American Journal of Pharmaceutical Education 2003; 67 (4) Article 108. RESEARCH ARTICLES Identification of the Service Quality Dimensions of Pharmaceutical Education David Holdford, PhD, and Anuprita Patkar, MSPharm Department of Pharmacy, Virginia Commonwealth University Objectives. Describe the perceptual dimensions of student assessments of the quality of their education. Link those dimensions with student satisfaction with their educational experience. Methods. A 37-item educational service quality instrument and a seven-item satisfaction scale were administered to 372 students in their final year of education in years 1999 to 2002. Results. Factor analysis identified 5 dimensions of service quality labeled resources, interpersonal behavior of faculty, faculty expertise, faculty communication, and administration. Stepwise regression analysis showed that all factors were significantly related to overall satisfaction with each explaining the following percent variance: faculty interpersonal behavior (47%), administration (8%), resources (4%), faculty communication (3%), and faculty knowledge (1%). Mean scores of service quality dimensions demonstrated significant improvement from 1999 to 2002 for four of the five dimensions. Conclusion. The 5 dimensions of educational service quality found in this study are unique to the services marketing literature. Student perceptions of faculty are multidimensional with faculty interpersonal behavior explaining the most variance in student satisfaction responses. Keywords: pharmaceutical education, service quality, assessment, student satisfaction from satisfaction primarily because it is qualityspecific, while satisfaction deals with quality and nonquality (eg, price) evaluations. INTRODUCTION Collecting student feedback about various aspects of pharmaceutical education has become an important part of educational outcomes assessment. Educators have developed numerous feedback mechanisms to assess student perceptions of their educational experiences. Focus groups, town hall meetings, course evaluations, and one-to-one discussions are but a few. Student feedback about educational services is useful for several reasons. It can be used in quality improvement programs to help educators recognize opportunities to improve services and establish positive student perceptions.4,5 Service quality assessments can also identify gaps between students’ perceptions of education and that of educators.5 Educators often have misconceptions of student attitudes because of the anecdotal manner in which student comments are received. Systematic service quality evaluations can help identify the incidence and strength of student feelings. Finally, service quality assessments can be used as a strategic tool for marketing educational programs.1,6 Pharmacy schools are in an increasingly competitive market for students. Schools that do not fulfill the needs and desires of their students will be at a competitive disadvantage in the marketplace. Some researchers have suggested that educators should evaluate student perceptions of educational service quality.1-3 Educational service quality is defined as a student’s overall evaluation of services received as part of their educational experience.3 It describes a variety of educational activities both inside and outside of the classroom including classroom instruction, faculty member/student interactions, educational facilities, and contacts with administration. Service quality differs Corresponding Author: David Holdford, PhD. Mailing Address: Department of Pharmacy, P.O. Box 980533, Virginia Commonwealth University (MCV Campus), Richmond, Virginia 23298-0581. Tel: 804828-6103. Fax: 804-828-8359. E-mail: david.holdford@vcu.edu. Dimensions of Service Quality The services marketing literature has made significant progress exploring fundamental questions regard- 1 American Journal of Pharmaceutical Education 2003; 67 (4) Article 108. ing service quality. One area that has received significant attention is the multidimensional nature of services. In a seminal research study, Parasuraman, Zeithaml, and Berry identified 10 dimensions of service quality - tangibles, reliability, responsiveness, competency, courtesy, communication, credibility, security, access, and understanding, based upon a series of focus group studies.7 Since that study, service quality measures have been used to assess a broad variety of services including physician, hospital, educational, banking, and dental. ample, a measure of service quality may have questions that assess the attributes of speedy service, reliability of service, and friendliness of service personnel. Respondents’ ratings of those