ACRL News Issue (B) of College & Research Libraries


January 1986 / 13

RESEARCH 
FORUM

Consumer research for college libraries: 
Doing the job in-house

By Jam es V. D i G iam battista
College Librarian and D irector o f Learning Resources 
Hawaii Loa C ollege

The Learning Resources C enter at H aw aii Loa
College entered its fifth year of operation in Au­
gust 1985. In addition to a new facility the C en­
ter’s academic library incorporated a num ber of
technological innovations th a t were new to the
College. These included DIA LO G and ORBIT
search capabilities, a com puter-based circulation,
catalog, and acquisitions system, and increased
purchasing of m icroform s. T he L e a rn in g  Re­
sources Committee of the faculty and the D irector
of Learning Resources concluded th a t this set of in­
novations, and the services based on them , were
now familiar to the College and should be given a
formative evaluation.

The Committee and the D irector found th em ­
selves involved in a program  of consumer research
designed to examine the library needs of various
user groups, their preferences for service, inform a­
tion needs, and adequacy of the current program .
It began the research w ith a survey directed at the
needs of College faculty and the curriculum .

For the purposes of research, the responses of all
faculty members were given equal weight. The
Committee did not distinguish between part-tim e
and regularly appointed instructors. T he over­
whelming m ajority of teachers at H aw aii Loa
(77%) hold regular, full-tim e appointm ents. The
decision making processes am ong a small faculty
of fifty-one are dem ocratic enough th a t concerns
voiced by any fa c u lty  m em b er are considered
widely.

As a small library, just over 50,000 volumes, the
premier aim of the unit is curriculum  support.

 

 
 
 
 
 
 
 
 

 
 
 
 
 
 
 
 
 
 
 
 
 

Staff members work to meet the constant dem and 
for resources dealing w ith current course topics. 
Fifty thousand relatively recent acquisitions go far 
in this cause. The areas of user needs included in 
the survey were reference services, both com puter- 
based and traditional desk services; reserve and 
circulation services; and collection devlopm ent 
and evaluation. Respondents were asked to evalu­
ate the book and periodical collections in their 
area of specialization and to anticipate any need 
for changes in subject area direction and coverage.

Bibliographic and library skills instruction were 
examined from four perspectives: the form at of in­
struction, responsibility for assuring an opp ortu ­
nity for students to acquire skills, planning link­
ages between BI and the subject area curriculum  
and evaluating the use of library skills across the 
curriculum . The Com m ittee decided th a t it was 
valuable for the library to have a profile of the be­
liefs of the instructional faculty about how biblio­
graphic instruction was best accomplished. V aria­
bles considered in developing this sequence of 
survey questions included w hether the BI should 
be delivered in individualized, small group, or 
class size form at? Should the job be done only by a 
librarian, in cooperation w ith the instructor, or 
did th e instructor prefer to do th e w hole job? 
Should the BI program  be planned as p a rt of a 
course, exist outside the course curriculum , or be 
in te g ra te d  in to  th e  course assig n m en ts c o m ­
pletely? Finally, our survey allowed us to assess the 
use of library research projects assigned by various 
instructors and gain a sam pling of w hat they felt



14 /  C &R L  News

were the most successful projects.
Since the installation of a computer-based ac­

quisitions, circulation, and catalog system in 1981, 
many features of technical support services had 
changed. Although the catalog is not online, it was 
im portant to gather some idea of the reaction to 
this possibility. Ordering and arrival notification 
processes had undergone the most change. Notifi­
cation was faster, but less personal. The staff felt 
the need to evaluate the effects of library technol­
ogy on our users.

Finally, the consumer survey process was the 
only im partial way to sample priorities for ser­
vices. Faculty respondents were asked to prioritize 
their need for reference and instructional services, 
collection development needs in the short and long 
term , and to rank the im portance of major catego­
ries of expenditure. This included reference and 
research services, collection developm ent, and 
technical services.

Directing and carrying out 
a library consumer survey

Survey developm ent and adm inistration is a 
planned process th at requires the participation of 
the community of respondents and a degree of pop­
ular support. In cooperation w ith a sponsoring 
committee the library must establish the need for 
information and the appropriate means of obtain­
ing it, w hether a mail-out or interview type survey. 
The committee membership should reach well be­
yond the library staff into the library’s constituen­
cies if it hopes to examine user needs thoroughly 
and impartially. The committee specifies the goals 
of the research and develops a list of the categories 
of information needed. The committee then ap­
points no more than two members to work w ith the 
director in drafting the instrument.

