Mediating effect of hardiness in the nexus of service quality and students’ satisfaction of public universities in Ghana

Abstract The education sector as a service industry, is increasingly recognising the importance of service quality, in their quest to heightened efforts to meet the ever-changing expectations of their students. The higher education landscape regards students as customers and therefore, gives them greater latitude in their demand for better academic programmes. This study assessed the impact of service quality perception on students’ satisfaction with educational services provided in public universities. The study further examined the mediating role of students’ academic hardiness in the relationship between service quality and students’ satisfaction with education services. A proportionate stratified sampling technique was used to select 1022 students from two public universities to respond to a survey questionnaire. The first research question was examined with descriptive statistics, specifically mean and standard deviation, percentages, and frequencies. However, the study hypotheses were all tested using a biased-corrected covariance-based structural equation modelling approach. The study revealed a high level of student satisfaction with education services provided in the public universities. The study also showed a significant positive impact of service quality perception on students’ satisfaction. Further, academic hardiness was found to have a positive indirect effect on students’ satisfaction through the quality of educational services provided by an institution. The finding that students’ satisfactions were affected by the quality of services provided by academic institutions implies that universities should strive to add to the quality of infrastructure and services to maintain students’ satisfaction, which has rippling effect on their learning and loyalty to the institution.


Abstract:
The education sector as a service industry, is increasingly recognising the importance of service quality, in their quest to heightened efforts to meet the everchanging expectations of their students.The higher education landscape regards students as customers and therefore, gives them greater latitude in their demand for better academic programmes.This study assessed the impact of service quality perception on students' satisfaction with educational services provided in public universities.The study further examined the mediating role of students' academic hardiness in the relationship between service quality and students' satisfaction with education services.A proportionate stratified sampling technique was used to select 1022 students from two public universities to respond to a survey questionnaire.The first research question was examined with descriptive statistics, specifically mean and standard deviation, percentages, and frequencies.However, the study hypotheses were all tested using a biased-corrected covariance-based structural equation modelling approach.The study revealed a high level of student satisfaction with education services provided in the public universities.The study also showed a significant positive impact of service quality perception on students' satisfaction.Further, academic hardiness was found to have a positive indirect effect on students' satisfaction through the quality of educational services provided by an institution.The finding that students' satisfactions were affected by the quality of services provided by academic institutions implies that universities should strive to add to the quality of infrastructure and services to maintain students' satisfaction, which has rippling effect on their learning and loyalty to the institution.

ABOUT THE AUTHORS
Isaac Amoako holds a doctor degree in Measurement and Evaluation, and is a tutor at the Atebubu College of Education.Currently, he is the head of Education Studies Unit of Atebubu College of Education.He has a research interest in issues around high stakes testing and quality assurance in higher education institutions.Eric Anane is an associate professor in Measurement and Evaluation.He is currently the director of the University of Cape Coast's Institute of Education.He has a research interest in "Assessment in Schools, Psychometrics, Teacher Education, and Cognition."Andrews Cobbinah holds a doctor degree in Research, Measurement and Evaluation.He is a Senior Lecturer with the Department of Education and Psychology at the University of Cape Coast.He is also a board member of the School of Educational Development and Outreach at the University of Cape Coast.Dr. Cobbinah's research interests revolve around the equating of test scores, Assessment in Schools, and critical thinking and peer-assessment skills training.

