Exploring educational service quality and improvement priorities from the viewpoints of health sciences students using the SERVQUAL model

In higher education institutes, there are competitive and tussling environments to provide students with a high educational and service quality. Service quality is the key to meet or exceed student expectations. The quality of health systems, therefore; would be directly dependent on the quality of medical sciences education. In the current study, the modied SERVQUAL model was applied to assess the quality of the educational service provided at King Abdulaziz University (KAU) in Saudi Arabia. This is a descriptive, cross-sectional study that was conducted within the Faculties of Medicine, Applied Medical Sciences, Dentistry, and Pharmacy. Through applying the modied SERVQUAL model, the gap between students' perceptions and expectations was calculated and correlated with other demographic data. The ratio of students’ perception to their expectation was calculated as a probe of student satisfaction in the current educational services to enable prioritizing items for improvement. University between the between The perception/expectation ratio was calculated as a of and be and for and expectation level of signicance was set at p < 0.05. in Saudi Arabia by using a modied version of the SERVQUAL model. The ndings of this study showed that all the dimensions were not up to the expectations of students. Improvements are needed across all ve SERVQUAL dimensions. P/E Ratio and service improvement matrix can be used as a guide to determine which item/dimension should be prioritized for improvement. Gender and faculty factors tend to have a signicant impact on some dimensions. The results are deemed to help the university to develop action plans to address all weaknesses. It is strongly recommended to replicate this study in other universities and in various programs. This might result in developing guidelines for the educational system improvement at the national level. In addition, future work venues may include:


Introduction
In higher education institutes, there are highly competitive and tussling environments to provide a highly valued educational and service quality [1]. Recently, higher education is dealing with high pressure to improve their quality and service, and to focus on stakeholder interests, and to improve student satisfaction [2]. The evaluation of higher education quality and service is essential to enhance educational value, to meet student satisfaction, and to give feedback on the effectiveness of educational plans and applications [3]. Service quality is the key to meet or exceed student expectations [4]. Many studies have been developed to measure student and stakeholder satisfaction [5]. Several tools and models have been developed to measure the quality services [6][7][8]. The SERVQUAL is a multidimensional model that has been widely used to measure quality services in many contexts, including service industries such as hospitality [9], banking [10], hospitals [11], mobile industry [12], and tourism [13]. In education, the model has been applied to business schools [14], and higher education institutions [4,[15][16][17].
The SERVQUAL quality management framework, which is developed by a group of American researchers [18] in 1985 is the most common method to measure quality in the service sector. The SERVQUAL is an instrument that has 10 dimensions of service quality, which were presented together with a model of service quality. They are accessibility, reliability, responsiveness, competence, courtesy, communication, credibility, security, understanding/knowing the customer, and tangibles [19]. In education, modi ed SERVQUAL instruments have been validated and used [20][21][22]. It has also been used and formed to LibQUAL, which is used to measure academic library service quality [23].
Literature studies [24,25] in higher education institutions revealed that students' perceived service quality is a precursor to student satisfaction. A high perception of service quality develops a positive image in the minds of students which then leads to satisfaction as well as to attract more students through word of mouth [26,27]. Students' withdrawal from higher education institutions was primarily attributed to not meet their expectation [28]. Nowadays, educational institutions exert huge efforts to meet customers' expectations, similar to what business organizations do, but some of them still lack customer awareness among the staff, and it has become a common drawback for many institutions [29,30].
The service quality gap is based on the difference between student's perceived services and their expectation as pointed out by Parasuraman [18], but the gap value does not provide a tool to prioritize which dimension or subdimension must be placed rst for improvement. Prioritizing the service improvement is of high importance for educational management for e cient empowerment of human resources and nancial capital. Recently, service improvement matrix has been used by several authors [1,31] to prioritize which dimension in educational services should be improved rst. Although there are some studies [32][33][34][35] dealing with quality services and student satisfaction in higher education institutions in the Kingdom of Saudi Arabia (KSA), an in-depth analysis of quality services with management education perspective is still needed. Herein, we focus on health science students' perception and satisfaction as health-related programs are in high demand by students in KSA compared to other programs. In this study, we employed a modi ed SERVQAUL model that consists of 27 items under ve dimensions to a) assess the educational services in the ve dimensions (reliability, empathy, assurance, responsiveness, tangibles), b) conduct quality service gap analysis on dimension and subdimension level, and c) set a guide to prioritize service improvement. This study also discussed the in uence of gender, college, and GPA factors on the service quality gap and student's perception to expectation ratio as an indicator of improvement priority. The gap helps identifying the areas requires improvement and ratio prioritizes improvements on the quality dimension and subdimension levels. The ndings will help the management develop action plans to improve the quality of institutions and its services, thereby achieving higher student satisfaction.