Once the draft survey instrum ent is complete, it 
is tested to see if it elicits the kind of inform ation 
sought. Responses from a small group representa­
tive of the sample is adequate basis for making any 
revisions in the instrum ent. The survey adm inistra­
tor, probably a member of the library faculty, then 
pulls a controlled sample of the population to be 
surveyed. This may include all undergraduate stu­
dents, all faculty members or any sub-group. It is 
im portant th a t the sample, or list of persons to be 
queried, represents the whole population and is not 
skewed tow ard gaining more responses from any 
one group.

Consumer surveys can be mailed to respondents, 
or respondents can be interviewed in person or by 
telephone. In a small institution w here the library 
has consistent and frequent contact w ith respon­
dents, a mail-out survey can generate significant 
response. This is particularly true if the survey ad­
ministrator is able to follow-up each non-response 
with a call or personal note.

On larger campuses and w ith surveys of the stu­
dent body, a mail-out questionnaire is likely to gen­

erate responses only from interested library users. 
T he opinions of non-users or in freq u en t users 
would go unnoticed. This is the best justification 
for conducting interview-type surveys either by 
telephone or in person.

Telephone surveys are a useful technique for re­
spondents w ith stable addresses and telephone 
numbers. A survey of faculty could be conducted 
easily by telephone, while doing the same w ith a 
student population, particularly off-campus stu­
dents, is problematic. Random sampling of stu­
dents in personal interviews is more productive. In 
either case, interviewers must be trained to apply 
the survey instrum ent and to produce consistent 
and complete surveys during the interview. Their 
ability to record accurately responses to open- 
ended items w ith speed is particularly im portant.

In most commercial survey research, telephone 
or personal interviews yield results th a t are prefer­
able to a mail-out type survey. Mail-out surveys to 
most respondents are kept simple, asking only di­
rect, multiple-choice questions. In general, they do 
not yield detailed results. From our experience, a 
detailed mail-out survey among college faculty is a 
reliable instrument. The level of education and 
sense of responsibility among faculty may be the 
distinguishing factor.

Compiling survey results

Once data collection is finished, only those mail- 
out surveys th a t have complete or reasonably com­
plete response are counted. Only those interview- 
type surveys which are fully complete are utilized, 
because the only reason for incompleteness is inter­
viewer error. Samples of over one hundred com­
pleted surveys should be keypunched and the com­
puter used to produce tables of item analysis and 
cross-tabulation. For responses of less than 100 usa­
ble surveys, it is practical to hand-tabulate results.

Responses to open-ended questions must be en­
coded to yield uniform  statistical results. Each 
open-ended response is placed in one category 
among five to ten possible choices and keypunching 
and item analysis proceed from the num ber of re­
sponses th a t fit into any one category. Finally, after 
analyses of response to each item in the question­
naire, tables are developed com paring responses to 
related items. Cross tabulations are developed to 
com pare the answers of various segments of re­
spondents to the same questions to examine the 
consistency of response and explain deviation.

In our survey, a significant num ber of faculty 
m em bers felt th a t students lacked a sufficient 
knowledge of library skills to complete their assign­
ments, yet faculty showed little interest in team- 
teaching library skills w ith a librarian or planning 
course assignments in conjunction w ith a member 
of the library faculty. This cross-tabulation dem ­
onstrates a perceived need for library skills among 
students and a failure of the course planning m ech­
anism to meet th a t need.





16 /  C &R L  N ew s

Report writing
R eporting research findings in a w ay th a t vali­

dates the techniques used to gather inform ation 
an d  makes those findings u n d e rsta n d a b le  is th e 
h e a rt of survey research. Any w ritten  rep o rt of the 
project should state clearly the goals of the research 
and w hy the survey was initiated. It should de­
sc rib e  th e  sa m p le  p o p u la tio n  a n d  re v ie w  th e  
m eth o d  used to select th e sam ple. Any u nusual 
problem s the researcher encountered in th e course 
of the survey th a t m ight affect the direction of the 
results is discussed openly. T he actual statistical 
findings are reviewed in detail. Responses to im ­
p o rta n t survey items are displayed in tables in the 
text of the rep o rt along w ith  cross-tabular d a ta  th a t 
fu rth er interprets the response to individual survey 
items. T he significance of survey results is discussed 
in term s of the survey objectives. As the statistics 
are revealed, the results build th eir case.

Executive sum m ary is perhaps th e most im por­
ta n t and neglected area of report w riting. Execu­
tive sum m ary distills th e significance of th e re ­
search for decision makers and interested persons 
beyond the research com m ittee. If the entire report 
is the record of the survey, th e executive sum m ary 
discusses the significance of th e findings for deci­
sions th a t m ay follow. It is not simply a set of rec­
om m endations, b u t a succinct tw o to four page re­
view of the results and an in terp retatio n  of their 
m eaning. E xecutive su m m ary  is m ost useful in 
keeping decision makers and those people affected 
by the results aw are of verifiable trends in user 
needs and preference. W ith in  a college or univer­
sity adm inistration, it is often th e basis for recom ­
m ending or supporting action and change.

