coleman.p65 Toward a TQM Paradigm 237 237 Toward a TQM Paradigm: Using SERVQUAL to Measure Library Service Quality Vicki Coleman, Yi (Daniel) Xiao, Linda Bair, and Bill Chollett This study provides the results of a survey conducted in the fall of 1994 by 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. The devised 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 quality in five dimensions: tangibles, reliability, responsiveness, assurance, and empathy. Survey results showed a discrepancy in the quality of the services 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 of Electronic Reference Services; e-mail: vcoleman@tamu.edu. Yi (Daniel) Xiao is Assistant Head of Electronic Reference Services; e-mail: danxiao@tamu.edu. Linda Bair is Senior Library Specialist and Chairperson of the Library Service Quality Committee; e-mail: l-bair@tamu.edu. Bill Chollett is System Administrator e-mail: b-chollett@tamu.edu. oncepts of total quality man- agement (TQM) indicate that evaluation of library service quality is based on customer perception.1 This raises the question: How do you measure service quality in terms of a provider �s perfor- mance with respect to a customer�s ex- pectations? For this study, SERVQUAL provided the solution. Developed in 1988 by Leonard L. Berry, A. Parasuraman, and Valarie A. Zeithaml, SERVQUAL is an instrument designed to measure service quality on the basis of a customer �s minimum, perceived, and desired levels of performance. This study examines the results of the SERVQUAL survey distributed at a large university. The objectives of the survey were: (1) to define library ser- vice quality, (2) to determine how to improve it, and (3) to assess the dimen- sions of quality most important to li- brary customers. For this study, service quality is determined as the discrepancy between the minimum, perceived, and desired levels of performance across five dimensions (tangibles, reliability, responsiveness, assurance, and empa- thy). Exceeding what customers expect from the service is the key to delivering high-quality service.2 TQM principles emphasize the use of a decentralized, team-oriented approach to address the issues causing the gaps in customer expectations and perceptions.3 Analy- 238 College & Research Libraries May 1997 tations: desired service (which reflects what customers want); adequate ser- vice (the standard customers are will- ing to accept); and predicted service (the level of service customers believe is likely to occur).5 Their research also sup- ports the theory that customers gener- ally use five factors, or dimensions, as their criteria for judging service quality. The dimensions as listed in table 1 are not mutually exclusive but provide a foundation 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. White and Eileen G. Abels had conducted a survey of the service marketing litera- ture for models and data-gathering in- struments measuring service quality.8 That study focuses on SERVQUAL and SERVPERF (an instrument used to gen- erate a performance-based measure of service quality), and assesses their ap- plicability to special libraries and infor- mation centers. The authors� findings show that the validity of the individual twenty-two items on the SERVQUAL survey has rarely been disputed. How- ever, there has been criticism of SERVQUAL 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 service quality in the library. Literature Review A review of the literature on service quality in libraries was conducted be- fore and after this study. The literature review prior to the study revealed only one documented source of a library actually using the SERVQUAL instru- ment. Françoise Hebert uses SERVQUAL specifically for the investi- gation of interlibrary loan service qual- ity in large public libraries in Canada. In her data analysis, she compares the library�s measures of interlibrary loan service quality to those of the respond- ing customers. Her study concludes with the finding that there is a mismatch be- tween library measures of interlibrary loan performance, based on fill rate and turnaround time, and customer mea- sures of quality, based on disconfirmation theory.4 Where Hebert focuses on one library service (in- terlibrary loan), the pur- pose of this study was to obtain from library cus- tomers an �aura� of over- all library service. First, it was necessary to under- stand the nature and de- terminants of customer expectations; under- standing customer expec- tations is a prerequisite for delivering superior service. Berry, Parasura- man, and Zeithaml sug- gest that there are three levels of customer expec- TABLE 1 Five Dimensions of Quality Defined7 Dimensions Definitions Reliability The ability to perform the promised service dependably and accurately. Responsiveness The willingness to help customers and provide prompt service. Assurance The knowledge and courtesy of employees and their ability to convey trust and confidence. Empathy The caring, individualized attention given to customers. Tangibles The appearance of physical facilities, equipment, personnel, and communi- cation materials. Understanding customer expectations is a prerequisite for delivering superior service. Toward a TQM Paradigm 239 and the number and generic nature of the dimensions. White and Abels con- clude that, with slight modifications, the survey can adequately reflect the range of values that library users attach to information centers. Methodology Background Texas A&M University (TAMU), located in 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. The Sterling C. Evans Library is the main library servicing the educational and re- search needs of the university and the local community. With more than two million volumes, four million micro- forms, and 13,000 periodical subscrip- tions, the library is the major informa- FIGURE 1 Evans Library Customer Satisfaction Survey (Part 1) We would like your impressions about Evans’ service performance relative to your expectations. Please think about the two different levels of expectations as defined below: Minimum Service Level: the minimum service performance you consider adequate Desired Service Level: the level of service performance you desire For each of the following statements, please indicate: (a) your minimum service level by circling one of the numbers in the first column; (b) your desired service level by circling one of the numbers in the second column; and (c) your perception of Evans’ service by circling one of the numbers in the third column. Q.2 Employees who are consistently courteous Q.