toward a tqm paradigm: using servqual to measure · pdf fileservqual to measure library...

Toward a TQM Paradigm 237

237

Toward a TQM Paradigm: UsingSERVQUAL to Measure LibraryService Quality

Vicki Coleman, Yi (Daniel) Xiao, Linda Bair, and BillChollett

This study provides the results of a survey conducted in the fall of 1994by the Sterling C. Evans Library to measure service quality. This gen-eral user survey provided feedback from customers on their minimum,perceived, and desired levels of service from an academic library. Thedevised measuring instrument is based on SERVQUAL, a service qual-ity survey created by Leonard L. Berry, A. Parasuraman, and Valarie A.Zeithaml. The SERVQUAL survey is designed to measure service qualityin five dimensions: tangibles, reliability, responsiveness, assurance, andempathy. Survey results showed a discrepancy in the quality of theservices provided by the library and those desired by its customers.

All four authors work at the Sterling C. Evans Library at Texas A&M University. Vicki Coleman is Head ofElectronic Reference Services; e-mail: [email protected]. Yi (Daniel) Xiao is Assistant Head of ElectronicReference Services; e-mail: [email protected]. Linda Bair is Senior Library Specialist and Chairperson ofthe Library Service Quality Committee; e-mail: [email protected]. Bill Chollett is System Administrator e-mail: [email protected].

oncepts of total quality man-agement (TQM) indicatethat evaluation of libraryservice quality is based on

customer perception.1 This raises thequestion: How do you measure servicequality in terms of a provider�s perfor-mance with respect to a customer�s ex-pectations? For this study, SERVQUALprovided the solution. Developed in 1988by Leonard L. Berry, A. Parasuraman,and Valarie A. Zeithaml, SERVQUAL isan instrument designed to measureservice quality on the basis of acustomer�s minimum, perceived, anddesired levels of performance.

This study examines the results ofthe SERVQUAL survey distributed at a

large university. The objectives of thesurvey were: (1) to define library ser-vice quality, (2) to determine how toimprove it, and (3) to assess the dimen-sions of quality most important to li-brary customers. For this study, servicequality is determined as the discrepancybetween the minimum, perceived, anddesired levels of performance acrossfive dimensions (tangibles, reliability,responsiveness, assurance, and empa-thy). Exceeding what customers expectfrom the service is the key to deliveringhigh-quality service.2 TQM principlesemphasize the use of a decentralized,team-oriented approach to address theissues causing the gaps in customerexpectations and perceptions.3 Analy-

238 College & Research Libraries May 1997

tations: desired service (which reflectswhat customers want); adequate ser-vice (the standard customers are will-ing to accept); and predicted service (thelevel of service customers believe islikely to occur).5 Their research also sup-ports the theory that customers gener-ally use five factors, or dimensions, astheir criteria for judging service quality.The dimensions as listed in table 1 arenot mutually exclusive but provide afoundation for understanding what cus-tomers expect from service providers.6

Upon completion of this study, an-other literature search was conducted.The authors found that Marilyn D. Whiteand Eileen G. Abels had conducted asurvey of the service marketing litera-ture for models and data-gathering in-struments measuring service quality.8

That study focuses on SERVQUAL andSERVPERF (an instrument used to gen-erate a performance-based measure ofservice quality), and assesses their ap-plicability to special libraries and infor-mation centers. The authors� findingsshow that the validity of the individualtwenty-two items on the SERVQUALsurvey has rarely been disputed. How-ever, there has been criticism ofSERVQUAL in other areas, specifically,the SERVQUAL scale�s theoretical base

sis of the survey results offers manag-ers and quality teams a guide for devis-ing strategic plans to improve servicequality in the library.