attributes are summed to get an overall score that indicates the degree to which the construct pertains to them. If no further analysis of the scale’s dimensionality is conducted, then researchers are left with analyzing the summed score and/or the individual items. This is problematic for several reasons. If only the summed responses are used, significant information is lost. A researcher who sums the ratings on 20 items in a service quality scale into a single score has no way of identifying which attributes of service contribute to the score. Alternatively, if only individual items are used, interpreting the results in an actionable way becomes a problem. When faced with scores for 20 items, it is difficult to identify which responses to which items are more important. This problem is compounded due to multicollinearity between items. Multicollinearity refers to the intercorrelation between items. It reduces the validity of regression analyses of survey results because regression models assume independence of items.12 A regression analysis of service quality items on overall satisfaction with service will violate this assumption because perceptions of individual items such as reliability, responsiveness, and competence will likely be correlated. In other words, a student’s perception of faculty member competence will be significantly associated with perceptions of faculty member friendliness, trustworthiness, and reliability. Therefore, the analysis can give erroneous regression coefficients that incorrectly state the importance of individual items. From that initial research, Parasuraman, Zeithaml, and Berry developed a service quality instrument called SERVQUAL, which evaluated consumer perceptions of services. Factor analysis of consumer responses to SERVQUAL resulted in a conclusion that there are 5 key dimensions of service quality.8 • Reliability - the ability to deliver promised services in a dependable, accurate manner. • Responsiveness - the willingness to help customers and provide prompt service. • Empathy - the degree to which customers are treated as individuals. • Assurance - the ability to inspire trust and confidence. • Tangibles - physical elements of the service such as facilities and equipment. Other researchers found that the 5 dimensional service quality model did not hold up under all service experiences. New studies have demonstrated that the specific number of service dimensions associated with service quality varies depending on the services provided and the circumstances under which they are rendered. Mels et al found 2 dimensions of service quality,9 while Carman suggested 7 dimensions of service quality.10 Cronin and Taylor argued against the conclusion that service quality is multidimensional and instead proposed that the construct of service quality is unidimensional.11 In the end, researchers have concluded that the dimensionality of service quality is situation specific. One way to deal with the problems above is by discovering the underlying perceptual dimensions of the construct measured by the scale.13 Perceptual dimensions describe how respondents organize and arrange perceptions about concepts in their mind. For example, respondents may group questionnaire items about dependability of service, keeping promises to customers, timely delivery of service, and accurate record keeping under a perceptual dimension that can be labeled as reliability. This means that respondents answer the questions in this dimension in a predictable pattern that differs from responses to questions for other dimensions like responsiveness or assurance. If all of the items in a scale can be grouped into separate uncorrelated dimensions, then analyses of scale responses are more valid. Perceptual Dimensions of Service Quality To understand the importance of this research, background is provided on measurement scales and dimensions. A scale that measures a psychological construct such as satisfaction or service quality is made up of individual questions that characterize different attributes of the construct. Each attribute provides specific information about the overall construct. For ex- The dimensional nature of service quality is an important issue because it can be used to explore the theo- 2 American Journal of Pharmaceutical Education 2003; 67 (4) Article 108. retical relationships between service constructs.13 For instance, knowing that service quality is made up of 5 dimensions can permit researchers to explore the relationships between each dimension and constructs such as satisfaction and loyalty. This is more challenging to accomplish if service quality is described at the individual attribute level. those in