Focus groups

S tru c tu re d  in terv iew s a n d  discussions, o ften  
called “focus groups,” can be useful in developing a 
survey instrum ent or in refining th e results of a p re­
viously adm inistered survey. O ne problem  w ith 
conducting consum er research in-house is th a t the 
persons designing the instrum ent ten d  to seek infor­
m ation in the channels w here they expect to find it. 
T here is no outside consultant to represent th e fresh 
and sometimes u n tu to red  perspective of th e library 
user. Focus groups can serve this purpose.

P articipants selected for a focus group discussion 
should be representative of th e sam ple and w illing 
to talk about th eir needs and perceptions of library 
services. The setting and atm osphere of the discus­
sion should be relaxed to encourage discussion. 
Most focus group sponsors serve food and beverages 
as refreshm ent and a stim ulant for conversation.

Discussion is conducted according to a set of p re ­
determ ined questions and the leader is responsible 
for guiding the progression of talk to m eet th e needs 
of the researchers. P articipants m ust be allowed 
the freedom to express opinions fully, w hile d ia ­
logue am ong group m em bers is kept to a m inim um  
unless it is useful in fu rth erin g  th e discussion. An

observer should be present for reliability, and to 
develop some record of response im p o rtan t to sur­
vey developm ent or in terp retatio n  of results.

Focus group results can and should be included 
in the research rep o rt and th e results detailed as a 
fu rth er refinem ent in m ethod.

In-house research vs. 
contracted services

As m any com m ercial and industrial companies 
are now  discovering, it is possible and feasible to 
conduct m eaningful consum er research as an in- 
house effort. Given th e often peculiar functioning 
of academ ic institutions, and the very specialized 
concerns of academ ic libraries, research there may 
best be conducted in-house.

T h e p ra c tic a l c o n sid eratio n s of in-house re ­
search lend themselves well to college and univer­
sity libraries. The research process, p articu larly  a 
m ail-out survey, occurs over a d u ratio n  of tim e. It 
requires p ro m p t and m ethodic follow -up, b u t will 
n o t m o n o p o lize an y  m e m b e r of th e  sta ff com ­
pletely until the statistics are developed and th e re­
p o rt w ritten . T he publics of most academ ic institu­
tions are easily identified and tracked. Pulling a 
sam ple should not be m ore difficult th a n  a trip  to 
th e  re g is tra r’s office for a c o m p u te r p rin to u t. 
W ithin most institutions, certainly small colleges, 
there is the sense of a need to respond w hen a ques­
tion is asked in fairness. Finally, research results 
are given greater w eight w hen they are developed 
by academics for use w ith in  the academ y. C on­
sum er research conducted in-house means th a t in­
dividual student or faculty m em ber’s responses re­
ally co u n t an d  are un d ersto o d  in th e ir context 
because the researchers are not outsiders.

Reasons for research
At H aw aii Loa College research was initiated to 

assess the full program  of inform ation services, in­
cluding library, audiovisual, and academ ic com­
p u tin g  services. Five years ago th e College built a 
new  facility th a t physically integrated all inform a­
tion service units and placed them  u n d er one ad ­
m inistration. T he C enter was to develop a com­
m on mission and identity am ong all units, along 
w ith  com m on service objectives. O ne goal of the 
survey was to assess if th a t mission h ad  been accom ­
plished.

T he survey results w ere used to set priorities for 
budgeting across all inform ation service units, and 
w ithin each u n it’s budget. In  th e L ib rary , faculty 
felt th a t co llectio n  d e v e lo p m e n t w as th e  o v er­
w helm ing priority and th a t reference and research 
services w e re  a d e q u a te  to m eet s tu d e n t needs. 
T h e re  w as less c o n cern  for lib ra ry  in stru c tio n  
across the curriculum  th a n  h ad  been anticipated.

In  a developing institution like H aw aii Loa Col­
lege, these considerations w ere crucial to budget­
ing a $500,000 g ran t th e L earning Resources C en­
ter received from  a p riv ate foundation. Faculty



January 1986 /  17

members, responding to a n u m b er of items on the 
Library’s holdings, felt th a t th e critical mass of m a ­
terials in the general circulating collection needed

to support th eir area of th e curriculum  was not yet 
av ailab le a n d  needed to be given first p rio rity  
above all other needs in any area of th e LRC.

The Tufts/EDUCOM data-sharing project 
for library statistics

By John A. D u n n  Jr.

Vice-President, Planning 
Tufts University

The Tufts University d ata-sh arin g  project sup­
ports college and university p lan n in g  and m an ag e­
ment by facilitating self-assessment an d  co m p ari­
sons w ith  p e e rs , u s in g  c o m p u te r - s u p p o r t e d  
data-aggregation an d  analysis techniques. A p ri­
mary advantage of th e d atab ase approach to li­
brary statistics is th a t it gives th e user access to d a ta  
on a more cu rren t basis th a n  is generally possible 
with paper surveys.