Introduction
Quality is an essential concept in modern society (Mattah et al., 2018;Sierra, 2012;Valencia-Arias et al., 2023), given that we can achieve optimal use of resources through the creation of quality evaluation.Education forms an essential part of society and obtaining a quality education is the pillar on which sustainable development is based (United Nations UN, 2015).This is so, given that the future of society's citizens will play out in accordance with the educational development they undergo.For this reason, the optimisation, improvement, and development of mechanisms that guarantee educational quality, and especially the quality of higher education, will assist future citizens who find themselves in a transitional period with regards to training for the world of work.Generally, in the broader service sector, service quality has been associated with profitability and a competitive advantage (Santhi & Ganesh, 2015).Though "service quality" is an abstract concept (Zietsman et al., 2019), Eresia-Eke et al. (2018) aver that in modern-day organisations, the quality of a service offering, and its consistent delivery could be what distinguishes a struggling organisation from a thriving one.This is meaningful because studies generally tend to suggest that good service quality typically correlates with higher levels of customer satisfaction (Yusoff et al., 2015).
Three methods of service quality measurement are frequently used in the literature on service quality.The evaluation of ongoing service comes first, followed by consideration of teachers' and students' perspectives on quality service.The first focus is on service quality.This concept has received a great deal of attention from a research perspective in the educational sector, especially in developed jurisdictions (Ham, 2003;Vergara et al., 2011).Lago et al. (2013) stated that quality is the outcome of comparing expectations of a service with the actual service received.Essentially, service quality is generally seen as the ability of the service organisation to meet or exceed customer expectations (Zeithaml & Bitner, 2003;Zeithaml et al., 2006).The emphasis on customer satisfaction has been rooted in their evaluation of services, according to the expectancydisconfirmation model (Oliver, 1977).That is, the organization's outcome (i.e., service or product) is seen as good compared to the anticipated expectations of the customer.This means that the measurement of quality lies in the satisfaction of individuals and in the value that they believe they receive.Authors such as Carvalho and De Oliveira (2010), Ledden et al. (2011), andSimpson (2012) explain that educational model dynamics place the student as the recipient of education and, as a result, convert them into the client receiving the service, with the second and third focus (i.e., student and teacher) being encompassed by the first (service).Mihanovi´c et al. (2016) observed that academic institutions are beginning to realise the significance of student satisfaction.Sokoli et al. (2018) define student satisfaction as an emotional state resulting from a comparison of their experience of services delivered and their expectations.Student satisfaction with education services is important because it influences their decision to continue with their studies at the institution.Indeed, levels of student satisfaction may also influence current students' propensity to recommend the institution to other potential students.Conversely, student dissatisfaction leads to negative word of mouth that would tarnish the reputation of the institution (Mansori et al., 2014).In essence, students' actions, or lack thereof, in this regard could affect student retention and loyalty, which are considered by Santhi and Ganesh (2015) as important indicators of service quality provision in an academic institution.
Service quality is a multi-dimensional phenomenon that is influenced by several factors.Within educational institutions, factors that could affect the service quality perceptions of students may include teaching ability, location and facilities of the institution, administrative issues, classroom aesthetics and practicability, relationships between lecturers and students, and so on (Mason et al., 2018).Ltifi (2013) observed that there is a lack of consensus among studies with respect to variables that substantially affect service quality, even though several studies have been conducted on service quality.This is so because the nature of the service quality construct is that it is subjective and contextual (Eresia-Eka, 2020;Izogo, 2017).The lack of agreement, which may be attributed to the subjectivity and context specificity of the quality construct, undermines the generalizability potential of service quality studies.In effect, the risk of drawing inaccurate conclusions is real if findings from previous studies conducted in different contexts are relied upon to infer what the precise situation may be within the context of Ghana.
The behavioural manifestations of the three hardiness attitudes (i.e., commitment, control, and challenge) can be extended to the service quality of higher education environment conversation.The moderating effect of commitment on academic performance may be demonstrated by students becoming deeply involved in their studies, seeing this as the best way to turn whatever they are experiencing into something that seems interesting, worthwhile, and important (Maddi, 2006).Such an attitude is likely to facilitate industriousness and a willingness to expend extra time and effort to meet academic goals.Students with high levels of control ought to be able to manage their studies, for example, by demonstrating good time management, prioritising those activities deemed most contributory to academic success, and taking responsibility for their own learning and development (Maddi, 2006).Challenge-related attitudes should moderate academic performance by allowing students to view potentially stressful situations as exciting and stimulating rather than threatening (Maddi et al., 2002).This ought to increase the likelihood of students accepting the difficulties associated with fulfilling academic course requirements and engaging in the process of working towards a degree, thus facilitating the positive process of growth through learning (Maddi, 2006).
Existing research has found that hardiness has some dimensional mediating influence on variables other than student satisfaction.For example, the commitment dimension is reported to have a mediating effect on first aid affect and self-efficacy (Yu & Liang, 2021), as well as on interpersonal teacher support and individual support variables (Lin et al., 2020).The challenge dimension of hardiness has been revealed to have an indirect effect on academic stress through a sense of belonging (Abdollahi et al., 2020).In addition, studies have shown the positive influence of academic hardiness on undergraduates' coping with university first-year stress (Maddi et al., 1996), student adjustment to university life (Mathis & Lecce, 1999), and student retention (Maddi et al., 2002), which may in part suggest students' satisfaction.Strong levels of hardiness attitudes among students enhance performance criteria such as creativity, wisdom, and fulfilment while also maintaining or enhancing physical and mental health (Maddi, 2006).However, there is little evidence to date regarding the mediating effect of hardiness in the relationship between service quality and students' satisfaction with service delivery.
Examining critically the literature, a considerable number of studies have investigated students' satisfaction (Amoako & Asamoah-Gyimah, 2020); service quality (Arënliu et al., 2021;Asante, 2015;Kwarteng, 2020;Mattah et al., 2018;Seniwoliba, 2014;Zakari, 2016); and hardiness (Dodoo & Surienty, 2021) within the Ghanaian context.These studies presented evidence of an appreciable level of students' satisfaction, service quality, and hardiness, respectively.However, the unique interaction of the three variables in terms of how hardiness indirectly affects students' satisfaction through service quality appears to be sparse.Specifically, the present study examined: (1) students' satisfaction levels with educational services; (2) the impact of service quality on students' satisfaction; and (3) the mediating effect of academic hardiness on the connection between students'satisfaction and service quality.
The present study would be valuable to stakeholders in education in several ways.For instance, feedback on students' satisfaction with service quality may serve as a prompt for the management of universities in Ghana to put in place quality mechanisms to meet the varied needs of students and global academic standards.Moreover, information on the indirect effect of hardiness on service quality perception and satisfaction may inform policy directives towards cognitivebehavioural enhancement programmes.This programme is widely believed to have the potency to improve hardiness among students, which translates into quality academic engagement for better preparation of human resources towards nation-building.