the SERVPERF service quality measure developed by Cronin and Taylor were used.11 The original 41 items in the ESQ instrument consisted of items that assessed student perceptions about learning resources in the college, faculty, school administration (ie, deans, department chairs, and office personnel), and perceptions of their educational progress. Four items from the original instrument relating to perceptions of educational progress were dropped from further analysis (see Analysis section for details). The final 37 items in the instrument are listed in Table 1. Another advantage of understanding the dimensionality of service quality is that it provides more actionable results. If a series of service quality questions can be distilled into a smaller number of uncorrelated dimensions, then service improvement strategies that target specific dimensions of service (eg, reliability or responsiveness) can be developed. Rather than consider individual item scores, researchers can look at summed scores for items within a dimension. Overall Satisfaction. A 7-item series of questions relating to overall satisfaction with various aspects of the educational experience were also asked of the students. A Chronbach’s alpha coefficient calculation of 0.91 showed the 7-item instrument to be internally reliable. The 7 satisfaction items are included in Table 2. When the dimensionality of educational service quality is discussed, researchers have assumed the 5 dimensional model proposed by Parasuraman et al.4,5 However, no known research has been conducted to date which has empirically identified the dimensions of either college or professional education. Administration of the Instrument. The ESQ survey was administered to P-4 pharmacy students before graduation at a university in the Southeastern United States. Four successive classes of P-4 students (graduating classes of 1999 to 2002) were surveyed during their on-campus day in the spring semester. Three hundred twenty-five students completed the survey out of a total of 372 P-4 students enrolled at the time of administration, for an 87% completion rate. This study examines the dimensionality of educational service quality building upon previous research by Holdford and Reinders.3 That research provided evidence of the reliability and validity of the instrument, but was limited to responses from a single fourth year pharmacy (P-4) class. Therefore, the sample was too small to permit multivariate statistical analysis of the data. Hypotheses. A number of hypotheses were proposed relating to expected relationships between service quality and satisfaction. The first (H1) was based upon the assumption that each dimension of educational service quality should have a significant relationship with overall satisfaction. H1: Resources, faculty member behavior, faculty member expertise, faculty member communication, and administration dimensions each are significantly associated with the 7item overall satisfaction measure. This research contains a sufficient sample size to permit analyses of the dimensionality of student responses. Holdford and Reiders proposed a priori a 3dimensional model of service quality consisting of student perceptions of school facilities, faculty member services, and administration services.3 In this research, an analysis was conducted to examine whether the original 3-factor model or some other model might be appropriate. This paper attempts to achieve the following objectives: The next hypothesis consisted of a series of subhypotheses that were used to provide evidence of the validity of student responses. The assumption was that if ESQ is a valid measure, then its dimensions should correlate in a predictable manner with individual satisfaction items (H2). Therefore, we hypothesized that resources should predict the single school facility’s satisfaction question better than any other dimension (H2a). At least one of the faculty member dimensions of behavior, expertise, and communication should predict student satisfaction with the items assessing teaching quality, faculty members, and perceptions of intellectual development better than facilities or administration (H2b, H2c, and H2d). It was not clear which faculty member dimension would be the best predictor, but it 1. Describe the perceptual dimensions of student assessments of the quality of their education. 2. Link those dimensions with student satisfaction with their educational experience. METHODS Measures Educational Service Quality (ESQ). An educational service quality instrument from Holdford and Reinders was used.3 Performance measures similar to 3 American Journal of Pharmaceutical Education 2003; 67 (4) Article 108. Table 1. Five Factor Model of Educational Service Quality (n=320) Percent Factor Variance Chronbach's alpha Loading Explained Mean (SD) 8.2 0.76 1.09 (0.72) 0.62 1.32 (0.81) 0.66 Factor 4 - Facilities 1. Up-to-date teaching tools & equipment. 