Methods
The study was conducted at the Faculty of Medicine (FOM), Faculty of Applied Medical Sciences (FAMS), Faculty of Dentistry (FOD), and Faculty of Pharmacy (FOP) of King Abdulaziz University in Saudi Arabia. The study design was descriptive, cross-sectional, survey based. The survey facilitated identi cation the students' perception and expectations of the educational environment at KAU.
The sample size was calculated by applying the following formula to estimate the number of the sample: Where n: sample size, z: standard variable corresponding to 95% con dence level, P: anticipated population proportion, and d: statistical precision. Using 95% con dence, d = 6% and P = 50% (as no prior information is available on population proportion). The estimated sample size is 266.
Study participants were handed hard copies of the survey. The SERVQUAL survey was originally adopted from Parasuraman et al [19] and modi ed and validated by Kebriaei and Akbari [36] and used by many studies [35][36][37].
A ve-point Likert scale was used to measure the perception and expectation of perceived quality in ve generic dimensions with 27 items: assurance (5 items), responsiveness (5 items), empathy (6 items), and reliability (7 items). The difference between students' perception and expectations was calculated to identify the gap by using the equation: The gap between these two opinions was correlated with other demographic data (age, gender, and current GPA).
For perception, students were asked to rate each item of current services on a scale from 1 to 5, where: 5 = very good, 4 = good, 3 = moderate, 2 = poor, and 1 = very poor. For expectation, students were asked to rate how important each item is to the quality of the service on a scale from 1 to 5, where: 5 = very important, 4 = important, 3 = moderate, 2 = less important, and 1 = least important. Most important is equal to the highest expectation and least important is equal to lowest expectation. Each item of SERVQUAL tool was scored from 1 to 5, where: 1 = very poor/least important and 5 = very good/very important.
The ratio of students' perception to expectation was calculated to prioritize the improvement for dimension and as a probe of students' satisfaction in the educational services. The item or dimension with lowest P/E requires the highest priority for improvement. The ratio of P/E is de ned as follows: When the mean score of perception is less than mean score of expectation, then P/E ratio is less than 1 implying that the perceived service does not ful l students' expectation and they see it unsatisfactory, and it requires improvement. When the P/E ratio = 1.00, it means that the mean score of perception equals expectation and the perceived service ful ls students' expectation, and they see it satisfactory and needs to be maintained to sustain its growth. When the value of P/E is higher than 1, that means score of the service provided is highly satisfactory to students. As the value of P/E ratio approaches 1.00, students become more satis ed with the service and improvement priority deceases and vice versa. Thus, the P/E ratio can be used to set out a quick guide for improvement priorities and can be used as a probe to evaluate students' satisfaction with the service.
The service improvement matrix was also developed as described in Chui et al study [1] to prioritize items for improvement. Brie y, a 2D-plot was generated by plotting the mean scores of perception (indicating satisfaction) versus mean scores of expectation (indicating importance) for each item (or dimension). The graph is divided into four quadrants based on overall mean scores of perception and expectation for items. The quadrant with items having mean scores of perception (satisfaction) lower than the overall mean score of perception (27 items) and mean scores of expectation (importance) higher than the overall mean score of expectation require prioritize improvement.
The data were analyzed using IBM SPSS 24.0 software (IBM Inc.). Descriptive statistics for demographic data were calculated, which include frequencies and percentages. Inferential statistics using nonparametric tests of Wilcoxon test for paired samples, Kruskal Wallis test, and Spearman correlation coe cients were conducted. The difference between students' perceptions and expectations was calculated to identify the gap between them. The perception/expectation ratio was calculated as a probe of students' satisfaction. The reliability of the survey instrument was determined by using Cronbach's alpha value and found to be 0.896 and 0.927 for perception items and expectation items, respectively. The level of signi cance was set at p < 0.05.  The difference between students' perception and expectation and the ratio P/E were calculated for all items and presented in Table 2. Furthermore, the overall gap and ratio of each domain were also calculated. There was a significant difference (p < 0.00) between perception and expectation for all items except for item 24 (Professional appearance of faculty members and staff). Negative gap was observed for all items except for item 24. The P/E ratio for all items was equal to or greater than 0.67. Item 24 scored the highest P/E ratio 1.20, with no significant difference between perception and expectation mean values. Figure 1 shows a direct linear relationship between P/E ratio and service quality gap (P -E) with high correlation coefficient (R 2 = 0.9412). As the value of the service quality gap (P -E) in educational services gets wider (more negative) the value of P/E becomes progressively less than 1. Comparing the score of overall quality gap calculated among the five quality dimensions using Wilcoxon test showed a significance difference. Table 3 summarized the results of Spearman correlation analysis between the components of service quality. Positive and significant correlations were revealed between the components. Reliability was strongly correlated with assurance and responsiveness. The significant difference between male and female students based on SERVQUAL dimensions was tested. It revealed that there was a significant difference in assurance and responsiveness based on gender (p < 0.05) as shown in Table 4. The P/E ratio showed significant difference in assurance, responsiveness, and empathy based on gender, indicating the difference between the gap and ratio values. Kruskal-Wallis test was used to test whether the faculty has a significant difference with the responses of SERVQUAL dimensions or not. Table 5 showed the significant differences between students' perceptions and expectations in assurance, responsiveness, reliability, and tangibles dimensions (p < 0.05). On the other hand, empathy has no significant difference (p > 0.05).