Literature ReviewA review of the literature on service

quality in libraries was conducted be-fore and after this study. The literaturereview prior to the study revealed onlyone documented source of a libraryactually using the SERVQUAL instru-ment. Françoise Hebert usesSERVQUAL specifically for the investi-gation of interlibrary loan service qual-ity in large public libraries in Canada.In her data analysis, she compares thelibrary�s measures of interlibrary loanservice quality to those of the respond-ing customers. Her study concludes withthe finding that there is a mismatch be-tween library measures of interlibraryloan performance, based on fill rate andturnaround time, and customer mea-sures of quality, based ondisconfirmation theory.4

Where Hebert focuseson one library service (in-terlibrary loan), the pur-pose of this study was toobtain from library cus-tomers an �aura� of over-all library service. First, itwas necessary to under-stand the nature and de-terminants of customerexpectations; under-standing customer expec-tations is a prerequisitefor delivering superiorservice. Berry, Parasura-man, and Zeithaml sug-gest that there are threelevels of customer expec-

TABLE 1Five Dimensions of Quality Defined7

Dimensions Definitions

Reliability The ability to perform the promisedservice dependably and accurately.

Responsiveness The willingness to help customers andprovide prompt service.

Assurance The knowledge and courtesy ofemployees and their ability to conveytrust and confidence.

Empathy The caring, individualized attentiongiven to customers.

Tangibles The appearance of physical facilities,equipment, personnel, and communi-cation materials.

Understanding customer expectationsis a prerequisite for delivering superiorservice.


and the number and generic nature ofthe dimensions. White and Abels con-clude that, with slight modifications, thesurvey can adequately reflect the rangeof values that library users attach toinformation centers.

MethodologyBackgroundTexas A&M University (TAMU), locatedin College Station, has approximately

44,000 students of which nearly 20 per-cent are graduate students. The univer-sity employs approximately 2,500 fac-ulty and 13,000 staff members. TheSterling C. Evans Library is the mainlibrary servicing the educational and re-search needs of the university and thelocal community. With more than twomillion volumes, four million micro-forms, and 13,000 periodical subscrip-tions, the library is the major informa-

FIGURE 1Evans Library Customer Satisfaction Survey (Part 1)

We would like your impressions about Evans’ service performance relative to yourexpectations. Please think about the two different levels of expectations as definedbelow:

Minimum Service Level: the minimum service performance you consideradequate

Desired Service Level: the level of service performance you desire

For each of the following statements, please indicate: (a) your minimum service level bycircling one of the numbers in the first column; (b) your desired service level by circlingone of the numbers in the second column; and (c) your perception of Evans’ service bycircling one of the numbers in the third column.Q.2 Employees who are consistently courteousQ.3 Employees who deal with customers in a caring fashion

My Minimum Service My Desired Service My Perception of Evans’When it comes to. . . Level Is: Level Is: ServicePerformance Is:

Low High Low High Low High NoOpinion

Q.1 Prompt serviceto customers 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 N

Q.4 Providing service at the promised timeQ.5 Employees who understand the needs of

their customersQ.6 Visually appealing materials associated with

the service (e.g., clear and concise forms)Q.7 Having the customer’s best interest at heartQ.8 Willingness to help customersQ.9 Maintaining error-free customer and catalog

recordsQ.10 Keeping customers informed about when

services will be performedQ.11 Providing services as promisedQ.12 Employees who instill confidence in

customers

Q.13 Employees who have the knowledge toanswer customers’ questions

Q.15 Dependability in handling customers’service problems

Q.14 Readiness to respond to customers’questions

Q.16 Performing services right the first timeQ.17 Visually appealing facilitiesQ.18 Giving customers individual attentionQ.19 Employees who have a neat, professional

appearanceQ.20 Convenient business hoursQ.21 Modern equipmentQ.22 Assuring customers of the accuracy and

confidentiality of their transactions


tion resource for TAMU students andfaculty.

The library offers its customersmany services, including orientations,tours and demonstrations, book andjournal circulation privileges, interli-brary loan services, remote computeraccess to the library computer system,access to computerized literaturesearches, access to audiovisual mate-rials and equipment, more than 150 mi-crocomputers, and other services. Astheir comments indicate, many of theseservices were considered during thecustomers� evaluation of the quality ofthe library�s service.