2. Physical facilities visually appealing & comfortable 3. Physical facilities convenient to students. 1.66 (0.93) 0.58 4. Electronic access to drug & health science information. 0.92 (0.77) 0.67 5. Computer laboratory an important asset. 6. Physical facilities readily available for use around-theclock. 0.62 (0.80) 0.77 (0.79) 0.61 0.56 Factor 2: Interpersonal Behavior of Faculty 7. Friendly & approachable. 1.01 (0.67) 0.71 8. Willing to help you. 0.97 (0.65) 0.74 9. Available outside of class. 1.18 (0.68) 0.74 10. Keep their promises. 1.23 (0.75) 0.55 11. Behavior instills confidence in students. 1.35 (0.82) 0.65 12. Sensitive to student confidentiality. 1.05 (0.74) 0.54 13. Honest with you. 1.10 (0.71) 0.62 14. Treat you with respect. 1.25 (0.77) 0.62 Factor 5: Faculty Expertise 15. Have the knowledge to answer your questions. 16. Are current with the developments in their area of expertise. 17. Know what topics are relevant to becoming a good pharmacist. 0.79 (0.59) 0.64 0.75 (0.64) 0.73 1.20 (0.75) 0.49 Factor 3: Faculty Communication 18. Consistent with their grading practices and what they tell you. 1.41 (0.84) 0.73 19. Explain things in a way that you can understand. 1.26 (0.67) 0.65 20. Have your best interests at heart. 1.37 (0.74) 0.64 21. Attempt to understand my specific needs. 1.41 (0.75) 0.75 22. Make clear what they expect of you. 1.26 (0.64) 0.69 23. Usually give me adequate feedback about my performance. 1.34 (0.77) 0.55 4 13.7 0.91 5.9 0.81 11 0.88 American Journal of Pharmaceutical Education 2003; 67 (4) Article 108. Table 1. Five Factor Model of Educational Service Quality (n=320) (Continued) Mean (SD) Percent Variance Chronbach's Factor alpha Loading Explained Factor 1: Administration 24. Show sincere interest in solving student problems. 1.16 (0.83) 0.77 25. Friendly & approachable. 1.10 (0.86) 0.83 26. Dependable. 1.12 (0.83) 0.81 27. Attempt to understand my specific needs. 1.20 (0.86) 0.8 28. Act promptly. 1.19 (0.80) 0.79 29. Willing to help you. 1.25 (0.87) 0.85 30. Honest with you. 1.19 (0.86) 0.81 31. Behavior instills confidence in students. 1.12 (0.85) 0.81 .95 (0.73) 0.72 33. Sensitive to student safety. 1.02 (0.76) 0.62 34. Keep students informed about issues that concern them. 1.09 (0.77) 0.71 35. Treat you with respect. 1.01 (0.79) 0.8 36. Have knowledge to answer your questions. 1.09 (0.81) 0.64 37. Responsive to student evaluations about the curriculum. 1.29 (0.89) 0.54 32. Sensitive to student confidentiality. Total Model 25 0.96 63.8 0.96 Note: Scales are 0 = strongly agree, 1 = agree, 2 = somewhat agree, 3 = disagree, 4 = strongly disagree. Five surveys could not be used in this analysis due to missing responses. was expected that at least one faculty member dimension would be most important. Finally, the administration dimension was hypothesized to predict student satisfaction with school administration better than with any other dimension (H2e). predict satisfaction with faculty members better than the dimensions school resources or administration. H2d: At least one of the faculty member service quality dimensions of faculty member behavior, expertise, or communication should predict satisfaction with intellectual development better than the dimensions school resources or administration. H2e: Administration should predict satisfaction with administration better than any other dimension of educational service quality. H2: Some dimensions of educational service quality should predictably correlate more highly with some individual satisfaction items than others: H2a: Resources should predict satisfaction with school facilities better than any other dimension of educational service quality. H2b: At least one of the faculty member service quality dimensions of faculty member behavior, expertise, or communication should predict satisfaction with the quality of teaching better than the dimensions school resources or administration. H2c: At least one of the faculty member service quality dimensions of faculty member behavior, expertise, or communication should RESULTS Data Analysis Descriptive Analysis. Demographic data were not collected on individual respondents because of concerns about confidentiality. In addition, demographic variables of gender, age, college education, degree, and grade point average did not have significant impact on 5 American Journal of Pharmaceutical Education 2003; 67 (4) Article 108. Table 2. Descriptive Statistics of Elements of Satisfaction with Pharmaceutical Education Question I am satisfied with the school’s facilities. Mean (SD) 1.16 (0.88) I am satisfied that the school provided me a high quality education. 