Research approach
This investigation employed a quantitative approach through a descriptive design.This approach was deemed necessary to obtain understanding and knowledge via statistical lenses (Creswell, 2012).Per the specific objectives of the study, specifically exploring the satisfaction level of students with education services, the impact of perceived education service quality on satisfaction of students, and the mediation role of academic hardiness in the relationship between perceived service quality and satisfaction, the quantitative research approach was the most befitting (Tabachnick, & Fidell, 2013).To sum up, this research approach became an option because the problem that necessitated the conduct of this study required the findings to be generalised from the sample of students to the population of students (Cohen et al., 2017).Students are the primary beneficiaries of education services, and thus, sampling students for a study of this nature was in order.Extant empirical studies in an attempt to investigate issues of this nature found the use of quantitative approach and the descriptive design relevant (Arenliu, et al., 2021;Hoque et al., 2021) which informed the choice of this approach for the current study.

Participants' selection
This empirical study included students from two public universities in Ghana (the University of Cape Coast (UCC) and the University of Education, Winneba (UEW) that were founded with the intention of training teachers and, in fact, run most education programmes.Even though in recent times in Ghana a lot more public universities have run education programmes, they are incomparable to the UCC and UEW in terms of the number of education programmes and the number of students in such programmes.Thus, regular education students at all academic levels in the "College of Education Studies" of UCC and the "School of Creative Art and Faculty of Ghanaian Language Education" of UEW became the focus of this investigation.Because all education students at both universities receive similar treatment and services, their characteristics were deemed equal based on their experiences with the services provided.On the foregoing premise, proportionate stratified random sampling (with the lottery method) was used to select 1022 participants from the targeted institutions to participate in this study.The Institutional Review Board (IRB) of the UCC approved the survey procedure after all ethical standards were scrutinised and duly met.In addition, study participants signed consent forms to register their willingness to be part of the study.