Research QuestionsService quality is measured in terms ofhow well a provider performs with re-gard to a customer �s expectations.9

Data gathered from the measurementof service quality can be used to an-swer the following researched ques-tions:

1. How do customers define libraryservice quality?

2. How can library service quality beimproved?

3. What dimensions of quality aremost important to library customers?

The Survey InstrumentThis survey is based on the modelSERVQUAL designed by Berry, Parasur-aman, and Zeithaml, a multiple-itemscale used to measure customer expec-tations and perceptions of service qual-ity.10 The survey is divided into two parts.Part 1 (see figure 1 for a partial samplesurvey) contains three sets of twenty-two questions each, where a customerindicates his or her minimum, perceived,and desired service acceptance levelson a scale ranging from a low of 1 to ahigh of 9. The twenty-two questions canbe divided into five sets, each of whichrepresents one of the five dimensions ofquality (reliability, responsiveness, assur-ance, empathy, and tangibles). Part 2 (see

figure 2 for a partial sample survey) al-lows the customer to give an overall rat-ing of library service quality and to weighfive different dimensions of quality per-taining to services offered in academiclibraries. Berry, Parasur-aman, andZeithaml speak in terms of a zone of toler-ance, which is the area between the mini-mum and desired acceptance levels.Quality is measured by determiningwhere perceived performance falls withrespect to the zone of tolerance. Scoresalso can be weighted to measure qualitywith respect to the dimension(s) most im-portant to the customer.

The original SERVQUAL surveyquestions were slightly modified to bemore specific to the library. They werelisted randomly so that sets of dimen-sional questions were not all groupedtogether. This was done to eliminateany biases a respondent may have hadwith respect to a particular dimension.

To ensure validity of the survey in-strument, Berry, Parasuraman, andZeithaml derived the five dimensionsfrom a systematic analysis of customerratings from hundreds of interviews inseveral service sectors. SERVQUALalso was tested in five different servicesectors: product repair and mainte-nance, retail banking, long-distance tele-phone, securities brokerage, and creditcards. The dimensions were determinedto be a concise depiction of the core cri-teria that customers apply in assessingservice quality.11

To supplement the raw data from thesurvey, the authors posed the followingquestion at the end of the survey: �Ifyou could say anything to the Directorof Evans Library, what would it be?� Bysoliciting input in this manner, the com-ments provided explanations for thesurvey raw data.

Data Collection and ProcessingA total of 525 surveys was sent out: 125each to faculty, staff, graduate, and un-dergraduate students; and 25 to com-


FIGURE 2Evans Library Customer Satisfaction Survey, Part 2

1. How would you rate the overall quality of service provided by Evans Library?(Circle one number below.)

Extremely Poor Extremely Good1 2 3 4 5 6 7 8 9

2. Listed below are five general features pertaining to academic libraries and theservices they offer. We would like to know how important each of these features isto you when you evaluate an academic library’s quality of service.

Please allocate a total of 100 points among the five features according to howimportant each feature is to you—the more important a feature is to you, the morepoints you should give it. Please be sure the points you give add up to 100.

1. The appearance of the library’s physical facilities, equipment, pointspersonnel, and communications materials.

2. The library’s ability to perform the promised services dependably pointsand accurately.

3. The library’s willingness to help customers and provide prompt pointsservice.

4. The knowledge and courtesy of the library’s employees and their pointsability to convey trust and confidence.

5. The caring, individualized attention the library provides its customers. points

Total Points Allocated 100 points

munity users. Surveys sent to facultyand students were distributed in pro-portion to the size of the respectiveschool (i.e., engineering, architecture,etc.). Names of survey recipients wererandomly generated by computer da-tabases. The above steps were takento ensure reliability of the survey re-sults. To encourage selected partici-pants to respond, the authors offeredthem small tokens of appreciation andthe option to receive the results of thesurvey if they so desired.

The responses from each surveywere manually entered into a comput-erized file manager. A printout wasmade of the raw data so that the au-thors could more easily double-checkentries for any errors. Once the datahad been double-checked, they wereimported into spreadsheet software for

the purpose of plotting graphs. It shouldbe noted that some of the survey par-ticipants had never visited the librarybefore. On the survey, they were askedto enter �No opinion� for all responseswhere they had no perception of ser-vice performance. Any averages cal-culated from the data were based onthe number of responses to a question.There was no numerical value attachedto the �No opinion� response.