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 No Opinion Q.1 Prompt service to 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 time Q.5 Employees who understand the needs of their customers Q.6 Visually appealing materials associated with the service (e.g., clear and concise forms) Q.7 Having the customer’s best interest at heart Q.8 Willingness to help customers Q.9 Maintaining error-free customer and catalog records Q.10 Keeping customers informed about when services will be performed Q.11 Providing services as promised Q.12 Employees who instill confidence in customers Q.13 Employees who have the knowledge to answer 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 time Q.17 Visually appealing facilities Q.18 Giving customers individual attention Q.19 Employees who have a neat, professional appearance Q.20 Convenient business hours Q.21 Modern equipment Q.22 Assuring customers of the accuracy and confidentiality of their transactions 240 College & Research Libraries May 1997 tion resource for TAMU students and faculty. The library offers its customers many services, including orientations, tours and demonstrations, book and journal circulation privileges, interli- brary loan services, remote computer access to the library computer system, access to computerized literature searches, access to audiovisual mate- rials and equipment, more than 150 mi- crocomputers, and other services. As their comments indicate, many of these services were considered during the customers� evaluation of the quality of the library�s service. Research Questions Service quality is measured in terms of how well a provider performs with re- gard to a customer �s expectations.9 Data gathered from the measurement of service quality can be used to an- swer the following researched ques- tions: 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? The Survey Instrument This survey is based on the model SERVQUAL designed by Berry, Parasur- aman, and Zeithaml, a multiple-item scale 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 sample survey) contains three sets of twenty- two questions each, where a customer indicates his or her minimum, perceived, and desired service acceptance levels on a scale ranging from a low of 1 to a high of 9. The twenty-two questions can be divided into five sets, each of which represents one of the five dimensions of quality (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 weigh five different dimensions of quality per- taining to services offered in academic libraries. Berry, Parasur-aman, and Zeithaml 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 determining where perceived performance falls with respect to the zone of tolerance. Scores also can be weighted to measure quality with respect to the dimension(s) most im- portant to the customer. The original SERVQUAL survey questions were slightly modified to be more specific to the library. They were listed randomly so that sets of dimen- sional questions were not all grouped together. This was done to eliminate any biases a respondent may have had with respect to a particular dimension. To ensure validity of the survey in- strument, Berry, Parasuraman, and Zeithaml derived the five dimensions from a systematic analysis of customer ratings from hundreds of interviews in several service sectors. SERVQUAL also was tested in five different service sectors: product repair and mainte- nance, retail banking, long-distance tele- phone, securities brokerage, and credit cards. The dimensions were determined to be a concise depiction of the core cri- teria that customers apply in assessing service quality.11 To supplement the raw data from the survey, the authors posed the following question at the end of the survey: �If you could say anything to the Director of Evans Library, what would it be?� By soliciting input in this manner, the com- ments provided explanations for the survey raw data. Data Collection and Processing A total of 525 surveys was sent out: 125 each to faculty, staff, graduate, and un- dergraduate students; and 25 to com- Toward a TQM Paradigm 241 FIGURE 2 Evans 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 Good 1 2 3 4 5 6 7 8 9 2. Listed below are five general features pertaining to academic libraries and the services they offer. We would like to know how important each of these features is to you when you evaluate an academic library’s quality of service. Please allocate a total of 100 points among the five features according to how important each feature is to you—the more important a feature is to you, the more points 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, points personnel, and communications materials. 2. The library’s ability to perform the promised services dependably points and accurately. 3. The library’s willingness to help customers and provide prompt points service. 