lASSIST Newsletter* Vol. 3t No. 3 (Summer 1979)

A LIBRARY-BASED REFERENCE SERVICE FOR MACHINE-READABLE CENSUS DATA;

THE CANADIAN EXPERIENCE

SLavko Manojlovlch

University of Windsor

Windsort Ontario

The Cens
source for
econom i c an
social scie
small propo
census data
copy form,
growing use
analysis w

machi ne-rea
the census
more import
library is

information
pus with t h

avai lable i

patron's i

the I ibrar
paper will
place of t h

academ i c I

followed b

potent i al
which I i b.r

alike may
of the M-R

us of C

the dem
d hous
nee re

rt ion
is pu
If we
of CO

e can
dable (

is beco
ant to
the ma
on the

e libra
nte rmed
nf ormat
y • s r

at temp
e M-R
i br ary »

y a di

problem
ar i ans
encount
census .

anad
ogra
1 ng
sear
of t

blis
ad

mput
se

M-R)
mi ng
rese
Jor

un
r i an
i a ry

i ona
esou
t to
cens

T

scus
s

and
er i

a i

phi c

data
c h.

he
hed
d to
ers

inc
arch
depo
i ver
act
be

I n

r ces
ex

us w

his
s 1 on
and

re

s a prime
« social*
used in
Only a

avai lable
in hard-
this the
for data

that the
er s i on of

reasingly
er s . The
sitory of

s i t y c am-
ing as an
tween the
eeds and

This
amine the
ithin the
will be
of the

pitfalls
searchers
aking use

IH£ MzE CENSUS AND JHE ACADEMIC
LIMAEI

The Census of Canada first
became available in M-R form in

1961 with the production of the
User Summary Tapes (U.S.T.). In

1971 the Public Use Sample Tapes
(P.U.S.T.) were produced in res-
ponse to the researcher's need for
micro-data. For those Libraries
which had access to the M-R census
tapes the referencing of the

patron's request for census infor-
mation required a consideration* on
the part of the librarian* of both
the hard-copy and the M-R census
resources.

The importance of the latter is

evident in the following situa-
tions:

1. The requested data is not
abailable in a printed
publication but may be
retrieved from a census
tape. Since the census
bulletins provide summary
data for a Limited number
of standard geographic
regions* they are unable
to satisfy data requestes
for small municipalities*
subdivisions of these
municipalities* or user-
created geographic areas.
City planning districts
are an example of the
Latter. The major fea-
ture of the U.S.T. (enu-
meration area series) is

that the record is based
upon the enumeration area
(E.A.) . This is the
smallest geographic unit
of analysis allowed by
the principle of confi-
dentiality which governs
the dissemination of cen-
sus data. Furthermore*
since E. A. boundaries
never cross those of lar-
ger recognized statisti-



lASSIST Newsletter, Vol. 3, No. 3 (Summer 1979)

complete a description
as inaicated by the
Statistics Canada
informat ion.

c) Furthermore* these
respondents were
asked* "Who processes
the user's request for
information from
tape?". The majority
response was the com-
puter centre (33%),
followed by the data

arch
the
ef f o

and
tre
ing
a V

user
with
of f

cert
(i.e
soc i

ive (2

rt of
the c

(13).
33% re
ar i a t i

h i mse
out th
acuity
ai n

ge
I ogy )

0%)
co-o
the

ompu
The

spon
on
If,
e as

me
depa
ogra

t hrough
perat 1 ve
library

t er cen-
rema i n-

ded with
of the
with or

s i stance
tubers in
r

t

ment s .

phy or

TABLE 1

Do you have any mac h 1 ne- re adab I e census tapes on your campus?

