College and Research Libraries


By A A R O N J . B R U M B A U G H 

Accrediting Associations and the 
College Library 
Dr. Brumbaugh is dean of the College of 

the University of Chicago. He is also 

secretary of the Commission on Institu-

tions of Higher Education of the North 

Central Association. 

TH A T A N E F F E C T I V E college must have a good library is generally accepted 
as axiomatic. So also is the assumption 
that the character and function of the 
library must be related to the nature of 
the college program. Any effort to deter-
mine the excellence of a college, what-
ever may be the purpose of the evaluation, 
must include, therefore, some measure of 
the quality of its library in relation to the 
college's purposes and program. 

T h e Commission on Institutions of 
Higher Education of the N o r t h Central 
Association has for many years published 
a list of accredited member institutions. 
T h i s list has been a means of informing 
institutions of higher education through-
out the country, and the public in gen-
eral, that the accredited colleges and uni-
versities provide an educational program 
of acceptable quality when judged by cri-
teria that have been adopted by the com-
mission. T h e library has always been 
included as one of the aspects of an insti-
tution to be considered for purposes of 
accreditation. 

Prior to 1934 many of the criteria 
employed by the commission in accrediting 

colleges were quite rigid and quantitative 
in nature. T h e criterion regarding the 
library, for example, provided t h a t : 

T h e college shall have a live, well-dis-
tributed, professionally administered library 
of at least 8000 volumes exclusive of public 
documents, bearing specifically upon the sub-
jects taught, and with a definite annual 
appropriation for the purchase of new books 
and current periodicals. It is urged that such 
appropriation be at least five dollars per stu-
dent registered. 

N e w criteria adopted in 1934 upon the 
recommendation of a committee that had 
devoted several years to a redefinition of 
the measures of institutional excellence 
under a subvention from the General Edu-
cation Board, stressed the importance of 
evaluating an institution in terms of its 
own objectives. T h e acceptance of this 
point of view led at once to a greater 
emphasis on the quality of the institution's 
program in terms of its purposes and called 
for a type of appraisal that would be more 
flexible than were the old criteria. Q u a n -
titative data were not abandoned; in fact 
more data of a quantitative nature were 
called for than previously, but these data 
were to be interpreted in relation to vari-
ous intangible and nonstatistical factors. 
T h i s emphasis upon the qualitative in con-
trast to the strictly quantitative character-
izes the plan of evaluation adopted in 1934 
and still in effect. 

T h e committee on evaluation referred 

206 COLLEGE AND RESEARCH LIBRARIES 



to above made a special study of the 
differentiating characteristics of fifty-seven 
institutions which in the combined judg-
ments of educators varied widely in the 
quality of their programs and which were 
representative of higher institutions in 
the N o r t h Central territory. T h e charac-
teristics thus identified were grouped under 
eleven general heads, one of which was the 
library. 

But the committee was still confronted 
with the question of what specific criteria 
might be employed in judging the excel-
lence of a college library. T h i s problem 
was attacked by first determining what 
criteria might be employed, then by finding 
the correlation between each criterion and 
other measures of institutional excellence. 
For example, the holdings of books and 
periodicals was chosen as one criterion. 
Six measures of holdings investigated were 
the number of volumes in the library; the 
number of periodicals subscribed t o ; the 
average annual additions to the library; 
the volumes added to the library in 1931-
32, the year immediately preceding the 
study; the number of «books held by the 
library that were included in a specially 
prepared checklist; and the number of 
magazines subscribed for that were in-
cluded in a checklist of periodicals. All 
of these measures gave significant statisti-
cal correlations with other measures of 
the excellence of the institution as a whole 
and with a composite rating of the excel-
lence of the library itself. Because the 
checklists of books and periodicals yielded 
a high correlation both with other meas-
ures of the library and with other criteria 
of institutional excellence, and because 
they provided an instrument that could be 
used most economically, they were adopted 
as the measures of library holdings. O t h e r 
measures arrived at by a similar tech-

nique were the average expenditure for 
books over a five-year period, and the an-
nual expenditure for library salaries 
weighted for enrolment. 

Libraries Compared 

For the purpose of securing comparable 
data, all member institutions were asked 
to apply the checklist on books and periodi-
cals to their respective libraries, to report 
their expenditures for books over a five-
year period, and their annual expenditures 
for library salaries. From these data was 
computed the percentile position of each 
institution on each of the four measures 
of library excellence. T h e distribution 
thus arrived at made it possible to com-
pare the library of an institution applying 
for- accreditation with the libraries of 
institutions that were already members of 
the association. Lest the inference be 
drawn that by this procedure the appraisal 
of a college library becomes merely a sta-
tistical procedure, it should be said that 
these percentiles have been and are re-
garded only as cues to the examiners in 
making an institutional survey for the 
purpose of determining eligibility to ac-
creditation. If the library stands high 
by the measures employed it is still looked 
into carefully by the examiners but not 
as critically as when it stands relatively 
low. 