The project has th ree com ponents: E D U C O M ’s 
Higher E ducation D ata-S h arin g  Service (HEDS) 
software; sets of definitions an d  ratios (d ata p ro ­
files) developed by Tufts U niversity w ith  th e guid­
ance of the m em bers; an d  collections of d a ta  con­
trib u ted  by th e  m e m b e r schools. T h e  H E D S  
software and th e d atab ase reside on an IBM m a in ­
frame at C ornell University.

The set of d a ta  to be collected is based on d a ta  
already b e in g  c o lle c te d  by A R L , A C R L , a n d  
LIBGIS, as well as by A rth u r Monke at Bowdoin 
for his college survey. It also goes beyond those and 
beyond th e ARL supplem entary questionnaire in 
the area of au to m atio n , and is m ore inclusive of 
other indicators of institutional size an d  ch aracter. 
The softw are allows th e com puter on w hich the 
database resides to perform  th e ratio  calculations 
for the user, so th a t th e o u tp u t includes ratios and 
trend indicators as well as raw  d ata.

Each user collects d a ta  for his or her ow n institu ­
tion following th e profile descriptions, an d  enters 
them using T elenet, TY M NET, B IT N E T  or other 
data com m unications netw orks. T he user can then 
obtain:

•  time-series d a ta  for his or her ow n institution, 
and for any other p a rtic ip a n t, including differ­
ences betw een those sets of d a ta  in absolute or p e r­
centage term s;

•  d ata for any given year for all institutions or 
for the set of schools specified (access to peer group 
data is by consent of th e  m em bers);

•  statistical m easures on each variable, for all in ­
stitutions or for th e set selected, as well as several 
types of graphic displays of th e d ata.

The user can en ter an d  p rin t out th e d a ta  in

“p u re tim e-sharing” m ode using an o rd in ary  te rm i­
nal or m odem . A lternatively, in “m icrocom puter- 
to -m ain fram e” m ode, th e user can em ploy sp read ­
sheet softw are (e.g., LOTUS 1-2-3), to en ter or 
extract d a ta  by file transfer, using th e m icrocom ­
p u te r for fu rth e r local analysis and graphics.

T he areas of d a ta  collection and analysis include 
financial statistics such as o perating incomes and 
expenses, endow m ents, p riv ate support, an d  b a l­
ance sheet changes; statistics on u n d e rg rad u ate a d ­
missions, enrollm ents (by level and by degree p ro ­
g r a m ) , s tu d e n t c h a rg e s  a n d  f i n a n c ia l a id ; 
institutional d a ta  in such areas as personnel an d  fa ­
cilities; sponsored research; libraries an d  faculty 
dem ographics. A profile on faculty com pensation 
is in p re p a ra tio n . Profiles include both th e base in ­
p u t d a ta  an d  a w ide variety of com puted ratios, 
grow th rates, an d  com parisons to n atio n al statis­
tics.

C u rre n t un iv ersity  p a rtic ip a n ts  are  B randeis, 
Carnegie-M ellon, C ornell, E m ory, G eorgetow n, 
NYU, Pennsylvania, Rochester, Southern M ethod­
ist, T u fts, T u la n e , V a n d e rb ilt an d  W ash in g to n  
University. College m em bers are A m herst, Bates, 
B o w d o in , B ry n  M a w r, B u c k n e ll, C a r l e to n , 
C larem ont-M cK enna, C lark, C olgate, C olorado, 
D ic k in so n , F r a n k lin  a n d  M a rs h a ll, G r in n e ll, 
H am ilto n , H averford, K alam azoo, Kenyon, L a ­
fa y e tte , L a w re n c e , L e h ig h , L ew is a n d  C la rk , 
M iddlebury, Mills, M ount Holyoke, O berlin, Po­
m ona, Reed, Scripps, Sm ith, St. Jo h n ’s (A nnapo­
lis), St. L aw rence, S w arthm ore, T rin ity  College, 
T rin ity  U n iv e rsity , U n io n , V assar, W ellesley, 
W esleyan, W h eato n , an d  W illiam s. Several other 
colleges an d  universities are considering joining.

P articip atio n  w ith in  th e tw o user groups is vol­
u n tary . O nly a few libraries now  have d a ta  in the 
system, b u t m ore have indicated a willingness to 
join actively.

F u tu re  directions include expanding and fu rth e r 
refining th e areas of d a ta  collection.

Interested lib rarian s m ay contact m e at (617) 
381-3274 to find o u t how  to p a rtic ip a te  most effec­
tively. ■ ■