SIZE OF HOLDINGS

MAJOR OTHER TOTAL

Response YES

NO

8 7 56% (15)

7 5 44% (12)

100% (27)

T

cate
of

act1
acce
whi c

f aci
the
exi s

or,
data
camp
Libr
Cana
tape
libr
eith
i nd i

hese o

that
ac adem
ve ly i

ss to
h are
Uty's
libra

tence
if it

from
us fa
ary .

da inf
s wer
ary .

er t h

V i dua L

bserva
the ov
ic I

nv I ve
the M

locat
tape

ry is

of th
is, th
tape

c i L i t y

A c c or
o rma t i

e rare
R equ

e c omp
f a c u L

1 1 on
er wh
i b ra
d in
-r c

ed i

lib
not

e M-
e pr
i s

ot
ding
on ,

ly
es t s

u t er
ty

s c Le

e I m i n

r i es
p rov

ensus
n th
ra r y .

a wa
R ce
ocess
per f o

he r

to
the

order
were
c en

membe

a rl y

g ma
are

idin
res

e CO

re
nsus
i ng
rmed
tha

Stat
M-R
ed b

m a

tre
r s

.

i nd i
-

j r i t y
not

g user
our ces
mput er
Either
f the
tapes

of the
by a

n the
i s t i c s

census
y the
de by
or by

Thi s

cou I

libr
the
it

mere
libr
t ow a

data
The
data
the
Rule
will
libr
M-R
of t

d p

ary'
M-R
is

ly
ary'
r ds
f il

ado i

f il

An

s (

i ngn
ary
r eso
he I

ossibly
s posit
census,
more I i

a re

s tradi
M-R in

es and
t i on o

e s in t

g lo-Ame
AACRII )

ess on
to acce
ur ces a

i b rary '

ac
ion

On
kely
flee
t i on
form
c omp
f a

he I

r i c a

s

th

Pt
s a

s ho

coun
with
the
th

t i on
al
at i

uter
cha

a t es
n

hou I

e p

the
leg

Idin

t

re
oth

at

indi
n i

pro
p t e r

t ed
Cat

d f

art
un 1 v

i t i m

gs.

for the
spec t to
er hand,
this is

of the
f f erence
nc ludi ng
grammes.
on M-R

i t ion of

a logui ng
oster a

of the
ersity's
ate part

- 57



lASSIST Newsletter, Vol. 3, No. 3 (Summer 1979)

expand their services
primary data reference w

reasonable in light of
to acquire adequate supp
ment at i on.

file
the
of

In
vide
he d

the
1 1 em

i

uld
thei
ort

s tea
data

secon-
either
d with
e s i red
i nf or-
pt to
nc lude
not be
r need
docu-

The library has the personnel
and the resources to provide a pri-
mary data reference service for the
M-R census. The success of this
venture depends on the following
condi t i ons :

1. The university community
must recognize thelibrary
as the resource centre
for all census informa-
tion* M-R and hard-copy.
Thust the library should
be prepared to promote
its expanded service by
conducting several infor-
mation and training semi-
nars for faculty* staff
and students on the
availability and use of
the M-R census

.

2. The li

take a

M-R c

rent ly
pus a

na t e

acquis
Stat is

ma t i on
of the
major
acqui r

of the
sar y a

cons i d

are s t

t e r

L i b r a r

b ra r

n in
ensu
ava

nd s

i t i o

tics
ind

i n

h

ed m

tap
dded
er
or ed

c e

y

.

y will
V en t or
s t a

i lable
hou Id

the
n of
Canad

i c ated
st itut

I di ng
ultipl
es * an
e xpen

that t

in t

n t r e s

At

nee
y of
pes

on
co-o

fu

ta
a in
that

ions
s

e CO
unne

se if

he t

he CO

d to
the

cu r-
c a m-
r d i

-

t u re
pes .

f r-
2 5%

with
had

pies
c es-
y ou

apes
mp u-
t ape
h e se

institutions tape
requests were made by as
many as 8 departments.

