College and Research Libraries


Catalog Subject Searches in the 
Yale Medical Library 

BY BENEDICT BROOKS AND FREDERICK G. KILGOUR 

THE PURPOSE of the present study was to 
obtain specific data concerning the heav-
iest use of the subject cards in the catalog 
of the Yale medical library. These data 
will serve in the design of a catalog com-
puterization project being carried out at 
the medical libraries of Columbia, Har-
vard, and Yale universities. 

A search of the literature reveals some 
thirty surveys on the use of the card cat-
alog,l but most do not contain informa-
tion useful to this investigation. Frarey2 
provides an excellent review of the studies 
of card catalog use, one of his summary 
statements being that subject use of the 
catalog had been found to be about equal 
to author use. Jackson gives the results of 
the only attempt to obtain comprehensive 
information on a statistically meaningful 
scale. He shows "subject searches" as 43 
per cent and "known-item searches" as 
57 per cent for special libraries, the low-
est percentage of "subject searches" for 
any of the library categories listed. The 
mean was 52 per cent "subject searches" 
and 48 per cent "known-item searches."3 
Spalding, 4 analyzing call slips at the Li-
brary of Congress, concludes that 25 to 
50 per cent use is made of the subject 
approach. Emily Brown5 investigated the 

1 Sidney L. Jackson, Catalogue Use Study, edited 
by Vaclav Mostecky. (Chicago: ALA, Resources and 
Technical Services Division, Cataloging and Classifi-
cation Section, Policy and Research Committee, 1958), 
p.51. 

2 Carlyle J. Frarey, "Studies of the Use of the 
Subject Catalog; Summary and Evaluation," in 
Maurice F. Tauber (ed.) The Subject Analysis of 
Library Materials (New York: School of Library 
Service, Columbia University, 1953), pp.147-66. 

a Jackson, op. cit., p.10. 
4 C. Sumner Spalding, "The Use of Catalog Entries 

at the Library of Congress," Journal of Cataloging 
and Classification, VI (Fall 1950), 95-100. 

s Emily Klueter Brown, The Use of the Catalogue 
in a University Library, (Master's thesis, University 

NOVEMBER 1964 

Mr. Brooks is a Cataloger and Mr. Kil-
gour is Librarian, Yale Medical Library. 

use of the card catalog at the library of 
the University of Wisconsin. She reported 
that the subject division received 13 per 
cent of the use while the author-title di-
vision received 87 per cent and that staff 
use accounted for 28 per cent of the total. 
Although the majority of those using the 
subject catalog used only one subject 
heading, the largest number of subject 
headings searched was seven, while the 
average was 3.8. 

MATERIALS AND METHODS 

A plot made of recorded circulation 
for the preceding year in the Yale med-
ical library served as the basis for select-
ing a week of heavy use. Since this plot 
showed a fairly steady high of library use 
during October and November, the choice 
fell to Monday, October 28, through Fri-
day, November 1, 1963, as the period of 
observation. 

The card catalog of the Yale medical 
library consists of a single large case, con-
taining in its two sides cards in a dic-
tionary arrangement constituting a single 
alphabet. An observer on the opposite 
side of the room, facing the single free 
end, has no difficulty simultaneously not-
ing activity at both sides. A continuous 
record made by . an observer as catalog 
of Chicago, Faculty of the Graduate Library School) 
60l. Another pertinent study of a university library 
with generally similar results on card catalog use is 
Robert L. Bovey and Satinder Kumer Mullick, "A 
Study of Library Usage," in Johns Hopkins Univer-
sity Research Library, Progress Report on an Opera-
t-bons Research and Systems Engineering Study, 
(Baltimore: 1963) Section IV, pp ;53-108. 

483 



60 

~ Staff' Use 
I PublicUse 

1-4 
::s 
0 
~ 
1-4 
<!) 

0-t 
40 '"0 

E -~ 
(/) 

= 0 
() 

(/) 

~ 30 
!J 
bL) 
0 
"; ..... 
Cd 
() 

'+-1 
0 
1-4 20 <!) 