Further analysis to compare the faculties based on P/E ratio was performed and results indicated a significant difference (p < 0.05) for the five dimensions except for empathy. Kruskal-Wallis test was used to test whether the GPA has a significant difference with the responses of SERVQUAL dimensions or not. Table 6 showed significant difference between students' perceptions and expectations only in reliability dimension. Interestingly, lowest GPA range scored the largest gap and lowest student P/E ratio. Improvement service matrix graph was developed and presented in Figure 2. Vertical and horizontal lines represent the mean scores of perception and expectation for 27 items.

Discussion
The results presented in Table 2 indicate that the mean scores of students' perception and expectation were signi cantly different for all items of services (p < 0.001) except for item 24 (Professional appearance of faculty members and staff). In all items, the quality gap values between students' perception and expectation are negative, indicating that students do not perceive that the actual services provided by the university meet their expectation except for item 24. The overall gap values for the ve dimensions exhibit the smallest and largest gap values for "tangibles" and "reliability" dimensions, respectively. The results highlight that there are substantial areas for improvement and growth in educational services. The reliability dimension re ects the consistency of institution or program in terms of performance, as indicated in a similar study by Legčević [38]. This nding is consistent with the result reported by other studies [5,31,39].
The P/E ratio for reliability was the smallest among other dimensions, which prioritizes it as the rst dimension to be tackled by decision makers to eliminate or reduce the gap value and, additionally, positively affect the quality gaps in other dimensions [30]. Conversely, other studies have shown that other quality dimensions had the greatest quality gap. Aldarmahi et al [35] employed the SERVQUAL model to assess the quality gap between student expectations and perceptions of the service quality in King Saud bin Abdulaziz University for Health Sciences in Saudi Arabia and found negative overall gap values for the ve dimensions with the largest gap for responsiveness, reliability, empathy, assurance, and tangibles. A study [40] in a private university in Thailand investigated service quality using SERVQUAL model and reported a negative gap between the students' perceptions and expectations in the 5 dimensions employed. The author recommended that higher educational institution in Thailand to upgrade facilities and equipment to reduce the gap and improve service delivery.
Ten out of the 27 items had gap in quality services less than − 1, which requires action plans to minimize or eliminate the gap between student perception and expectation. In reliability dimension, the largest gap was − 1.60 in item 20 (Give higher scores if students attempt more), indicating that most of the students feel that their extra efforts and attempts do not necessarily lead to a better grade or score. Items 18 and 19 also identi ed areas for improvement in presenting the content material comprehensibly and informing students about their exam results, assignments, and homework in the assigned or promised timetable. Low student perception (satisfaction) in item 23 (Ful lling responsibilities by faculty members and staff in the promised time) suggests that faculty and staff need additional training and attending workshops and seminars to meet students' expectation. In assurance dimension, item 2 (Qualifying students for future job) had a gap of -1.36 indicating that students do not feel that the university quali es them for future jobs, which requires further improvement in the curriculum to include what students can do in the future rather than focusing on mere teaching of the content. This requires the university to set out clear learning outcomes from the program level to lecture level and share them with students.
In responsiveness dimension, two items (7 and 9) had gaps less than − 1 about the curriculum and easy accessibility of administrators to receive students' suggestion and opinions. Students do not feel that their views and suggestions are taken into consideration when structuring or updating the curriculum. This also provides a room to improve the way of collecting students' feedback on the contents and teaching, learning, and assessment strategies of courses, and how this feedback is evaluated and become part of the curriculum development cycle.
In empathy dimension, two items (12 and 13) had gaps less than − 1 about exibility of faculty members to understand students' situation under certain conditions and the convenience of class hours. The allotted time for classes was not satisfactory where the students perceived a low mean score of 2.82 (low satisfaction) while they were expecting high score of 4.45 (high importance).
The tangibles dimension has the smallest gap on quality services presented to student based on their opinions. Only one item (Up to date material and educational equipment) had a gap less than − 1, indicating the need to update the material.