Results are analyzed by determin-ing where the perceived data points fallwith respect to the zone of tolerance.Any data points falling within the zoneindicate that the customer finds the ser-vice tolerable, those falling below thezone indicate that the customer�s mini-m u macceptable service requirements havenot been met; and those falling above


the zone indicatethat the customer�sdesired service levelhas been ex-ceeded.12 Each of thetwenty-two ques-tions is treated as aseparate entity andhas its own zoneof tolerance. There isno set range forthe zone of toler-ance because it isbased sole-ly on the customer�s mini-mum acceptable service level and de-sired level of performance. However,the goal is that the perceived scores willeither equal or surpass the minimumlevel of performance.

To better illustrate the zone of toler-ance concept, the authors used bargraphs. (Graphs are explained ingreater detail in the section on surveyresults.) Question numbers are repre-sented on the X-axis, and levels of per-formance (minimum, perceived, anddesired) for each series are graphed onthe Y-axis.

Calculations(Note: Instructions for calculations areattributed to Zeithaml, Parasuraman,and Berry.13)

Unweighted SERVQUAL Scores. Un-weighted scores are those scores calcu-lated without taking into account the rela-tive importance that customers attachedto a dimension. They are calculated byaveraging customers� SERVQUALscores on the statements comprising adimension. For example, if N customersresponded to a SERVQUAL survey, theaverage SERVQUAL score along eachdimension is obtained by following thesesteps:

1. For each customer, add theSERVQUAL scores on the statements per-taining to the dimension and divide thesum by the number of statements makingup the dimension.

TABLE 2Averages by Category

Surveys Average Average AverageReceived Minimum Perceived Desired

Faculty 35 6.3 6.3 8.0Staff 23 6.4 6.0 7.9Graduate 64 6.3 6.2 8.0Undergrad. 70 6.2 6.2 8.0Community 6 5.5 7.3 8.2

Total/Avg. 198 6.1 6.4 8.0

2. Add the quantity obtained in stepone across all N customers and dividethe total by N. To obtain an overall aver-age of service quality, average theSERVQUAL scores over the five dimen-sions (i.e., summed and divided by five).

Weighted SERVQUAL Scores.Weighted scores take into account therelative importance of the dimensions.The following steps were taken to cal-culate the weighted scores:

1. For each customer, add theSERVQUAL scores on the statements per-taining to the dimension and divide the sumby the number of statements making upthe dimension.

2. For each customer, multiply theSERVQUAL score for each dimension(obtained in step one) by the importanceweight assigned by the customer to thatdimension. The importance weight is thepoints the customer allocated to the di-mension divided by 100.

3. For each customer, add theweighted SERVQUAL scores (obtainedin step two) across all five d i -mensions to obtain a combinedSERVQUAL score.

4. Add the scores (obtained in stepthree) across all N customers and di-vide the total by N.

Survey ResultsNearly 200 completed surveys were re-turned. The overall response rate was38 percent. The highest response camefrom the students, where more than 50


percent responded. The survey scaleranges from a low of 1 to a high of 9. Bytaking the average of the twenty-twoquestions (which encompass all five di-mensions), the average levels of accep-tance are shown in table 2.

Of those responding to the survey,more than 83 percent had visited thelibrary before. Nearly 50 percent of thefaculty and student respondents visitthe library weekly. More than 60 per-cent of the responding staff and com-munity users visit the library on amonthly basis. Table 3 shows the fre-quency of visitation based on thegroups surveyed.

In Part 2 of the survey, participantsranked dimensions by allocating 100points among them to indicate eachdimension�s importance. Reliabilityranked first, with responsiveness aclose second. Empathy was determinedto be the least important dimension. Incomparison to �other studies� con-ducted by Berry andcolleagues, the dimen-sions were ranked in thesame order with the ex-ception of tangibles andempathy. The relativeimportance of dimen-sions for both studieswas similar for respon-siveness, assurance,and empathy, as shownin table 4.14

In the portion of thesurvey allotted for di-

rect comments from the cus-tomer to the library director, veryspecific feedback was provided.Most survey participants ex-pressed areas of dissatisfactionwith the library. Topics com-mented on included: difficulty infinding resources, circulation poli-cies, unpleasant library staff, ex-penses associated with duplica-tion services, outdated re-sources, building accessibility

hours, and missing/lost materials. Themore positive comments pertained to theattitude and work ethics of the librarystaff.