In this procedure the checklist of books 
and periodicals is very important. F o r 
that reason something more should be said 
about the compilation of the checklists 
and about some of the problems that arise 
in using them. From several lists of titles 
considered suitable for college libraries 
that were available at the time our check-
list was constructed, the Shaw list of books 
for college libraries was chosen as the 
basis for the checklist. T h e use of any 

JUNE, 1941 207 



checklist involves the sampling process. 
Sampling is necessary because obviously 
it would be asking too much were librari-
ans requested to check a complete list 
such as Shaw's containing 14,000 titles. 
Moreover, were the full list of 14,000 
titles checked by the librarians of 280 or 
more institutions, the amount of statistical 
work required in the office of the secretary 
of the Commission on Institutions of 
Higher Education in deriving normative 
data would be prodigious. Furthermore, 
P r o f . William Randall found a high cor-
relation between the holdings in subdivi-
sions of long book lists and the holdings 
in the list as a whole. For example, using 
the entire Shaw list, Prof. Randall found 
that the number of titles that were 
held by 103 colleges when correlated 
with the holdings from the section 
on history in the same list, gave a co-
efficient of correlation of + . 9 4 . Similarly 
high coefficients were found for other 
sections. Reference books were found to 
be particularly useful in constructing a 
sampling list because they are basic to the 
whole college program and are less af-
fected by differences in curricular offerings 
than are departmental lists. A sampling 
list of 507 reference books was therefore 
constructed by the committee on evaluation 
and has been used in the same form since 
1 9 3 4 . 

Periodicals Checklist Made 

A similar instrument was needed to 
evaluate the periodical holdings of college 
libraries. D r . Eugene Hilton constructed 
a list of periodicals by pooling the votes of 
instructors as to the relative importance 
of various journals to the teaching of 
standard courses in each of twenty-six 
different college departments. T h i s list, 
like the Shaw book list, was too long to be 

used in its entirety; consequently only the 
upper 10 per cent of the titles were se-
lected from each of the twenty-six de-
partmental lists. T h i s list containing 304 
titles was reduced still fu rth er, giving a 
comparatively short checklist that was 
found valid for identifying weak periodical 
collections but not valid for differentiating 
collections of over a hundred titles. 

T h e validity of any sampling list of 
books or periodicals is impaired when that 
list is used by a library as a purchasing 
list. T h e numerous requests for copies 
of the library checklists and the unusually 
high rating of some libraries which were 
found by examiners to be far inferior to 
expectations based on their percentile 
standings on the four measures of excel-
lence, led to the conclusion in 1939 that 
a new checklist was needed. 

Among the criticisms of the original 
list that were taken into account in attack-
ing the problem of making a new checklist 
were ( a ) that the original list was not 
sufficiently flexible to take cognizance of 
differentiated types of institutions, e.g., 
the junior college,«the technical school, or 
the liberal arts college whose offerings 
were restricted to certain fields; ( b ) also 
that not enough account was taken of the 
library needs of institutions whose pro-
grams are built around certain basic re-
ligious ideas. 

T h e book lists selected for the purpose 
of deriving a new sampling list were 
Shaw's supplement to his List of Books 
for College Libraries, p u b l i s h e d in 1 9 3 1 ; 
t h e Selected Bibliography of Engineering 
Subjects, prepared by the Engineers' Coun-
cil for Professional Development in 1937; 
a n d t h e Catholic Books for College Li-
braries, provided by the Library Com-
mittee of the National Catholic Education 
Association. A preliminary list composed 

208 COLLEGE AND RESEARCH LIBRARIES 



of about one fourth of the 3800 titles in 
Shaw's supplementary list with some addi-
tions from the other lists was constructed. 
T h e books in this list were grouped by 
subjects, twenty-six in all, corresponding 
to the grouping in Shaw's original list. 

Holdings Computed 

About 260 institutions cooperated in 
trying out the new checklist and in criti-
cizing it after they had employed it. N e w 
percentiles of holdings on this list were 
computed, but the procedures employed 
differed from those used in computing in-
stitutional percentiles on the basis of the 
original checklist. Instead of finding the 
percentage of books held from the whole 
list, each institution's standing was com-
puted on the basis of its holdings in 
relation to the departments in which it 
gave instruction. T h e books in the fields 
in which the institution gave no instruc-
tion were therefore omitted before the 
percentage of holdings was computed. 
T h i s plan has the advantage of taking 
cognizance of the specialized programs of 
some institutions that do not and probably 
should not be expected to offer the full 
range of courses found in a majority of 
institutions in the association. I n other 
words, the computation was made on the 
basis of titles in the subject-matter fields 
in which work is actually offered. For 
example, one institution reported that 
astronomy and engineering and Catholic 
religion were not offered in its program. 
T h e number of titles on the booklist in 
these three fields is 135, so this number 
was subtracted from the total number of 
titles, 1032, to arrive at a base for com-
puting this institution's percentage of hold-
ings. T h i s institution consequently ranked 
higher than an institution holding the 
same number of volumes out of the com-