0.84 (0.74) I am satisfied with the quality of teaching. 1.07 (0.83) I am satisfied with the extent of my intellectual development since enrolling in this school. 0.98 (0.77) I am satisfied with the faculty of this school. 1.06 (0.81) I am satisfied with the administration of this school. 1.01 (0.78) I am satisfied with the school’s curriculum. 1.13 (0.84) Items combined 1.09 (0.62) Note: Scales are 0 = strongly agree, 1 = agree, 2 = somewhat agree, 3 = disagree, 4 = strongly disagree. Chronbach’s alpha for the 7 items in the satisfaction scale = 0.91. (n=320) Table 3. Demographic Description of Student Population Surveyed (N =372) Variable N (%) Gender Male 101 (27%) Female 271 (73%) Mean age (y) 28 Race Caucasian 241 (65%) Asian American 71 (19%) African American 45 (12%) Other (mixed, Hispanic, or no response) 15 (4%) Married 59 (16%) With children 26 (7%) Education prior to pharmacy school 2 years 37 (10%) 3 or more years without a degree 112 (30%) 4 year degree or more 223 (60%) Note: The above statistics describe the student population who were eligible to take the test. Only 325 students completed the questionnaire for an 87% completion rate for eligible students. student evaluations of service quality.3 However, it was possible to provide some description of the surveyed student population based upon the overall demographics of each P-4 class. These descriptions are provided in Table 3. Mean scores and standard deviations for student responses to the service quality instrument are shown in Table 1. The scale varied from 0 (strongly agree) to 4 (strongly disagree) meaning that smaller the number, 6 American Journal of Pharmaceutical Education 2003; 67 (4) Article 108. explained 11% of variance and had a Chronbach’s alpha of 0.88. It was distinct from factor 2, interpersonal faculty member behaviors, because it addressed issues of communication only, while factor 2 examined a broader range of faculty member behaviors. the better the assessment. Mean scores ranged from 0.62 to 1.66 with student responses extending from 0 to 4 for most questions. All analyses of the collected data were conducted using SAS for Windows, Release 8.02. Exploratory Factor Analysis of the ESQ Factor 4 consisted of 6 items and was labeled resources. The resources section included questions relating to the facilities of the school and explained 8.2% of variance. The Chronbach’s alpha reliability coefficient was 0.76 and consisted of all items originally designed to assess school resources. This provides further evidence that students responded to the instrument as expected. A principal components exploratory factor analysis with varimax rotation was performed on the original 41 items in the ESQ.13 Principal components factor analysis is a statistical technique that transforms data from one set of variables into a smaller set of uncorrelated factors. An orthogonal varimax rotation was conducted because it maximizes the amount of variance described by a factor and minimizes the correlation between factors.14 Factor 5 was labeled faculty expertise. It consisted of questions of faculty member knowledge and expertise in their areas of teaching. Faculty member expertise explained 5.9% of variance and had a Chronbach’s alpha of 0.81. It was distinct from the other faculty member factors, 2 and 3, because they addressed faculty member behaviors distinct from expertise. Factor analysis of the 41 items revealed a 5-factor structure that explained 63.8% of total variance. The criteria for retaining the 5 factors were eigenvalues greater than one and the ability to describe and label each factor. Individual items in the ESQ were retained for further analysis if they had factor loadings greater than 0.45 and fell into 1 of the 5 interpretable factors. Four items relating to student perceptions of intellectual development were dropped from the ESQ because they did not load onto an easy to identify factor. To assess the reliability of responses, Chronbach’s alpha coefficient was calculated for the variables within each factor solution. All factors had coefficients greater than or equal to 0.70, indicating evidence of reliability.15 The final 37 items in the ESQ and the 5 factors, their loadings, and their Chronbach’s alphas are listed in Table 1. Mean scores and standard deviations for the 5 dimensions over 4 years are shown in Figure 1. Mean scores show a slow but steady improvement in student perceptions of faculty members’ interpersonal behavior, administration, and faculty