Student satisfaction scale
The Student Satisfaction Scale (SSS) is a unidimensional scale developed by Brady et al. (2002).The scale was developed to measure the overall satisfaction of students with the educational experience that they have received in a particular school context.There are six items on this scale.All the six items were positively worded with factor loads ranging from .5 to .7.Some of the items were: "I am pleased to be enrolled as a student at my university because there are enough facilities that support students' learning," "My choice to enroll in my university was a wise one because they provide all that I require," "The education services like library services, security services, accommodation services, etc., of my university are most essential for current tertiary education," and so on.
The scale is a five-point Likert kind of scale, ranging from 1 = "strongly disagree" to 5 = "strongly agree."Internal consistency reliability, as measured by the Cronbach Alpha index, was.92.The subscale was adapted for the current study by rewording all the items for simplicity of reading and meaning to yield more consistent scores (Coleman, 2013).

Higher education service quality scale
The higher education service quality scale (HESQUAL) developed by Teeroovengadum et al. (2016) was adapted and used for measuring educational service quality within the selected universities.The original scale was made up of two dimensions with 15 items on a Five-Point-Likert scale (that is, very good, good, credit, poor, and very poor).The first dimension of the scale was "Functional Quality," which assessed university students' perceptions of the quality of educational services provided by their institution.This subscale had the largest number of items (i.e., 9) of the entire HESQUAL scale.Some of the items that participants were to rate as very good, good, credit, poor, or very poor were "Attitudes and behaviour of administrative staff," "Administrative processes," "General Infrastructure," "Attitude and behaviour of most lecturers," "Adequacy of support facilities like libraries, laboratories, ICT laboratories," etc.The internal consistency of the subscale measured by Cronbach's Alpha was.90.
The second dimension of the HESQUAL scale was named "Transformational Quality," which reflected the students' belief about how their university had helped change their academic lives to get some relevant skills.Some of the items were: "My university has helped me to acquire selfconfidence," "my university has helped me to acquire critical thinking skills," "my university has helped me to acquire problem-solving skills in my field of study," etc.This sub-dimension had six items that were all positively framed.The internal consistency index of the sub-dimension was.87.The HESQUAL scale was adapted and used in this present study.In the adaptation process, the items were reworded with simpler words, but the meanings of the original items were maintained in the adapted items.

Academic hardiness scale
The Academic Hardiness Scale (AHS) was adapted for this study.The ultimate goal of the AHS was to measure the students' ability to withstand stress in the academic environment without permanent damage, either emotional, psychological, or physiological.The AHS has three (3) hypothetical sub-dimensions with 38 items in the original scale developed by Benishek et al. (2005).AHS is a four-point Likert scale, ranging from Strongly Agreed (SA) to Strongly Disagreed (SD).The first dimension was dubbed "commitment," which measured the willpower of students to remain involved in events or situations around them in an academic setting.Some of the items on the commitment dimension were "I take my academic work as a student seriously," "I work hard for good grades," "I am involved in all my class learning activities," etc.This sub-dimension had 8items with a Cronbach Alpha reliability index of.82.
The second sub-dimension was also named "Control," which measured the students' personality style of having the ability to manage relevant life issues through the application of knowledge, imagination, and choices.Some of the items on the control sub-dimension were: "often times, if I perform poorly, I doubt my abilities as a student, and I do not ascribe it to external forces," "I often find it difficult to bounce back to normal from academic disappointment, so I dwell on experience to act," and "I often become less motivated to study when I do not get the right grades, so I do not procrastinate learning."This sub-dimension was made up of 3-items with a Cronbach Alpha reliability index of.80.Finally, the third sub-dimension, which was "Challenge," measured the students' inner power to see stress as a normal part of academic life and, as a result, take advantage of it to learn.The subdimension contained three items and had a Cronbach alpha coefficient of.77.Some of the items on the challenge dimension are: "I often avoid courses that require extra work from a student," "I do not find it needful to pursue a course if I am not confident, I will do well in that course," and "I enroll in classes for courses in which I can do well."Six items were negatively worded, so their scores were reversed.The scale was adapted for the purpose of this current work.In the adaptation process, all negatively worded items were changed to positively worded ones.In this study, 14 items, which reflected all the items on the original scale, were used.
Finally, the three scales, that is, SSS, HESQUAL and AHS were put together as a single questionnaire and piloted among 150 education students of Akenten Appiah Minka University of Skills Training and Enterpreneural Development (AAMUSTED), Mampong campus.The results of the pilot test helped in testing feasibility of the data collection approach, the checking of the possible errors of the items on the questionnaire (Leedy & Ormrod, 2005).