plete list but which offered a wider range 
of subject matter and therefore had a 
higher base for the computation of the 
percentage of holdings. In applying this 
technique the question arose whether an 
institution that offered no instruction in 
a particular field but reported that its li-
brary contained books in that field should 
be given credit for those holdings. For 
example, one institution referred to above 
offers no instruction in astronomy or en-
gineering, yet holds ten titles from the lists 
in these two fields. Should these ten titles 
be subtracted from its holdings in figuring 
its final percentage? O n the premise that 
a good college library should include some 
volumes outside the institution's immediate 
curriculum, it was decided that these vol-
umes should be included in its holdings. As 
a matter of fact, while this procedure 
seemed to make allowance for good general 
collections and saved considerable clerical 
work, it made little difference in the final 
picture. T h a t is to say, the data showed 
that a library with a large percentage of 
holdings in its fields of instruction is likely 
to hold more books outside of those fields 
than the library whose holdings in its fields 
of instruction are more limited. 

Variations in Ranking 

T h e results of this experiment in de-
vising a new sampling list have been grati-
fying in some respects, disappointing or 
at least perplexing in others. I t seems 
quite clear from the data that the method 
employed takes more adequate cognizance 
of the adequacy of the library in the in-
stitution offering a specialized or restricted 
program. O n the other hand, the corre-
lation between the percentile ranks of 
institutions based on holdings on the new 
checklist and the percentile ranks based 
on holdings on the earlier checklist is 

JUNE, 1941 209 



sufficiently low to be a matter of some 
concern. Some of the more marked vari-
ations are illustrated by citing the cases 
of a few institutions. 

1934-35 P e r -
tile R a n k centile 

I 35 
2 3 2 9 
3 4 3 8 
4 7 5 i 
5 74 
6 IS 6 6 
7 1 6 6 6 
8 19 60 
9 6 5 

1 0 58 
11 23 6 5 
12 2 8 6 9 
13 70 
14 74 
IS 35 78 
1 6 38 7 6 
17 42 74 
1 8 5i 70 
1 9 57 8 1 
20 59 8 0 
2 1 71 6 
2 2 74 43 
23 74 2 0 
2 4 8 2 19 
2 5 2 

Reasons for Variations 

Several explanations of these variations 
are suggested. First, the two checklists 
may measure different aspects of the li-
brary. T h e 1934 checklist probably f u r -
nished indices of the relative adequacy of 
the holdings of libraries without giving 
any special weight to recent additions. 
T h e supplement to the Shaw booklist in-
cludes no books published prior to 1930. 
It may be therefore that some libraries 
that ranked high in 1934-35 have not 
been adding new publications as assidu-
ously as have the libraries that ranked 
lower. O r perhaps the lists measure two 
different phases of a library, the former 
indicating holdings of standard reference 
works, the latter the acquisition of new 
books in various fields of instruction. O n 
the assumption that the latter may be the 
case, we are asking institutions applying 
for accreditation to apply both checklists 
to their libraries if they stand low in their 
holdings on the 1939 list. 

Validity of Lists 

Second, it is possible that the problem 
with which we are confronted grows out 
of a lack of validity of either or both 
checklists. W e have assumed that the 
original book lists from which our check-
lists were derived were valid sources. 
W e have no basis now for questioning 
this assumption. W e have endeavored to 
make our sampling list adequate, but must 
make f u r t h e r analyses before we can prove 
its adequacy. 

T h i r d , it is possible that some insti-
tutions used the 1931 Shaw list as a pur-
chasing list, undoubtedly a legitimate use 
to make of it, as a consequence giving these 
institutions a higher rating than others 
that did not follow the list in making 
their purchases. T h e supplement was not 
available as a guide in purchasing, perhaps 
leading to a shift in the positions of some 
of the institutions. W e are quite certain 
that some institutions accredited since 1934 
made a similar use of our own checklist, 
even though we have repeatedly advised 
against such use, thereby giving these 
institutions an undeservedly high rank. 

F u r t h e r exploratory studies will be 
made with a view to determining the 
validity of our new checklist and to iden-
tify the factors accounting for the change 
in the relative ranks of some institutions 
for which we see no good reason now. 