member expertise. Resources showed improvement after the first year and no further improvement in later years. Faculty member communication scores remained consistent over the 4year assessment period. All dimensions except faculty member communication demonstrated statistically significant improvements in later years when compared to first year scores. Factor 1 was labeled administration and was composed of 14 items. It explained 25% of the variance in student responses to ESQ, and the 14 items had a total Chronbach’s alpha of 0.96. Items in this factor consisted of all of the questions relating to the quality of services provided by school administration. The fact that the 14 items were originally meant to assess school administration gives some evidence that students responded as expected. Hypothesis Testing Stepwise regression analyses were conducted to test the hypotheses proposed in the methods section. The results of the analyses are presented in Table 4. All 5 dimensions of service quality had significant relationships with overall satisfaction with the educational experience providing evidence to support H1. The analysis showed the extent to which the following service quality dimensions predicted overall satisfaction: faculty member interpersonal behavior (47%), administration (8%), resources (4%), faculty member communication (3%), and faculty member expertise (1%). Factor 2 consisted of 8 items and was labeled interpersonal behavior of faculty. It explained 13.7% of variance and had a Chronbach’s alpha of 0.91. The factor was labeled interpersonal behavior of faculty because it contained items that described behaviors associated with development of interpersonal relationships between students and faculty members (eg, willing to help, treat you with respect). Analyses of service quality dimensions on individual satisfaction items provided evidence to support all of the proposed hypotheses of H2. The resources service quality dimension predicted satisfaction with facilities (H2a) better than any other dimension of educational service quality. Faculty member behavior was most predictive of satisfaction with quality of teaching Factor 3 was labeled faculty communication and comprised 6 items that described the quality of communication between faculty members and students. It 7 American Journal of Pharmaceutical Education 2003; 67 (4) Article 108. Trend analysis of ESQ dimensions 1.6 Mean Likert Scores 1.4 1.2 1 1999 Mean 2000 Mean 0.8 2001 Mean 0.6 2002 Mean 0.4 0.2 0 Admin Behavior Communic Resource Fac. Expertise Figure 1. Means and Standard Deviations of ESQ Dimensions by Year of Administration. (H2b) and satisfaction with faculty members (H2d), while faculty member expertise best predicted perceptions of intellectual development (H2c). Finally, Factor 1, administration predicted the administration satisfaction item best of all dimensions providing support for H2e. ing of dimensions called resources, faculty member, and administration (Figure 2). The result for resources was similar to that of the tangibles dimension of Parasuraman et al, while faculty member and administration dimensions consisted of assessments of the reliability, responsiveness, assurance, and empathy of faculty members and administration services. No hypotheses were developed for relationships between service quality dimensions and overall satisfaction with the quality of education or the curriculum because there were no clear reasons to suggest one dimension over any other. Factor 1, administration, predicted student satisfaction with the quality of their education, while faculty member behavior best predicted student satisfaction with the school’s curriculum. Factor analysis of the data showed a more complex dimensional structure of educational service quality than the proposed 3-dimensional model and a different 5-factor model than that of Parasuraman et al.17 The resources and administration dimensions were found to be unidimensional, but perceptions of faculty members were multidimensional consisting of interpersonal behavior, expertise, and communication. DISCUSSION Examinations of the relative importance of each service quality dimension on satisfaction found that faculty member interpersonal behaviors were most important in explaining overall student satisfaction with their education as measured by the 7-item satisfaction instrument. Attributes associated with faculty member interpersonal behavior included perceptions of faculty member friendliness and approachability, willingness to help, availability, honesty, ability to instill confi- Educational service quality consists of 5 dimensions, which are all significantly associated with