Procedure
After scrutiny and subsequent approval from the Institutional Review Board (IRB) of the University of Cape Coast, with reference number CES-ERB/UCC-edu/v5/21-87, data collection began at the two public universities chosen.Some administrative officers and faculty members were helpful in sensitising students to the exercise during this exercise.The data collection exercise was done using a face-to-face approach, where copies of the questionnaire were distributed to the chosen students in their lecture halls.During the administration of the questionnaire to the students, such terms as consent, confidentiality, anonymity, volition, freedom of withdrawal, and protection from emotional harm or discomfort were strictly adhered to.For instance, students were asked to sign the consent form as a sign of their willingness to be part of the study.Students were also prompted not to write their name or anything that could identify them on the questionnaire.Furthermore, students were told that they had the right to withdraw from the study at any time, without giving any reason.The data collection exercise was done in 2021 and lasted for eight weeks.

Data analysis
After the data collection, out of the 1022 questionnaires distributed, 1019 (representing 99.7% response rate) were filled as expected and were deemed fit to be used.Again, the entire set of data was inspected for appropriateness and correctness through screening and cleaning to check for outlines.Mean score was computed on respondents' responses to the items of the survey instrument, the frequency counts, and percentages of the various levels of satisfaction were computed and presented for the first objective.A covariance-based structural equation modelling (CB-SEM) approach was performed to assess the education service quality impact on satisfaction and the mediating effect of academic hardiness on the relationship between service quality and satisfaction for the second and third objectives, respectively.Prior to the analysis of the data, traditional linear regression analysis was preferred; however, the inefficiencies of traditional regression, such as the inability to estimate measurement error (Civelek, 2018), as well as the procedure's inability to estimate indirect effect relationships (as in the case of objective 3 of this study) (Civelek, 2018;DeVellis, 2017), made the use of CB-SEM the obvious option.Once again, CB-SEM's robustness in terms of adhering to strict parametric protocols paved the way for its adoption over variance-based structural equation modelling (also called PLS-SEM) (Civelek, 2018).

Preliminary analysis
To satisfy the measurement model requirement of the CB-SEM approach used in this study, the fitness, convergence and discriminant validity of the data were inspected.Tables 1, Tables 2, 3,  Tables 4, 5 and Table 6 show the various indices achieved for the above-named measures.1, is evidence of the data meeting acceptable fitness criteria across the various indices of absolute, comparative and parsimonious fitness.The various indices were within the standard benchmark of .90 or better (Byrne, 2011;Hair et al., 2010).On an account of reliability and validity of the scales, CFA results in Tables 2 to 6, adequately exhibit factors loads indicators for the various dimensions pegged at .5 to .7 threshold (Civelek, 2018).Factors loads below .5 were discarded prior to data collection.The average variance extracted (AVE) indicators for the multidimensional scales were middle way through to convergence, the internal consistency reliability ranged from .5 to .80 which are deemed appropriate (Pallant, 2011).Once again, for the multidimensional scales, that is, HESQUAL and AHS, Tables 4 and 6 showed evidence of moderate   Questionnaire items, SQ1-SQ15, (see Appendix A); *Items to be discarded; Overall alpha = .88discriminant validity.Clearly, the presence of moderate discriminant, convergence and good interconstruct correlation suggest an appreciable level of construct validity for purposes of estimation of the structural model (Fornell & Larcker, 1981).

