the same text

and give uniform exams, there is considerable

flexibility in planning the curriculum.

Instructors are encouraged to experiment with

materials and share their successes or failures

with colleagues. As the importance of public

data sources became recognized within the LID,

the department began discussing methods of

including this information resource in the

curriculum. Frankly, a complete answer has not

been arrived at, although several configurations

have been used.

Cleariy, undergraduates do not need

sophisticated knowledge of public data, but they

do need some awareness of the role of raw data

in the information process, of the availability of

data for secondary analysis, and where data can

be obtained. The inclusion of data sources in

the curriculimi reinforced some of the

conceptual goals of the course as well. Master}'

of the development of search strategies is an

important objective of the course, and data files,

due to the lack of bibliographic control, provide

an excellent example of non-standard

information search strategies. The process of

searching for data is, of necessity, very different

from the index searching model which

undergraduates come to assimie is relevant in all

situations.

Awareness of raw data files and of what

constitutes an authoritative soiu^ce in this field

reinforces another objective of the course, that

of enhancing the students' ability to evaluate the

information they consume on a daily basis.

Polls and government reports based on data are

constantly presented in newspapers, magazines,

etc. Questioning the data on which these

reports are based is an important attribute of

the educated citizen in private or professional

life. By including data as an information

resource in the research course, we are

equipping students the better to handle this

important task.

Working with my colleagues, I developed

several different instructional modules which

presented numeric data and the different

formats in which they come. The lectures

contained explanations of machine-readable data

files and the nature of secondary analysis.

Basically, 1 wanted students to understand how
data are used in business and academic research

and the retrieval methods used to find data

files. These lectures were tied to different units

of the course, depending on the instructor:

marketing in one case, social science research in

another. In each case, the lectures built on

what the students had already learned about

information access and retrieval as well as

emphasizing the evaluation of sources. The

lectures always included a demonstration of data

use in an area that would pique students'

interest

During these demonstrations, students were

encouraged to participate in the development of

a hypothesis and the testing of it with data at

hand. Although this technique was borrowed

from modules developed for sociology courses,

the emphasis was placed on information

dissemination and evaluation issues rather than

Summer 1988



iassist quarterly — 9

on research methodologies. These

demonstrations were very effective in making

students aware of the whole process of research

transmission, but it is not clear that the best

combination of subject year and lecture format

has yet been found.

Despite their success, these lectures have not yet

become a standard part of the course for

several reasons. As with every academic

institution, we are understaffed. It is not always

possible for me to do these lectures at the

appropriate time in each instructor's syllabus.

Nor do the instructors necessarily have the time

in a single semester course to devote a full

lecture to data files. The amount of material

we would like to cover in a course is far

greater than can be managed in a semester.

Topics are constantly being juggled as we search

for the optimum mix. However, the

Department has a commitment to including the

identification and retrieval of data files in the

basic information course because we feel that

use of these sources is an important information

skill. Data users must first be able to find data,

and the bibliographic instruction course can

equip them with the skills to do the job

efficiently and cost-effectively.n

Summer J988