satisfaction. Although other researchers have proposed a 5dimensional structure for service quality, the dimensions in this study are unique. Parasuraman et al proposed the 5 service quality dimensions of tangibles, reliability, responsiveness, assurance, and empathy (Figure 2).16 The service quality instrument in this study proposed a priori a 3-dimensional model consist- 8 American Journal of Pharmaceutical Education 2003; 67 (4) Article 108. Table 4. Stepwise Regression of Service Quality Dimensions with Satisfaction Satisfaction Items I am satisfied with the school’’s facilities. I am satisfied that the school provided me a high quality education. I am satisfied with the quality of teaching. I am satisfied with the extent of my intellectual development since enrolling in this school. I am satisfied with the faculty of this school. I am satisfied with the administration of this school. I am satisfied with the School's curriculum. Items combined Faculty Faculty Faculty Resources Behavior Expertise Communication Administration (R2) (R2) (R2) (R2) (R2) 0.12 0.01 0.28 0.04 0.08 0.01 0.28 0.02 0.02 0.39 0.07 0.27 0.03 0.01 0.01 0.06 0.45 0.02 0.01 0.03 0.02 0.56 0.02 0.29 0.04 0.01 0.07 0.04 0.47 0.01 0.03 0.08 Values shown are partial R-Square values from regressing service quality dimensions with individual satisfaction items. Figures in bold indicate the dimension which explains the greatest portion of variance in responses toward the satisfaction item in that row. Only significant relationships between service quality dimensions and satisfaction items are shown. Service Dimensions In Literature A Priori Dimensions Study Dimensions Tangibles Resources Resources Reliability Faculty Behavior Faculty Responsiveness Faculty Expertise Assurance Faculty Communication Empathy Administration Administration Figure 2. Dimensions of Educational Service Quality. satisfy or dissatisfy service customers. It is important that faculty members pay particular attention to how they deal with these opportunities if they wish to develop good professional relationships with the students and enhance overall satisfaction with the school. dence, and respectful behavior. These attributes deal with the one-to-one interactions between students that determine faculty member/student relationships. In services marketing, these interactions are called “moments of truth” where an opportunity exists to either 9 American Journal of Pharmaceutical Education 2003; 67 (4) Article 108. It is also not known whether this set of dimensions of service quality is exhaustive. Although the model explained 63% of student satisfaction with their education, there still may be other dimensions that have been missed. For example, some researchers have argued for a dimension that assesses the outcome of services.18 Educational service quality deals primarily with the processes of education (eg, the manner in which it is provided) rather than the outcomes (eg, skills developed, extent of learning). Another important dimension of education, curriculum, was not addressed in the ESQ. Although some dimensions indirectly address issues associated with curriculum, the curriculum itself was not directly assessed in the ESQ. School administration was the next most important determinant of student responses to the multi-item satisfaction instrument. School administration assessed student perceptions of administration reliability, responsiveness, expertise, and other aspects of service. Students receive extensive exposure to administration in their fourth and final year of professional school because administration coordinates clerkships, oncampus experiences, and communications between the school and students when they are off campus. This gives the administration multiple opportunities to experience “moments of truth” with students and influence student satisfaction. Faculty member communication followed school administration in importance. Communication attributes deal with how faculty members set expectations of students and respond to student efforts in a fair manner. Faculty member communication is also important in student satisfaction because it is essential in preventing conflicts and establishing trust. Many student complaints about faculty members result from student perceptions that they have been treated unfairly.17 These conflicts often result because faculty members and students differ about what is expected of students. When expectations and consequences are not consistently communicated to students through syllabi, policies, grading, words, and actions, conflict is a likely result. CONCLUSIONS This study demonstrated that the ESQ scale is a valid and useful tool for assessing