The
libr
copy
wh ic

ac ce
c omp
p r oc
r equ
some
e lem
An i

a p

such
t han
expe
mass
A mi

of a

such
help

ma j

ar ies
com

h th
s th
uter

.

essi n

ests
st a

enta r

nt rod
rogra

as

sat i

rt i se
aging
ni -se

st a

as S

ful.

or i t y
have a

put er
ey can
e un i V

Howev
g of

will
f f trai
y prog
u c t or y c

mm i ng
PL/1 wi
sf y the

necess
data f

r

ss i on on
tistical
AS would

of the
hard-

terminal
use to

ersity's
er , the

user
require

ni ng in
ramm i ng .

ourse in
language
II more
computer
ary for
om t ape .

the use
package
also be

A

vice
tine
user
is d

copy
Sec
a re
tied
t i on
wi I I

may
use

I i b rar
for th

t advan
. Firs
i rec t ed

and M

nd, the
f erence
in the

The
focus u

arise i

of the M

y-ba
e M-
t age
t, t

to a

-R
requ
libr
art

rema
pon
n t

-R c

sad
R c

s f

he p

n i

c ens
est
aria
of

inde
the

he r

ensu

refer
ensus
or t

at ron
nt egr
us c

is pr
n who
quer

r of
prob

e f ere

enc e

has
he
• s r

ated
oLle
oces
is

y ne
this
Letns

nc i n

ser-
dis-

census
equest
har d-

ct i on.
sed by
qua I i-

go t i a-
paper
which

g and

"solutions
PROBLEMS AND

Thelogical analysis of a typical
census query into its component
parts can provide us with a frame-
work for comprehending the underly-
ing problems associated with refer-
encing census summary data.
Statistics Canada has developed the
PQRST classification as the basis
of an indexing and retrieval system



lASSIST Newsletter* Vol. 3» No. 3 (Summer 1979)

—' —
• o

— — o

CO CO
CO CO Ol ^j
en en o en

CO CO
en *. UD en
en o O en

CJ COo o CO *nJo en O o

ro ro
-c» CO CO
en o o en

TO
n>

Ol -b -^ U3^ 4- en o en O
cu
a-

—t

r+ ,-i

^ o en eno o o O
D.

n
ro
a>

C»J ro
(T> o en o o
(o

o en o o

en en en o

o o o o

- 61



lASSIST Newsletteri Vol. 3, No. (Summer 1979)

major realignment o't

boundaries. Witness
the recent emergence
of the "regional muni-
cipality" which
replaced certain coun-
ties in the province
of Ontario.

Code - A geographic
area is identified in
the M-R census via an
assigned code. The
introduction of the
"Standard Geographical
Classification" in
1975 produced a com-
plete revision of
these codes in the
1976 census. L 16 ]

T

of
chan
the
grap
I ys i

the
cont
numb
bod i

E.A.
boun
mine
C ana
c i pa
with
trol
no
w h i c

geog
the
det a

to c

pr i a

the
m en t

a gen

he
the

ge s

fact
hie
s m

ro I

er
es .

an
da r i

a b

da

»

I b

in p

sin
h de
r aph
requ
i I

onsu
t e p

re sp

ce
c i es

i den
se
i s h

th
I eve
ay I

j ur i

of
of

Fo
c e

e s a

y

whe
ound
ro V i

Sine
g I e

s c r i

i c c

i red
one
It

ubli
ec t i

part

t if i

pot
ampe
at a

I o

i e

sd i c

one
gove
r ex
n su s

r e

Stat
rea s

a r i e

n c i a

e th
do

bes
hang

le

i s

the
c at i

ve g

men t

cation
en t i a I

red by
geo-

f ana-
wit h i n

t i on a L

of a

rnm ent
amp I e t

tract
deter-
i s t i c s

m u n i
-

s are
I c on-
er e is

cum ent
these

e s in
ve I of
forced
app r o-

ons of

ove r n-

s and

d)

A.'s.C
Theor e

wou Id

create
area
years .