4. The knowledge and courtesy of the library’s employees and their points ability 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 faculty and students were distributed in pro- portion to the size of the respective school (i.e., engineering, architecture, etc.). Names of survey recipients were randomly generated by computer da- tabases. The above steps were taken to ensure reliability of the survey re- sults. To encourage selected partici- pants to respond, the authors offered them small tokens of appreciation and the option to receive the results of the survey if they so desired. The responses from each survey were manually entered into a comput- erized file manager. A printout was made of the raw data so that the au- thors could more easily double-check entries for any errors. Once the data had been double-checked, they were imported into spreadsheet software for the purpose of plotting graphs. It should be noted that some of the survey par- ticipants had never visited the library before. On the survey, they were asked to enter �No opinion� for all responses where they had no perception of ser- vice performance. Any averages cal- culated from the data were based on the number of responses to a question. There was no numerical value attached to the �No opinion� response. Results are analyzed by determin- ing where the perceived data points fall with respect to the zone of tolerance. Any data points falling within the zone indicate that the customer finds the ser- vice tolerable, those falling below the zone indicate that the customer�s mini- m u m acceptable service requirements have not been met; and those falling above 242 College & Research Libraries May 1997 the zone indicate that the customer �s desired service level has been ex- ceeded.12 Each of the twenty-two ques- tions is treated as a separate entity and has its own zone of tolerance. There is no set range for the zone of toler- ance because it is based 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 will either equal or surpass the minimum level of performance. To better illustrate the zone of toler- ance concept, the authors used bar graphs. (Graphs are explained in greater detail in the section on survey results.) Question numbers are repre- sented on the X-axis, and levels of per- formance (minimum, perceived, and desired) for each series are graphed on the Y-axis. Calculations (Note: Instructions for calculations are attributed 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 attached to a dimension. They are calculated by averaging customers� SERVQUAL scores on the statements comprising a dimension. For example, if N customers responded to a SERVQUAL survey, the average SERVQUAL score along each dimension is obtained by following these steps: 1. For each customer, add the SERVQUAL scores on the statements per- taining to the dimension and divide the sum by the number of statements making up the dimension. TABLE 2 Averages by Category Surveys Average Average Average Received Minimum Perceived Desired Faculty 35 6.3 6.3 8.0 Staff 23 6.4 6.0 7.9 Graduate 64 6.3 6.2 8.0 Undergrad. 70 6.2 6.2 8.0 Community 6 5.5 7.3 8.2 Total/Avg. 198 6.1 6.4 8.0 2. Add the quantity obtained in step one across all N customers and divide the total by N. To obtain an overall aver- age of service quality, average the SERVQUAL scores over the five dimen- sions (i.e., summed and divided by five). Weighted SERVQUAL Scores. Weighted scores take into account the relative importance of the dimensions. The following steps were taken to cal- culate the weighted scores: 1. For each customer, add the SERVQUAL scores on the statements per- taining to the dimension and divide the sum by the number of statements making up the dimension. 2. For each customer, multiply the SERVQUAL score for each dimension (obtained in step one) by the importance weight assigned by the customer to that dimension. The importance weight is the points the customer allocated to the di- mension divided by 100. 3. For each customer, add the weighted SERVQUAL scores (obtained in step two) across all five d i - mensions to obtain a combined SERVQUAL score. 4. Add the scores (obtained in step three) across all N customers and di- vide the total by N. Survey Results Nearly 200 completed surveys were re- turned. The overall response rate was 38 percent. The highest response came from the students, where more than 50 Toward a TQM Paradigm 243 percent responded. The survey scale ranges from a low of 1 to a high of 9. By taking the average of the twenty-two questions (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 the library before. Nearly 50 percent of the faculty and student respondents visit the library weekly. More than 60 per- cent of the responding staff and com- munity users visit the library on a monthly basis. Table 3 shows the fre- quency of visitation based on the groups surveyed. In Part 2 of the survey, participants ranked dimensions by allocating 100 points among them to indicate each dimension�s importance. Reliability ranked first, with responsiveness a close second. Empathy was determined to be the least important dimension. In comparison to �other studies� con- ducted by Berry and colleagues, the dimen- sions were ranked in the same order with the ex- ception of tangibles and empathy. The relative importance of dimen- sions for both studies was similar for respon- siveness, assurance, and empathy, as shown in table 4.14 In the portion of the survey allotted for di- rect comments from the cus- tomer to the library director, very specific feedback was provided. Most survey participants ex- pressed areas of dissatisfaction with the library. Topics com- mented on included: difficulty in finding resources, circulation poli- cies, unpleasant library staff, ex- penses associated with duplica- tion services, outdated re- sources, building accessibility hours, and missing/lost materials. The more positive comments pertained to the attitude and work ethics of the library staff. The unweighted SERV- QUAL scores across each individual di- mension (tangibles, reliability, respon- siveness, assurance, and empathy) are shown on the graphs in figures 3�7. On the graphs, a blank gap indicates room for possible improvement, shaded ar- eas show acceptable service, and solid black areas indicate unacceptable ser- vice. Figure 1, Tangibles, is the only di- mension where the library consistently performed within the zone of tolerance (i.e., perceived scores are between de- sired and minimum). Figure 8 shows the scores 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 3 Frequency 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 13.6 Graduate 24.2 50.0 24.2 1.6 Undergrad. 