,.0 

s 
::s z 

10 

MTWThF MTWThF MTWThF MTWThF MTWThF MTWThF MTWThF MTWThF MTWThF MTWThF MTWThF MTWThF MTWThF MTWThF MTWThF MTWThF 
Hr. Ending 0900* 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 

M Monday, Oct. 28, 1963 W Wednesday, Oct. 30, 1963 F Friday, Nov. 1, 1963 
T Tuesday, Oct. 29, 1963 Th Thursday, Oct. 31, 1963 * Represents 0 hr. as library opens at 0830 

TRAFFIC AT THE CARD CATALOG OF THE YALE MEDICAL LIBRARY 



trays were consulted during all the hours 
of library opening (0830 to 2400 each 
day) of the observation period produced 
the information on traffic at the card cat-
alog. This record reflects the "transac-
tions" alone, disregarding information on 
users. Since the authors felt that a gross 
distinction could be made between the 
type of use made of the card catalog by 
the library staff of the technical service 
departments of cataloging and acquisitions 
(hereafter referred to as "staff") and that 
of the public plus other library personnel 
(hereafter referred to as "public"), the 
record makes this distinction. The only 
other distinction made in the record con-
cerns the number of trays consulted by a 
particular user before leaving the card 
catalog, designated here as a "session." 

Since the Yale medical library has a 
dictionary catalog, the observer could not 
record information concerning approach 
by subject, author, etc. Therefore, the 
authors designed an interview question-
naire to elicit these types of data. This 
form was purposely kept simple so that 
as much informality and "naturalness" 
could obtain in the interviews as possible. 
Interviews xarely required more than one 
or two minutes. Greater control in tabu-
lating the results was possible since but a 
single interviewer filled out all the ques-
tionnaire forms. The interviewer set up 
a schedule of two-hour periods, during 
which he intended to interview all users 
of the card catalog. The recording of total 
traffic coincided with these two-hour pe-
riods so that the two parts of the study 
might serve as a check on each other. 
Each day the interviewer was on hand for 
six hours, staggered over various times 
during the library day so that every hour 
from 0830 to 2400 appeared on the 
schedule at least once. A few persons did 
escape being interviewed during three of 
the busier periods, but their number was 
small. The authors feel that the data thus 
obtained are statistically reliable since 
they agree closely with unpublished re-

N 0 V EMBER '1 9 6 4 

suits obtained during random interview-
ing periods which spanned a month. 

RESULTS 

The accompanying figure depicts the 
traffic at the Yale medical library card 
catalog during the period of observation. 
First to be noted is that no clear pattern 
of card catalog use emerges, particularly 
in regard to staff. The least ambiguous 
result is that the hour 1400 through 1459 
(recorded as 1500) comprises the peak 
use of the card catalog by the public. 

Sessions at the card catalog consisted 
usually of one, two, or three drawers 
consulted, but the record was thirty (a 
member of the public checking forty-two 
separate periodical titles to see if the 
library possessed them) . Staff members 
not infrequently consulted twenty or more 
trays before quitting the card catalog. 

The table, analyzing the searches of 
public and staff in the card catalog by 
approach used, lists separately those 
searches in which subject approach was 
utilized and those in which subject ap-
proach was not utilized. It must be em-
phasized that nowhere do these figures 
reflect number of card catalog users (or 
interviews conducted) , but rather the unit 
is the "search," here construed as all 
steps involved in locating a known item, 
or references to material on a subject. 
Thus, for example, a person who wished 
material on the subject of tuberculosis as 
it affects the spleen and who looked un-
der "Tuberculosis" and "Spleen" was con-
sidered to have performed a single search. 

The table indicates that 501 searches 
occurred during the thirty hours of inter-
viewing and that these searches involved 
60 1 card catalog entries, making ap-
proaches 120.0 per cent of searches. 
These 50 1 searches are divided nearly 
equally between 251 for the public rep-
resenting 292 entries or 116.3 per cent 
of searches, and 250 for the staff repre-
senting 309 entries or 123.6 per cent of 
searches. Other figures to be commented 