The unweighted SERV-QUAL scores across each individual di-mension (tangibles, reliability, respon-siveness, assurance, and empathy) areshown on the graphs in figures 3�7. Onthe graphs, a blank gap indicates roomfor possible improvement, shaded ar-eas show acceptable service, and solidblack areas indicate unacceptable ser-vice. Figure 1, Tangibles, is the only di-mension where the library consistentlyperformed within the zone of tolerance(i.e., perceived scores are between de-sired and minimum). Figure 8 shows thescores averaged over the five dimen-sions; reliability shows the highest mini-mum and desired service scores indi-cating that customers expect the most,and will tolerate the least, in this dimen-sion.

TABLE 3Frequency of Visitation

to the Evans Library

n=198 Daily Weekly Monthly Never

Faculty 12.5% 46.9% 37.5% 3.1%Staff 4.5 18.2 63.6 13.6Graduate 24.2 50.0 24.2 1.6Undergrad. 24.3 52.9 21.4 1.4Community 0.0 16.7 66.7 16.7

TABLE 4Relative Importance of Service Quality Dimensions

Relative ImportanceThis Study Other Studies

Dimension Rank n=198 n=1,936

Reliability 1 26.7% 32%Responsiveness 2 23.6 22Assurance 3 18.8 19Tangibles 4 16.2 11*Empathy 5 14.8 16** Note differences in rankings of “This Study” vs “Other Studies.”


FIGURE 3Graph Showing Tangibles

FIGURE 4Graph Showing Reliability


FIGURE 5Graph Showing Responsiveness

FIGURE 6Graph Showing Assurance


FIGURE 7Graph Showing Empathy

FIGURE 8Graph Showing the Average for Each Dimension


FIGURE 9Graph Showing the Weighted Average for Each Dimension

In figure 9, the scores for each di-mension were weighted based on thepoints customers allocated to the di-mension. This graph shows the poten-tial opportunity for improving quality ofservice and suggests where emphasisshould be placed toward continuous im-provement. Respondents weighted re-liability as the most important and em-pathy the least important dimension.The importance of reliability is readilyapparent by comparing the downwardshift of the other four dimensions asseen in figure 9. Overall, the library con-sistently performed at or above theminimum acceptance level, with theexception of the reliability dimension.

ConclusionInterpretation of the tables and graphsshows that the survey provided answersto the original research questions.

Those questions were:1. How do customers define library

service quality?2. How can library service quality be

improved?3. What dimensions of quality are

most important to library customers?

Customers define service quality asthe extent of discrepancy between theirexpectations or desires and their per-ceptions.15 In this study, discrepanciesprevailed along dimensions of reliabil-ity, responsiveness, assurance, and em-pathy. Only the tangibles dimension(appearance of physical facilities,equipment, personnel, and communica-tion materials) surpassed the custom-ers� minimum acceptable service level,and this dimension was viewed by cus-tomers as the second least importantfor measuring quality. Tangibles was


the only dimension where the libraryperformed consistently within the zoneof tolerance.

For this study, customer ranking ofthe relative importance of the five ser-vice quality dimensions is as follows: re-liability, responsiveness, assurance, tan-gibles, and empathy. These findings arenot unique in that other studies haveconsistently shown similar results, withthe exception of tangibles and empa-thy. Reliability relates to the library�s

ability to perform promised services de-pendably and accurately. It was rankedas the most important dimension formeasuring quality; unfortunately, per-ception of the library�s performance waslowest for this dimension. Empathy (thecaring, individualized attention the li-brary provides its customers) was theleast important of the five dimensions.The library definitely needs to improvein the areas of reliability and respon-siveness because of their importanceto customers.