corres
for en
reveal
of the
f i cat i

mit th
of equ
all c

t i s t i c

p ro V i a

I at i on
c i f i ed
Census
Fi le

become

17]
t ical
enab I

ove
Unf

ponde
umera
that
bou

ons
e i de
i va le

a s es .

s Ca
e spe
s for
area

Ma
s ho

ne c e

Ly»
e a

com
r

ort u

nee
t ion
the

ndar
o n

nt if

n t u

Lie:
nada
e i a I

us
s f

s t er
uld
ssa r

this
user to
parable
census

nat e ly

»

lists
areas

nature
y modi -

ot per-
i cat i on
nits in

Sta-
doe s

t abu-
e r- spe-
r om the

Data
this

y«

Time: A

is conduct
years with
sus occur
5-year int
next major
be in 196
sufficient
that the
contains o

demographi
and some
housing o

questions.

major c

ea eve
a mini
ring at
e r va I .

census
1. I

to
m i n i - c

n ly the
c oues

addit
r eco

ensus
ry 10

-cen-
the
The

will
t is
not e

ensus
core

t i ons
i ona L

nomi c

CONCLUSION

The -"Official
Lists" of each census
identify geographic
areas in terms of E.

The first sec tion of this paper
described the current si tua t i on
sur roundi ng the accessibi I ity of
the M-R census at Canadian un i ve r-
sities and made a case f or the
creation of a centralized census
cat a reference service with in the
library. The discussion of the
problems related to census data use
unoer scored the need for a Library
or information professional to
assist the researcher in the ac qu i

-



lASSIST Newsletter, Vol. 3. No. 3 (Summer 19 79)

ROUND II DELPHI QUESTIONNAIRE INSTRUCTIONS

PLEASE READ CAREFULLY

In the last issue of the Ne w s,j^et.t.ejr you were asked to respond to Round
I of a Delphi study designed to aia our thinking about future
developments relating to the distribution and archiving of machine
readable data. That questionnaire was completely open-ended and was
structured to elicit the ideas and concerns (in a number of categories)
of the members of lASSIST. The responses from Round I were usee in the
development of the Questionnaire presented in this issue -- a closed
questionnaire consisting of forty-five "event" statements.

For this Round of the Delphi we are asking that you respond to each
event statement with three different ratings or "Questions" (designated
Question 1, Question 2t and Question 3). Questions 1 and 3 are seven
point rating scales extending from "Low" to "High". Question 2 requests
that you make an estimate of the Year in which the event will take place
or begin. With these three Questions we are attempting to do the
followina:

1. Assess the importance of each event
(Question 1).

to the members of lASSIST

2. Forecast the approximate date by which the event will occur or
begin to change (Question 2).

3. Estimate the Likelihood that the event will actually occur
(Questions).

When the data are evaluated, median ratings on each of the three
questions will be used as an estimate of the collective viewpoint of the
lASSIST membership.

When using the seven point ratings scales
scales should be interpreted as follows:

for Questions 1 and 3 the

1. Very low importance or Likelihood.
2. Moderately Low importance or likelihood.
3. Low importance or likelihood.
4. Neutral importance or Likelihood (a 50/50 chance).
5. High importance or Likelihood.
6. Mooerately high importance or likelihood.
7. Very high importance or likelihood.

To use the scales, simply CIRCLE the number which best approximates your
viewpoint concerning the event statement.

Our primary concern is in forecasting directions in the nineteen
eighties so your estimates of aates shouLo generally be in the time
period 1980 - 1990. It may be, however, that for a particular event you
believe that the oate will be Later than 1990 -- if so, indicate the
date. If you think the event will never happen, enter "5999" as the



PAGE t

QUESTION 1 QUESTION 2

HOW IMPORTANT IS ABOUT WHAT YEAR WILL
THIS EVENT TO THE THIS EVENT TAKE
DELIVERY OF SERVICE? PLACE OR BEGIN?

QUESTION 3

WHAT IS THE LIKELI-
HOOD THIS EVENT WILL
TAKE PLACE?