the quality of educational services. Student responses to the scale can be used to identify sources of satisfaction and dissatisfaction with educational experiences. This research found 5 dimensions of educational service quality. Students thought that all 5 dimensions were important in determining satisfaction with their education, but that faculty members’ interpersonal behavior was most important. The remaining dimensions in the order of importance that the students attributed to them were administration, resources, faculty member communication, and faculty member expertise. Faculty member expertise was the least important determinant of student satisfaction with their educational experience. Faculty member expertise addresses knowledge, currency of knowledge, and relevance of what is taught. Although it may seem surprising that students put least emphasis on faculty member expertise, this result may be situation specific. In this study, students appeared to have fewer problems with faculty member expertise than with faculty member behavior and communication. Students gave 2 of the 3 faculty member expertise items the highest ratings of all items in the instrument. In a different setting or situation, expertise could be a greater source of conflict and therefore impact student satisfaction more. Assessments of the final P-4 year are particularly important for pharmacy schools to conduct because these perceptions are the ones that students will take with them after graduation. Although these perceptions will probably change over time as students become pharmacists and professional experiences reshape perceptions of the value of educational experiences, the attitudes developed in the final year are likely to be long lasting. REFERENCES 1. Allen J, David D. Searching for excellence in marketing education: the relationship between service quality and three outcome variables. J Marketing Educ. 1991; 13:47-54. 2. DiDominico E, Bonnici J. Assessing service quality within the educational environment. Education. 1996;116:353-60. 3. Holdford DA, Reinders TP. Development of an Instrument to assess student perceptions of the quality of pharmaceutical education. Am J Pharm Educ. 2001; 65:125-31. 4. Anderson E. High tech vs. high touch: a case study of TQM implementation in higher education. Manage Serv Qual. 1995; 48-56. 5. Pariseau S, McDaniel J. Assessing service quality in schools of business. Internat J Qual Reliable Manage. 1997;14:204-18. Limitations This study exclusively examined responses by fourth year pharmacy students in a single geographic location. The ability to draw similar conclusions to other students in other years of education, types of education, or geographic locations is restricted. In addition, evaluations of educational quality by other important stakeholders such as educators, employers, and the public are not assessed. 10 American Journal of Pharmaceutical Education 2003; 67 (4) Article 108. 6. Ford JB, Joseph M, Joseph B. Importance-performance analysis as a strategic tool for service marketers: The case of service quality perceptions of business students in New Zealand and the USA. J Serv Marketing. 1990; 9: 171-86. 7. Parasuraman A, Zeithaml V, Berry L. A conceptual model of service quality and its implications for future research. J Marketing. 1985; 49:41-50. 8. Parasuraman A, Zeithaml V, Berry L. SERVQUAL: A multi item scale for measuring consumer perception of service quality. J Retail. 1988; 64:12-40. 9. Mels G, Boshoff C, Nel D. The dimensions of service quality: the original European perspective revisited. Serv Industry J. 1997; 17:173-89. 10. Carman JM. Consumer perceptions of service quality: an assessment of SERVQUAL dimensions. J Retail. 1990; 66:33-55. 11. Cronin JJ, Taylor SA. Measuring service quality: a reexamination and extension. J Marketing. 1992; 56:55-68. 12. Grapentine T. Managing multicollinearity. Marketing Res. 1997; 9:10-21. 13. Grapentine T. Dimensions of an attribute. Marketing Res. 1995; 7:19-27. 14. Hatcher L. A Step-By-Step Approach to Using the SAS System for Factor Analysis and Structural Equation Modeling. Cary, North Carolina: SAS Institute Inc.; 1994. 15. Nunnally JC. Psychometric Theory. 2nd ed. New York, NY: McGraw-Hill; 1978. 16. Parasuraman A., Zeithaml VA, Berry LL. Refinement and reassessment of the SERVQUAL scale. J Retail. 1991; 67:420-50. 17. Holdford DA, Lovelace-Elmore B. Applying the principles of human motivation to pharmaceutical education. J Pharm Teach. 2001;8:18. 18. Powpaka S. The role of outcome quality as a determinant of overall service quality in different categories of services industries: an empirical investigation. J Serv Marketing. 1996; 10:5-25. 11