485 



SEARCHES OF PUBLIC AND STAFF IN THE CARD CATALOG OF THE YALE 
MEDICAL LIBRARY BY APPROACH USED 

PUBLIC* STAFFt ALL UsERS 

A B c D E F G H J K 
-------- -- ------ ----

gj gj ~ gj gj ~ 
..c: ..c: + ..c: ..c: + 
~~ 

() - () ...... () ;... < ~ ~ ..c: ;... ril ft 0 ~ t 5 Q) Q) ..c: 
<ll 

0 --.: ~ ~ <ll Q) ..-< <ll <ll () <ll Q) ~ --...... . ..... ~ ...... ~ ~ ...... . ..... 0 ~ 0 
0 0 ;... o.g s X 0 0 "' o..C s ::: o;o ::: ::s ~ ~ 0. ::s ::s ~ ~ 0. ~ <ll 0 Q) ~ <ll 0 Q) "'0 ~ 

Q) 

~ ~ ~ ~ !::J) ..c: 
!::J) 

~ ~ ~ ii:: !::J>..C: 
!::J) X !::J) X 

0 ~ () 2 <'$ , ..... 0 ~ () 2 ~ E-t s Ill 
]:E~ 

"0 ..... <'$ 1:0 lQ "'d •1"""1 ..p.) "0 ..... Ill ~0 1~ ]~-+~ ,!d > 0 3 
011 C<l ,!d > () ,!d > 0 

3 
"0 ::s 

<'$ 0 ..... Ill-;... () oS 0 -~ <'$-;... () C<l ~- () lQ 
~ ~ .g ~ ~ g; ;... Q) > ,.Q ~ ~ g; "' <'$ 0 ;... APPROACH TO CARD CATALOG 0 Q) ;... ~ ::s 0 Q) ;...Q Q) P=l .... <ll P=l .e <'$ E-t P-<- P=l· ... <ll P=l .e <'$ E-t P..- 0~ P..----- -------- ----

Subject - 68 68 27.1 - 25 25 10.0 93 18.6 
Author 132 4 136 54.2 218 9 227 90.8 363 72.4 
Periodical title 51 2 0 51 20.3 5 1 6 2.4 57 . 11.4 
Other title . 26 2 28 11.2 3.8 3 41 16.4 69 13.8 
Editor . . . . . 5 0 5 2.0 1 1 2 0.8 7 1.4 
Corporate added entries 2 0 2 0.8 5 0 5 2.0 7 1.4 
Series 1 0 1 0.4 1 1 2 0.8 3 0.6 
Personal added entries 1 0 1 0.4 0 0 0 0.0 1 0.2 
Cross references+ - - - - 1 0 1 0.4 1 0.2 

-------- -------- ----
Total . 218 74 292 116.3 269 40 309 123.6 601 120.0 

------ -------- ----
Searches represented 2062 45 251 100.0 231 19 250 100.0 501 100.0 

* "Public" is taken here as including all card catalog users except members of the library staff in the cata-
log department and acquisitions department only, who constitute "staff" in this table. 

t This figure includes 42 titles, counted as 42 searches, because the individual was checking to see if the 
library possessed as many journals. 

:t: Included here as a special category for catalogers only where the card catalog was consulted specifically to 
check on a cross reference ; use of cross references by others was regarded as approach by author, etc., as 
appropriate and . is so listed. 

on are the percentages in columns D, H, 
and K. The public uses subject cards more 
than the staff-27 .1 per cent for the pub-
lic, 10.0 per cent for staff. Conversely, 
there is a higher use of the author ap-
proach by the staff-90.8 per cent as 
compared to 54.2 per cent for the public. 
On the basis of figures in the table the 
percentage of subject use of the card 
catalog, calculated on the basis of the 
number of searches is 12.8 over-all or 
17.9 if the public alone is considered. Re-
garding the subject searches, twenty-five 
of the public's forty-five, and five of the 
staff's nineteen, consisted of consultation 
of a single subject only; the largest num-
ber of subjects consulted in a single search 
was four, by a member of the public. 