KEY
KEY

1 LOI=LOW IMPORTANCE
3 LOL=LOW LIKELIHOOD

HII=HIGH IMPORTANCE
HIL=HIGH LIKELIHOOD

1. A significant increase in the use of oroprietary restrictions
on the dissemination of data.

LOI HII

12 3-4567 YOUR ESTIMATE
( )

LOL HIL

12 3 4 5 6 7

2. Expanded use of restrictions (copyrights* restrictive
contracts) on the use of software usea for archival and
analytical purposes.

LOI HII LOL HIL
YOUR ESTIMATE1234567 ( ) 1234567

3. An expansion of the use of "non-standard" formats for
disseminating data.

LOI HII

12 3 4 5 6 7

LOL
YOUR ESTIMATE

( )

HIL

12 3 4 5 6 7

4. A decline in the ability of archivists to promote the use of

standardized systems for defining data elements (variables).

LOI HII

12 3 4 5 6 7

YOUR ESTIMATE
( )

LOL HIL

12 3 4 5 6 7

5. A decline in the ability of archivists to promote the use of

standardized systems for classifying and retrieving machine
readab le data.

LOI HII LOL HIL
YOUR ESTIMATE1234567 ( ) 1234567

6. A precipitous- increase in the amount of machine readable
data.

LOI HII

12 3 4 5 6 7

YOUR ESTIMATE
( )

LOL HIL

12 3 4 5 6 7



PAGE 6

QUESTION 1 QUESTION 2 QUESTION 3

HOW IMPORTANT IS ABOUT WHAT YEAR WILL WHAT IS THE LIKELI-
THIS EVENT TO THE THIS EVENT TAKE HOOD THIS EVENT WILL
DELIVERY OF SERVICE? PLACE OR BEGIN? TAKE PLACE?

KEY 1 LOI^LOW IMPORTANCE
KEY 3 LOL=LOW LIKELIHOOD

HII=HIGH IMPORTANCE
HIL=HIGH LIKELIHOOD

13. Expansion of storage space requirements for data files.

LOI HI!

1 2 3.- 't 5 6 7

LOL HIL
YOUR ESTIMATE

( ) 12 3 4 5 6 7

14. Standardization of criteria for evaluating the importance of
data files for inclusion in an archive.

LOI HII

12 3 4 5 6 7

LOL HIL
YOUR ESTIMATE

( ) 12 3 4 5 6 7

15. The rising cost of traditional methods for disseminating
information (paper» printing) alter methods and techniques of

dissemination.
LOI HII LOL HIL

YOUR ESTIMATE1234567 < ) 1234567
16. Expansion of the variety of users requiring machine readable

data files.

LOI HII

12 3 4 5 6 7

YOUR ESTIMATE
( )

LOL HIL

12 3 4 5 6 7

17. Expansion of commercial archival services for machine
readable data.

LOI HII

12 3 4 5 6 7

YOUR ESTIMATE
( )

LOL HIL

12 3 4 5 6 7

18. Development of more varied funding techniques for the
development of machine readable data archives.

LOI HII

12 3 4 5 6 7

YOUR ESTIMATE
( )

LOL HIL

12 3 4 5 6 7



PAGE 8

QUESTION 1

HOW IMPORTANT IS
THIS EVENT TO THE
DELIVERY OF SERVICE?

QUESTION 2

ABOUT WHAT- YEAR WILL
THIS EVENT TAKE
PLACE OR BEGIN?

QUESTION 3

WHAT IS THE LIKELI-
HOOD THIS EVENT WILL
TAKE PLACE?

KEY
KEY

1 LOI=LOW IMPORTANCE
3 LOL^LOW LIKELIHOOD

HIUHIGH IMPORTANCE
HIL=hIGH LIKELIHOOD

25. Increased need for access to machine reaaabLe aata files
through technical systems which are transparent to the end
user.

