MACHINE READABLE ARCHIVES:

USER SURVEY
by

SUE GAVREL

Machine Readable Data Archives
Public Archives of Canada

This article is an abridged version of a

Stavko Manoj'loviah of the Social Science
Western Ontario, London, Ontario, Canada.

Background

In November, 1981, the Machine
Readable Archives Division undertook
an analysis of the requests and
inquiries it had received for its

services over the last few years.
It became evident during the analysis
that although the division provided
a great deal of information on the
establishment of machine readable
programs, procedures and practices
used to archive machine readable
records, the actual holdings of the
division were not being used. There
were several possible reasons for the
lack of use of the files, but the
main conclusion from the analysis was
that the MRA did not have sufficient
information on who the user population
was and therefore was not distributing
information about its holdings to the
appropriate areas. The users of the
traditional archival records were not
the same as those who would use
machine readable data. It was decided
that a survey of our existing users
and potential users should be under-
taken.

The Social Science Computing
Laboratory of the University of
V/estem Ontario was contracted to
undertake the survey. Several steps
were involved: the drafting of the
questionnaire, the collection of the
data and the analysis. The project

report written by S. Paula Mitchell and
Computing Laboratory , the University of

spanned the period February 1, 1982

to October 30, 1982 and the total cost
of the project was $12,478.07, which
included staff charges, computing
services, printing, stationery and
postage.

The main objective of the survey
was to obtain information on the

needs of Canadian social scientists
who created or used machine readable
data. The results of the survey would
be used by the MRA to help plan an
expanded public service program and

determine priorities for the alloca-
tion of resources. A secondary aim
has been to identify sources of

machine readable data in Canada.

Specific objectives of the survey
were as follows

:

• to provide a measure of the relative
usefulness of a number of products
and services to the user community;

• to provide an indication of the

preferred mode of acquisition for

machine readable data files and

the preferred type of codebook;

• to identify the location of machine
readable data files for the pur-

pose of expanding a union list of

machine readable data files in

Canada and identifying files of

historical research value and
national significance falling



within the archival mandate of the

MRA;

• to provide an indication of the

public's awareness of the holdings
and services of the MRA;

• to provide an indication of the

need for a national organization
to coordinate selected information
services for users of machine
readable data (including special
workshops and/or conferences)

;

• to provide a mailing list of

individuals interested in receiv-
ing the I'tRA's publications or

announcements; and

• to identify the need for products
or services not covered by the

first objective noted above.

Survey Population and Data Collection

The survey population was derived
from two social science information
systems, the Canadian Register of

Research and Researchers in the Social
Services (Social Science Register)
developed by the Social Science Com-
puting Laboratory, and the Canadian
Directory Service of Social Scientists
(SocScan) , established by the Social
Science Federation of Canada between
1976 and 1978. The Register super-
cedes the SocScan database.

The target population was derived
from these two data bases by selecting
variables defining methodology and
research orientation. Although the

target population does not include all
social scientists who are current or

potential users of machine readable
data, the large number (4,065) does
provide a good representation of users
of social science data in Canada.

The questionnaire, designed to be
as simple and short as possible,
contained 15 questions. Data were

collected using a mail-out/mail-back
process. A covering letter explaining
the objectives of the survey was sent
with the questionnaire. A follow-up
reminder was sent to all non-respondents

three weeks (15 working days) after the
initial mailing.

FINDINGS

Survey Response

As mentioned earlier, 4,065 ques-
tionnaires were sent out. Of the 4,065
questionnaires, 82 were returned as
undeliverable mail, thus making the
actual survey population 3,983.
Response to the survey was very good
with an overall response rate of 48.95
percent, or 1,950 completed question-
naires .

Survey Results

The following section provides a

brief summary of results. Findings
are reported on a question-by-question
basis.

Ql. Prior to receiving this question-
naire, were you aware of the

activities and services of the
Machine Readable Archives?

The results of the survey indicate
that awareness of the activities
and services of the MRA among social
scientists is low. Of the 1,930
respondents answering this question,
14.2 percent indicated that they
were aware of the Division's acti-
vities and services prior to

receiving the questionnaire packet.

There was no significant difference
in the level of awareness between
respondents in the academic and
non-academic sectors (14.2 percent
and 14.1 percent, respectively).

02. Do you use machine readable
(computer) data in your research,
teaching, or other activities?

Of the 1,908 social scientists who
responded to this question, 1,293
(67.8 percent) reported that they
are users of machine readable data,

and 615 (32.2 percent) reported that

they do not use machine readable
•~ ~~~,—^ continued



data. Given the quantitative
orientation of the survey popula-
tion, it is somewhat surprising
that close to one-third do not use

machine readable data. However,
the very high interest shown by

this "non-user" group in receiving
information on the services and
holdings of the MRA (76.8 percent
wished to be on the MRA mailing
list) suggests that the survey was
on target in terms of reaching
quantitative social scientists who
are potential users of the MRA.

It is possible that some re-

spondents in this "non-user" group
found the question ambiguous:
either limited the inclusive time
frame of "use" to the present, or
interpreted the question to mean
use of secondary data rather than
their own primary data.

The distribution of users of
machine readable data geographically
and by sector of employment is

proportional to the distribution
of all respondents in these two
respects. Except for Ontario and
Quebec, the percentage distribution
by province of social scientists
who use machine readable data does
not vary by more than plus or minus
0.6 percent relative to the percen-
tage distribution by province of
all respondents. In Ontario, there
are 1.9 percent more users relative
to the percentage of social scien-
tists from Ontario constituting the
total respondent group, and 3 per-
cent fewer in Quebec.

The distribution of users by
employment sector (academic/non-
academic) varies by only 0.3 percent
relative to the sectoral distribution
of social scientists in the respond-
ing population.

Q3. Is there a facility within your
department or organization which
provides information and/or other
services for machine readable data?

1,269 users of machine readable data
responded to this question. A very

high percentage (93.2 percent)
reported that there is a local
facility which provides machine
readable data services; 86 users
(6.8 percent) reported that there
is no local facility providing such
services

.

04. Please identify the facility within
your department or organization
which provides information and/or
other services for machine readable
data.

More than half (53.1 percent) of

1,172 users who are availed of a

local facility are serviced by a

computing center only. While the

range of services provided by any
one computing center may approach
the full range of information,
access, and utilization services
provided by specialized local
facilities, computer centers typi-
cally provide limited support ser-
vices in these areas. It is a

significant finding that such a

high proportion of machine readable
data users are availed of limited
informational support services for
finding and using machine readable
data. 369 Respondents (31.5 percent)
indicated that more than one facility
provides local machine readable data
services. 25.1 percent of users
have a specialized local data support
unit (such as a data archives) avail-
able to them. For 23.3 percent of

users, the local research library
integrates some level of service for

machine readable data with other
services

.

Q5. Please indicate the potential use-
fulness of each of the following
information products or services
to your research, teaching, or
other activities:

a) Catalogue of Holdings of the
Machine Readable Archives;

-^~- Continued on page 22