The zone of tolerance is the area be-tween the minimum and desired accep-tance levels of performance. Of thelibrary�s overall perceived performancefor the twenty-two questions, fourteen(64%) of the scores fell within the zoneof tolerance, six (27%) fell below the mini-mum acceptance level, and two (9%)were averaged right at the minimumacceptance level. Most of the scoreswithin the zone of tolerance were nearerthe minimum than the desired level ofservice. None of the perceived scoreswas above the desired service level. Al-together, the data clearly indicate thatthere is a mismatch between the priori-ties expressed by the library�s custom-

ers and the levels of service qualitydelivered by the library.

Topics for Further InvestigationSeveral variables may have impactedthe outcome of this investigation. Forexample, an individual�s frequency ofvisitation to the library may have influ-enced survey scores. Also, servicesprovided to customers vary by cus-tomer status (i.e., faculty, undergradu-ate, etc.). Those only receiving baselineservices may have evaluated the librarymore harshly in some areas. The surveyinstrument was rather lengthy, and per-haps some individuals did not take thetime to thoroughly read through it. Forfuture surveys, more effort should beplaced on increasing the response rateof 38 percent. A higher response rate willensure more confidence in the validityof the survey data as they apply to alllibrary customers.

Various teams, referred to asClouds, have been created in the libraryto address issues of service quality per-tinent to library customers. The termClouds is used because each team iscomposed of personnel from variousunits; hence, membership shifts acrossthe organization. The teams have de-vised strategic plans to bring the presentlevel of service up to the customers�desired level. A library advisory com-mittee consisting of students, faculty,and library administrators has beencreated. This committee will review therecommendations of the Clouds to en-sure that the issues most urgent to li-brary users are being addressed. A fu-ture survey, scheduled for spring 1997,will show the impact of the teams onlibrary service quality.

Comments from survey participantsindicated that library users do not lookat the operation of individual units withinthe library system but, instead, view thelibrary as one enterprise. For example,survey participants expressed dissat-isfaction with book availability within the

Reliability was ranked as the mostimportant dimension for measuringquality; unfortunately, perception ofthe library�s performance waslowest for this dimension.


Notes

1. Rosanna M. O�Neil, Total Quality Management in Libraries: A Sourcebook (Englewood,Colo.: Libraries Unlimited, 1994), 2.

2. Valarie A. Zeithaml, A. Purasuraman, and Leonard L. Berry, Delivering Quality Service:Balancing Customer Perceptions and Expectations (New York: Free Pr., 1990), 19�26.

3. O�Neil, Total Quality Management in Libraries, 6�7.4. Françoise Hebert, �Service Quality: An Unobtrusive Investigation of Interlibrary Loan

in Large Public Libraries in Canada,� LISR 16 (1994): 3�21.5. Valarie A. Zeithaml, Leonard L. Berry, and A. Parasuraman, �The Nature and

Determinants of Customer Expectations of Service,� Journal of the Academy of MarketingScience 21, no. 1 (1993): 1�12.

6. Zeithaml, Purasuraman, and Berry, Delivering Quality Service, 19�26.7. Zeithaml, Berry, and Parasuraman, �The Nature and Determinants of Customer Expec-

tations of Service,� 1�12.8. Marilyn D. White and Eileen G. Abels, �Measuring Service Quality in Special Librar-

ies: Lessons from Service Marketing,� Special Libraries 86, no. 4 (1995): 36�45.9. Zeithaml, Purasuraman, and Berry, Delivering Quality Service, 19.

10. Ibid., 181�86.11. Ibid., 26.12. Leonard L. Berry, interview by Linda Bair and Bill Chollett, Texas A&M University,

June 1994.13. Zeithaml, Purasuraman, and Berry, Delivering Quality Service, 176�77.14. Ibid., 28.15. Ibid., 19.

library but did not specifically attack theCirculation Department, which is re-sponsible for shelving, or CollectionDevelopment, which reorders missing

materials. The challenge for the EvansLibrary is to provide great service as asystem by continually improving uponthe services provided by each unit.

toward a tqm paradigm: using servqual to measure · pdf fileservqual to measure library...

Documents