LOI HII LOL HIL
YOUR ESTIMATE12i't567 ( ) 1234567

26. Expanded training of potential end users concerning the
availability and use of machine readaole data files.

LOI HII

12 3 4 5 6 7

YOUR ESTIMATE
( )

HIL

12 3 4 5 6 7

27. Dissemination of data files through networks rather than by
tapes or other magnetic media.

LOI HII

12 3 4 5 6 7

LOL
YOUR ESTIMATE

( )

HIL

12 3 4 5 6 7

28. Establishment of centralized and hierarchical data networks
to improve access and to reduce cost of access to machine
readaole data.

LOI - HII LOL HIL
YOUR ESTIMATE1234567 ( ) 1234567

29. Expanded use of microfilm or microfiche techniques in the
dissemination of documentation and other relevant
informat ion.

LOI HII LOL HIL
YOUR ESTIMATE1234567 ( ) 1234567

30. Use of cassette tapes or diskettes used in conjunction with
microcomputers (or with larger mainframes) for the
a i s sem i na t i on of information.

LOI HII LOL HIL
YCUR ESTIMATE1234567 ( ) 1234567



PAGE 10

QUESTION 1 QUESTION 2 QUESTION 3

HOU IMPORTANT IS ABOUT WHAT YEAR WILL WHAT IS THE LIKELI-
THIS EVENT TO THE THIS EVENT TAKE HOOD THIS EVENT WILL
DELIVERY OF SERVICE? PLACE OR BEGIN? TAKE PLACE?

KEY
KEY

1 LOI=LOU IMPORTANCE
3 LOL=LOW LIKELIHOOD

HII=HIGH IMPORTANCE
HIL=HIGH LIKELIHOOD

37. Greater concern for problems oi human engineering in the
production of equipment designee! for data retrieval.

LOI HII

1 2 3 't 5 6 7

YOUR ESTIMATE
< )

LOL HIL

12 3 4 5 6 7

38. Development of software and hardware for dealing more
automatically with natural language materials.

LOI HII

12 3 4 5 6 7

LOL HIL
YOUR ESTIMATE

( ) 12 3 4 5 6 7

39. Expanded deployment of f ont -i ndependent readers for
converting printed materials into machine readable form.

LOI HII

12 3 4 5 6 7

LOL
YOUR ESTIMATE

( )

HIL

12 3 4 5 6 7

40. Expanded concern on the part of machine readable data
archivists with commercial and governmental data needs.

LOI HII

12 3 4 5 6 7

LOL
YOUR ESTIMATE

( )

HIL

12 3 4 5 6 7

41. Expansion of the scope of lASSIST to meet the demands of

users other than researchers.

LOI HII

12 3 4 5 6 7

LOL
YOUR ESTIMATE

( )

HIL

12 3 4 5 6 7

42. Substantial resources devoted to the development of less
expensive methods of information delivery.

LOI HII

12 3 4 5 6 7

LOL
YOUR ESTIMATE

( )

HIL

12 3 4 5 6 7



PAGE 12

51« In what region do you reside'

C D 1» Western Europe
C D 2. Eastern Europe
C ] 3. Canada
C ] 4. The United States
[ ] 5. Other (Where)

52. Institutional Affiliation;

C 3 1. Academic Institution
C 3 2. Commercial Organization
C 3 J. Governmental Agency
C ] 4

.

Independent data archive
L 1 5. Other (What)

53. Regardless of the type of agency for which you work» are you
employed within a department which has:

C 1 1. Primary objectives other than data archiving?
C ] 2. Does primarily data archiving and servicing?

54*55. How many employees devote substantial amounts of time to
data archiving? C j

Comment s



lASSIST Newsletter, Vol. 3, No. 3 (Summer 1979)

Official 117. Statistics Canaoa
List, 1971.

Statistics Canada. Corres-

eoQaina kllkzllkk EQumsnaiifin
Area Numbers,

65