24.3 52.9 21.4 1.4 Community 0.0 16.7 66.7 16.7 TABLE 4 Relative Importance of Service Quality Dimensions Relative Importance This Study Other Studies Dimension Rank n=198 n=1,936 Reliability 1 26.7% 32% Responsiveness 2 23.6 22 Assurance 3 18.8 19 Tangibles 4 16.2 11* Empathy 5 14.8 16* * Note differences in rankings of “This Study” vs “Other Studies.” 244 College & Research Libraries May 1997 FIGURE 3 Graph Showing Tangibles FIGURE 4 Graph Showing Reliability Toward a TQM Paradigm 245 FIGURE 5 Graph Showing Responsiveness FIGURE 6 Graph Showing Assurance 246 College & Research Libraries May 1997 FIGURE 7 Graph Showing Empathy FIGURE 8 Graph Showing the Average for Each Dimension Toward a TQM Paradigm 247 FIGURE 9 Graph Showing the Weighted Average for Each Dimension In figure 9, the scores for each di- mension were weighted based on the points customers allocated to the di- mension. This graph shows the poten- tial opportunity for improving quality of service and suggests where emphasis should 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 readily apparent by comparing the downward shift of the other four dimensions as seen in figure 9. Overall, the library con- sistently performed at or above the minimum acceptance level, with the exception of the reliability dimension. Conclusion Interpretation of the tables and graphs shows that the survey provided answers to 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 as the extent of discrepancy between their expectations or desires and their per- ceptions.15 In this study, discrepancies prevailed 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 important for measuring quality. Tangibles was 248 College & Research Libraries May 1997 the only dimension where the library performed consistently within the zone of tolerance. For this study, customer ranking of the relative importance of the five ser- vice quality dimensions is as follows: re- liability, responsiveness, assurance, tan- gibles, and empathy. These findings are not unique in that other studies have consistently shown similar results, with the exception of tangibles and empa- thy. Reliability relates to the library�s ability to perform promised services de- pendably and accurately. It was ranked as the most important dimension for measuring quality; unfortunately, per- ception of the library�s performance was lowest for this dimension. Empathy (the caring, individualized attention the li- brary provides its customers) was the least important of the five dimensions. The library definitely needs to improve in the areas of reliability and respon- siveness because of their importance to customers. The zone of tolerance is the area be- tween the minimum and desired accep- tance levels of performance. Of the library�s overall perceived performance for the twenty-two questions, fourteen (64%) of the scores fell within the zone of tolerance, six (27%) fell below the mini- mum acceptance level, and two (9%) were averaged right at the minimum acceptance level. Most of the scores within the zone of tolerance were nearer the minimum than the desired level of service. None of the perceived scores was above the desired service level. Al- together, the data clearly indicate that there is a mismatch between the priori- ties expressed by the library�s custom- ers and the levels of service quality delivered by the library. Topics for Further Investigation Several variables may have impacted the outcome of this investigation. For example, an individual�s frequency of visitation to the library may have influ- enced survey scores. Also, services provided to customers vary by cus- tomer status (i.e., faculty, undergradu- ate, etc.). Those only receiving baseline services may have evaluated the library more harshly in some areas. The survey instrument was rather lengthy, and per- haps some individuals did not take the time to thoroughly read through it. For future surveys, more effort should be placed on increasing the response rate of 38 percent. A higher response rate will ensure more confidence in the validity of the survey data as they apply to all library customers. Various teams, referred to as Clouds, have been created in the library to address issues of service quality per- tinent to library customers. The term Clouds is used because each team is composed of personnel from various units; hence, membership shifts across the organization. The teams have de- vised strategic plans to bring the present level of service up to the customers� desired level. A library advisory com- mittee consisting of students, faculty, and library administrators has been created. This committee will review the recommendations 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 on library service quality. Comments from survey participants indicated that library users do not look at the operation of individual units within the library system but, instead, view the library as one enterprise. For example, survey participants expressed dissat- isfaction with book availability within the Reliability was ranked as the most important dimension for measuring quality; unfortunately, perception of the library�s performance was lowest for this dimension. Toward a TQM Paradigm 249 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 Marketing Science 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. 1 0 . Ibid., 181�86. 1 1 . Ibid., 26. 1 2 . Leonard L. Berry, interview by Linda Bair and Bill Chollett, Texas A&M University, June 1994. 1 3 . Zeithaml, Purasuraman, and Berry, Delivering Quality Service, 176�77. 1 4 . Ibid., 28. 1 5 . Ibid., 19. library but did not specifically attack the Circulation Department, which is re- sponsible for shelving, or Collection Development, which reorders missing materials. The challenge for the Evans Library is to provide great service as a system by continually improving upon the services provided by each unit.