The results showed a narrower service gap in educational services as reported by female students compared to male students in the ve dimensions and a statistically signi cant difference was revealed in two dimensions: assurance and responsiveness. The perception-to-expectation ratio for female students was higher than male students in all dimensions and a signi cant difference was shown in three dimensions: assurance, responsiveness, and empathy. The results indicate higher satisfaction of female students on the service quality provided. Different results were reported in the literature regarding the impact of gender on service quality and students' satisfaction. Ansary et al [41] had reported no signi cant effect of gender on service quality measured for 250 students from different nationalities in private higher education institution in Malaysia. Ham and Hayduk [42] showed that there is no signi cant difference between male and females with perceived service quality even though the ndings do show that males are more satis ed compared to females. Ilias et al [43] reported that there are no signi cant differences between male and female students with both quality service and satisfaction.
Aldarmahi et al [35] had reported that there was a signi cant difference between male and female undergraduate students in their quality gap values in the empathy, tangibles and responsiveness dimensions in a study conducted in King Saud bin Abdulaziz University for Health Sciences in Saudi Arabia. The greatest gap (dissatisfaction) was reported for female students in responsiveness and for male students in empathy and tangibles. Soutar and McNeil [44] found that there is a signi cant relationship between gender and satisfaction with service quality as males are more satis ed than females. Our nding agrees with those reported by Osman et al [45] that there was a signi cant difference in satisfaction between male and female students with females had a higher level of satisfaction compared to males. More research efforts are necessary to understand the differences among students' experiences to improve services offered to them, which consequently lead to student satisfaction.
In comparison between the students of the four colleges, the lowest P/E ratio and gap were reported in assurance dimension for FOM students, in responsiveness dimension for FAMS students, in empathy dimension for FOP students, and in reliability dimension for FOD students. Students at different colleges respond differently to the services provided by the university. Statistically signi cant differences were revealed among the students of the four colleges in assurance, responsiveness, tangibles, and reliability dimensions. Empathy was the only dimension that did not show any signi cant differences among the various colleges' students. Yusoff et al [46] studied the in uence of the program of study and other demographic factors on dimensions driving student satisfaction factors in Malaysia and found that commerce-tourism programs and international business programs are less satis ed compared to other programs within business schools.
Students with lowest GPAs showed greatest quality gaps (more negative) in all domains, which requires taking actions to reduce that gap with this group of students. However, no statistically signi cant differences were detected between students with different GPAs. Furthermore, students with lowest GPA exhibited lowest P/E ratio, re ecting their low satisfaction with the provided services. Our results are consistent with those reported by Ham and Hayduk [42] and Ilias et al [43].
The gap can identify the areas for improvement and P/E can probe the extent of the student's satisfaction and helps university administration to set priority improvement strategy not only for the whole dimension but also for each item in the survey. Items with the lowest P/E ratios (e.g., 13 and 20) should be given priority for improvement. In the same line, the developed improvement matrix graph (Fig. 2) indicated that the items in quadrant B perceive low satisfaction compared to its mean (3.32) and high importance compared to its mean (4.28), thus they should be given priority for improvement. Quadrant D items should be improved while item 24 must be maintained (perception > expectation). Quadrant A should be improved while its items present the best services in students' perception and importance. In a similar study, Chui et al [1] have developed service improvement matrix to prioritize the dimensions/items in SERVQUAL instrument for necessary and important improvement in private university in Malaysia. They found that the top priority should be directed to empathy dimension.

Limitations
The data collected in this study re ect the responses of health sciences students enrolled in four faculties during one semester. Including more faculties would have strengthened the results and improved the generalizability of the study ndings. Another limitation is that our study depended only on quantitative data. Adding a qualitative component (through focus groups or structured interviews) would have consolidated the ndings and added to the credibility of the study.

Conclusion And Recommendations
This study investigated gap and ratio of students' perceptions to expectations of educational service quality in four health sciences education faculties at King Abdulaziz University in Saudi Arabia by using a modi ed version of the SERVQUAL model. The ndings of this study showed that all the dimensions were not up to the expectations of students. Improvements are needed across all ve SERVQUAL dimensions. P/E Ratio and service improvement matrix can be used as a guide to determine which item/dimension should be prioritized for improvement. Gender and faculty factors tend to have a signi cant impact on some dimensions. The results are deemed to help the university to develop action plans to address all weaknesses. It is strongly recommended to replicate this study in other universities and in various programs. This might result in developing guidelines for the educational system improvement at the national level. In addition, future work venues may include: conducting service quality measurement at program, university, and national levels on a continuous basis to understand students' perception and expectation of current service and monitor them overtime; invoking institution self-study approach based on SERVQUAL periodically to minimize/eliminate the gaps and improves the services; and conducting comparative studies between private and public higher educational institutions for the quality service and student satisfaction to identify the differences and capitalize on the strength points of difference.  Figure 1 Correlation between service quality gap (P -E) and P/E ratio for 27 items. Linear line represents linear regression tting with equation P/E=0.2768Gap + 1.1021, R2 0.9412.