DISCUSSION 

Although, as has been mentioned 
above, no clear pattern emerges from the 

figure regarding times when the card cat-
alog in the Yale medical library is most 
used (except for the 1400 through 1459 
maximum as far as the public is con-
cerned) , the pattern for the public, to-
gether with the less clear pattern for staff 
can provide figures to help predict what 
size of load a computerized catalog should 
be prepared to handle. The high point 
seen at 2200 W does not give a picture 
of a typical Wednesday at that hour, for 
it includes the single thirty-tray session, 
the extreme for the observation period as 
noted above, but nonetheless points up 
the kind of peak which must be taken in-
to account in deciding on capacity for 
computer operation. 

Results of the interviewing constitute 
the table broken down by approach em-
ployed in searches. There was no attempt 
to separate main entries from others, but 
on the whole the approaches called "au-

486 COLLEGE AND RESEARCH LIBRARIES 



thor" and "periodical titles" represent 
main entries, while other approaches are 
added entries. The authors' strong im-
pression is that users of the Yale medical 
library card catalog tend to seek their 
known item by means of the most direct 
or "obvious" approach. 

What has been said above for the pub-
lic applies to the staff as well, except that 
the staff makes particularly heavy use of 
the author approach, largely because of 
the extensive checking of items done by 
the acquisitions staff. The cataloging staff's 
use of the card catalog is predominantly 
to determine authority, for in the Yale 
medical library the public catalog is the 
authority for names and subject head-
ings. 

Additional information to come from 
the interviews must remain qualitative in 
nature because the limitedness of the sub-
ject use of the card catalog in this inves-
tigation does not permit statistically mean-
ingful results. Nevertheless, the following 
indications are clear. First, the public 
seeks material in English almost exclusive-
ly, and at any rate preferentially. Second, 
publication dates of material sought by 
the subject approach parallel closely the 
findings of Kilgour6 regarding recorded 
use of books in the Yale medical library, 
indicating the heavy preference for locat-
ing as recent material as possible. Third, 
a considerable number of the . public 
choose to locate a known item through 
the subject approach. Fourth, some indi-
viduals looked up a subject in the card 
catalog to learn the classification of the 
topic, or, in other words, as a quick guide 
to shelf location of material. And fifth, 
location of a particular form of material 
(e.g. periodicals of biochemistry, text-
books of embryology) constitutes a defin-
able group of subject searches. 

CONCLUSION 

The peak load of catalog use between 

6 Frederick G. Kilgour, "Recorded Use of Books in 
the Yale Medical Library,'' American Documentation, 
XII (October 1961), 266-69. 

NOVEMBER 1964 

1400 and 1500 hours can serve as a guide 
for predicting the amount of use that 
must be provided when a machine is avail-
able to do the searching. 

An author catalog is required for the 
operation of a library, for the staff must 
be able to determine the library's holding 
before deciding to purchase or add a 
book to the collection. On the other hand, 
the subject catalog is designed primarily 
for public use. However, this study shows 
that the public makes relatively little use 
of the subject catalog at a time when 
public demand for more and more infor-
mation is soaring. One conclusion to be 
drawn from these observations is that the 
traditional card catalog having an aver-
age of about 1.6 subject entries per 
book7 • 8 does not furnish references ade-
quate to meet new demands. 

SUMMARY 

An investigation of the use of subject 
cards in the catalog of the Yale medical 
library took place to aid in designing a 
catalog computerization project. Data 
gathered by continuous observation dur-
ing a week of heavy use indicate that, in 
general, the peak traffic load at the card 
catalog occurs between 1400 and 1500 
hours. On the basis of short interviews of 
card catalog users, 501 searches, which 
broke down almost equally between pub-
lic-251, and sta:ff-250, yielded sixty-
four or 12.8 per cent subject searches in 
all. If only the public is considered, there 
were forty-five or 17.9 per cent subject 
searches. 

In a significant number of cases the 
purpose of the subject search is to locate 
a known item, obtain a quick guide to 
the shelves through the classification num-
ber, or to find a particular form of ma-
terial (e.g., periodicals). • • 

7 Peter M. Sprenkle, and Frederick G. Kilgour, 
"A Quantitative Study of Characters on Biomedical 
Catalogue Cards-A Preliminary Investigation," 
American Documentation, XIV (July 1963), 202-206. 

8 Henry J. Dubester, "Studies Related to Catalog 
Problems,'' Th e Library Quarterly, XXXIV (January 
1964)' 97-105. 

487