Results in Table
Table 7 provides descriptive details of the variables of the study.
The descriptive statistics of the data showed that the hardiness dimensions had mean values ranging from 2.8 to 3.4, with the students exhibiting a higher trait of commitment and a lower trait of control.The mean score for service quality was 4.0, with a standard deviation of .86.The mean satisfaction of students was 3.9, with a standard deviation of 1.1.All the skewness and kurtosis values for the variables were within acceptable ranges.

Undergraduate students' satisfaction level of educational services
This research objective examined the students' satisfaction level with the educational services delivered by their universities.The mean of the means for each student was computed, and mean scores between 0 and 5 were obtained.The values between 0 and 2.9 were categorised as low satisfaction, mean scores greater than 2.9 but less than 3.5 were classified as moderate, and mean scores greater than 3.5 were deemed as high satisfaction.The cut-offs for mean scores used in this study were those proposed by Brady et al. (2002) for summative interpretation.The details of the responses are shown in Table 2.
The results, as shown in Table 8, show that a larger proportion of the students reported a high level of satisfaction for education services (n = 513, 50.34%).A relatively smaller proportion of students reported moderate (n = 295, 28.95%) and low satisfaction (n = 211, 20.70%), respectively.This implies that most of the students were satisfied with the education services that they experienced at their university.

Service quality impact on students' satisfaction of educational services
The research objective sought to explore the impact of the perceived service quality provided by the university on students' overall satisfaction.A summary of the results is provided in Table 9 and Figure 1.
Figure 1 shows that service quality predicts students' satisfaction with educational services by 0.29, with an error variance of 67.07 for service quality and 15.30 for the students' satisfaction.Table 9 also shows that service quality is a significant predictor of students' satisfaction with educational services [B =.29, Boot 95% CI (.451570)].The finding (B = .29)suggests that service quality explains positively students' satisfaction with educational services.The implication is that better (i.e., higher) services rendered in an academic institution may lead to an increase in student satisfaction within the academic institution.In addition, the data indicates that service quality explains 26% of the variance in the students' satisfaction with educational services.

Mediating effect of academic hardiness on the connection between the students' satisfaction and service quality
This objective investigated the mediating role of academic hardiness in the relationship between service quality and satisfaction among students.A summary of the results is provided in Table 4 and Figure 2.
Table 10 and Figure 2 show that, aside from the control dimension of hardiness, B = .000,Boot 95% CI (−.001001), commitment, B = .025,Boot 95% CI (.014038), and challenge, B = .010,Boot 95% CI (.004019), significantly mediate service quality and the students' satisfaction with educational services.This implies that commitment and challenge dimensions of hardiness have a positive indirect effect on the relationship between service quality and students' satisfaction.Path analysis of hardiness in the relationship of service quality and satisfaction.

Discussion
The study investigated the satisfaction level of undergraduate education students' participants.
Findings revealed that most of the students were satisfied with the education services that they experienced at their various campuses.This assertion is accepted because of the higher proportion of students found in the category of high satisfaction (refer to Table 2).The finding further implies that students perceive the education services that they have experienced so far to be of high quality.Students often get satisfied when the university's services meet their expectations (Lago et al., 2013;Yusoff et al., 2015;Zeithaml & Bitner, 2003).For an academic institution, the students' satisfaction with educational services tells how efficient and effective that institution is.This may mean that the academic institutions are able to meet the varied needs of their students.The foregoing assertion aligns with the expectancy-disconfirmation theory by Oliver (1977), which posits that customer satisfaction or dissatisfaction comes from his or her comparison of performance with predetermined standards.With certainty, student satisfaction will increase when academic resources (both electronic and non-electronic), institutional infrastructure, and other administrative services meet the expectations of students.
As part of the objectives, the study also explored the relationship between service quality and students' satisfaction with education services provision.Results from the study showed that service quality had a positive impact on student satisfaction.This implies that students were more satisfied because they judged the services provided by the universities to be of high quality.The finding further implies that, at any point that an academic institution increases the quality level of their services, the students get highly satisfied, which would subsequently affect their retention and commitment to studies (Maddi, 2006).The study's findings that service quality predicts student satisfaction are consistent with previous research (Amoako & Asamoah-Gyimah, 2020;Mattah et al., 2018), which discovered a similar trend.As previous studies (Amoako & Asamoah-Gyimah, 2020;Mattah et al., 2018) had a limited focus on services quality (i.e., measured only infrastructure and technological quality), the current study looked at service quality from a broader perspective (i.e., capturing all what those institutions used in the study provides).Students' satisfaction is therefore a function of the quality of services provided by an academic institution.
Finally, on the issue of the mediating effect of academic hardiness in the relationship between service quality and satisfaction, the study discovered that, apart from the control subdimensions of hardiness, commitment, and challenge, all three fully mediated the relationship between service quality and the students' satisfaction with educational services.This implies that, if students do not act on what they believe (commitment) to be creative and innovative (challenge) toward the problem at hand, knowledge of control over situations may not amount to anything.This is to say that hardy students get satisfied easily when the service provided is of high quality.Even though the current study's aim was not the same as Yu and Liang's (2021) study, the results of both studies were similar in the sense that both studies found commitment, a dimension of hardiness, to be a significant mediator variable.The finding also corroborated other findings in the literature on the same issue of the mediating effect of the commitment and challenge dimensions of academic hardiness (Lin et al., 2020;Abdollahi, Panahipour, Akhavan-Tafti, & Allen, 2020).