Changes in Periodical Holdings 

Fortunately, the same situation does not 
obtain for the periodical checklist. T h e 
data we have indicate that variations in 
the rank on holdings of periodicals reflect 
actual changes in the libraries rather than 
limitations in our measuring stick. It 
appears from an examination of the in-

(Continued on page 276) 

210 COLLEGE AND RESEARCH LIBRARIES 



U.S. Labor Department. Women's Bureau. 
T h e Woman W o r k e r . [Vol. I, No. i ] 
Feb. 10, 1921 to date. Washington, 1921 
to date. Bimonthly. 

H a s a section " R e c e n t P u b l i c a t i o n s " which lists the 
n e w p r i n t e d a n d processed material issued by the 
U . S . W o m e n ' s B u r e a u , as well as o t h e r f e d e r a l a n d 
state p u b l i c a t i o n s on t h e s u b j e c t of w o m e n in 
i n d u s t r y . Occasional a n n o t a t i o n s . 

U.S. Library of Congress. Documents Di-
vision. Monthly Check-list of State 
Publications. Vol. 1, No. 1, J a n . 1910 
to date. Washington, 1910 to date. 

T h e official a n d most complete checklist of s t a t e 
publications, both n e a r - p r i n t a n d p r i n t e d . Limited 
to material deposited in the L i b r a r y of C o n g r e s s . 
I n c l u d e s publications of the D i s t r i c t of Columbia, 
the t e r r i t o r i e s a n d possessions as well as those of 
associations of state officials a n d regional organiza-
tions. 

U.S. President. National Resources Plan-
ning Board. Bibliography of Reports by 
State and Regional Planning Organiza-

tions. No. 1, Jan./April 1937 to date. 
Washington, 1937 to date. Irregular. 
Processed. 
I n c l u d e s both processed a n d p r i n t e d p u b l i c a t i o n s of 

official p l a n n i n g bodies. Occasionally a c u m u l a t e d 
n u m b e r is issued. N o t a n n o t a t e d . 

U.S. State Department. T h e Department 
of State Bulletin. Vol. 1, No. 1, July 1, 
l939 to date. Washington, 1939 to date. 
Weekly. 
I n c l u d e s t h r e e lists of n e w f e d e r a l p u b l i c a t i o n s : 

" R e g u l a t i o n s , " " L e g i s l a t i o n , " a n d " P u b l i c a t i o n s . " 
L i m i t e d chiefly to p r i n t e d m a t e r i a l b u t n o t confined 
to t h e publications of t h e U . S . S t a t e D e p a r t m e n t . 
Not a n n o t a t e d . 

U.S. State Department. Publications of the 
Department of State. April 1, 1930 to 
date. Washington, 1930 to date. Q u a r -
terly. 

Confined to t h e p r i n t e d m a t e r i a l issued by the 
U . S . S t a t e D e p a r t m e n t , but c u m u l a t i v e with each 
issue f r o m O c t o b e r i , 1929. I n c l u d e s m a p s . N o t 
a n n o t a t e d . 

Accrediting Associations and the College Library 

(Continued from page 210) 

dividual library reports that more periodi-
cals are n o w received on the average by 
college libraries t h a n was the case five 
years ago. T o maintain the same relative 
standing on holdings of periodicals that it 
had five years ago, it was necessary there-
fore for any given institution to increase 
the number of periodicals received. O n e 
college, for instance, has substantially the 
same percentile rank n o w that it had in 
1934, but at present it receives 145 peri-
odicals t h a t appear on our checklist, as 
compared w i t h 93 on the 1934 list. I n 
1934 this library subscribed for a total 
of 199 periodicals; n o w it receives 260. 
W e have found a sufficiently close agree-
ment between the number of holdings on 
our periodical checklist and the total n u m -
ber of periodical subscriptions to be satis-
fied t h a t in this field our checklist is a 
valid measure. I t is of course much 
easier to validate a measure of periodical 
holdings than of book holdings because the 

total n u m b e r of periodicals available is 
comparatively small. 

I have made this somewhat lengthy and 
detailed statement for several reasons: 
First, to i n f o r m college librarians of the 
techniques we have employed in develop-
ing checklists of books and periodicals as a 
measuring device; second, to let the li-
brarians k n o w that w e recognize the 
soundness of some criticisms of the check-
lists t h a t they have m a d e ; third, to provide 
information upon the basis of which the 
criticisms of some faculty members and 
administrative officers may be a n s w e r e d ; 
f o u r t h , t o assure librarians and administra-
tive officers that we make no a r b i t r a r y or 
u n w a r r a n t e d use of the data which we 
secure through our checklists; f i f t h , to 
enlist the continuation of the cooperation 
which librarians have so generously given 
us in attacking the difficult problem of 
developing adequate measures of the rela-
tive excellence of college libraries. 

276 COLLEGE AND RESEARCH LIBRARIES