Limitations
As only two public universities and a sample from the "Bachelor of Education programme" were used, the generalisation of this is restricted.Therefore, the generalisation of the findings to a larger student population and other academic programmes should be done with caution.As in the case of most survey studies, study participants are likely to either exaggerate or overreport on the study outcome variable (i.e., satisfaction) due to inherent social desirability concerns often associated with self-reported measures.

Practical implications
The finding that students' satisfactions were affected by the quality of services provided by academic institutions implies that universities should strive to expand and improve the quality of infrastructure and academic services to maintain students' satisfaction, which has implications for their learning and loyalty to the institution.Again, the finding that hardiness has a positive effect on service quality and satisfaction implies that management of universities should deepen academic counselling initiatives and build capacity toward the improvement of students' hardiness.This can be done through the implementation of hardiness training programmes (e.g., cognitive-behavioural therapy).The therapy focuses on cognitive reconstruction, which involves identifying and disputing irrational or automatic negative thoughts.Hardy students are able to learn new skills; they see effort as a major prerequisite for success; and they are less likely to worry about failure because mistakes are perceived as part of learning.
Higher educational institutions (especially, the study centers) interested in sustaining appropriate standards and students' satisfaction must place emphasis in improving hardiness among students.This is the sole idea which appears to be missing in the service quality literature and on which this study has relevance.

Conclusions
The study has revealed that majority of the students were satisfied with the educational services of the universities under consideration.The finding implies that students consider the educational services (i.e., teaching services, library services, security services, accommodation services, etc.) provided by the universities to be of high quality.Feedback from students (who are the primary customers of an education system) is a valuable source of information that directs management and other stakeholders toward policies about quality assurance.The study also discovered that the quality of service provided by universities influenced students' satisfaction.Clearly, quality academic service provision overlaps with satisfaction in a manner that creates a conducive academic environment for effective teaching and learning.As strong personality characteristics, commitment and challenge were identified as key influencers of students' satisfaction with educational services provision.Hardy students regard activities as engaging and enjoyable (i.e., requiring commitment) as well as significant learning stimuli (challenge).Since severe academic stress can also compromise a quality academic experience, the propensity to discover positive purpose in academic life at the university is a distinguishing trait of hardiness.For purposes of widening the scope of knowledge around the issue of student satisfaction with education services, future researchers are encouraged to examine how age or gender moderate satisfaction and service quality perception among students.

Figure 1 .
Figure 1.Service quality predicts satisfaction.Path analysis of service quality predicting satisfaction of education service.

Figure
Figure 2. Mediating effect of hardiness on the relationship between service quality and satisfaction.

Table 6 . Discriminant validity of AHS Dimensions Commitment Control Challenge
*Values in the parenthesis are square root of AVEs.