Validation of a Research Instrument to Measure Generation Y Parents’ Perception of Service Quality and Effect on Satisfaction and Word of Mouth in International Schools

is quantitative research aimed to (1) validate an adapted instrument used for measuring service quality in private hospitals to measure the perceived service quality of K-12 international schools and (2) implement the instrument to assess service quality’s impact on word of mouth and satisfaction. It was necessary to contextualize the scale items through item objective congruence test using industry experts. Scale items were adjusted to reect the service provided by teachers, sta, and leadership of the school.e resulting 27 scale items for service quality were shown to be contextually valid and internally reliable. e instrument was then implemented to measure parental service quality’s eect satisfaction and word of mouth.e survey was piloted by 33 parents and veried for internal consistency before being administered to 422 Generation Y parents. e results showed that the modied instrument was reliable and valid. e results showed that service quality had a direct and positive eect on both satisfaction and word of mouth but it had a greater eect on satisfaction.


Introduction
Schools are good at assessing the academic achievement of their students. Evaluating academic achievement is essential for schools. However, as a service industry, schools must also ensure that the quality of their service meets the expectations of parents and students. Schools will routinely monitor the satisfaction level of sta , parents, and students. Higher education institutes often measure the quality of the service they deliver by asking their students [1][2][3][4][5]. However, K-12 schools face the challenge that their students are not their direct customers; parents are the ones paying the tuition. As such, there is often a gap in measuring the service quality as perceived by their parents. Yet, satisfaction and perceived service quality are essential for service industries to be successful in an ever-increasing competitive market [6,7].
As a result, schools must seek to understand the needs of their parents and students allowing them to adapt and improve so that they can not only meet but also even exceed their expectations. erefore, there exists a need for K-12 schools to develop an instrument for measuring the perceived service quality of their parents.
is study sought to develop and validate an existing service quality instrument to make it contextually relevant and statistically reliable for use with parents of K-12 international schools.
Objectives for this research were, rst, to develop and validate the service quality instrument according to the speci c context of Generation Y parents in e International School of Macao, and second, to use a reliable modi ed instrument to measure Generation Y parents' perception of service quality and the e ect on parent satisfaction and word of mouth.

Research Question.
Can parent perception of service quality in K-12 international schools be reliably measured and what effect does service quality have on parent satisfaction and on parent word of mouth?

Research Objectives.
e objective of the current study was to develop a reliable instrument for measuring service quality by adapting and contextualizing one used in private healthcare. e second objective was to implement and validate the instrument by determining the effect that service quality has on parent satisfaction and parent word of mouth.

Service
Quality. Service quality is most often based on the expectancy disconfirmation theory [8][9][10]. A review of the literature suggests that two service quality models are typically accepted. Parasuraman [10] and colleagues proposed the SERVQUAL model for measuring service quality that consisted of 5 dimensions: tangibles, reliability, responsiveness, assurance, and empathy. is model has been used widely by many researchers [11][12][13][14][15] and specifically to measure service quality in higher education [2,[16][17][18]. Gronroos [9] developed a service quality model specifying the technical and functional aspects of service quality [9,19].
Research into service quality has been used in many industries but its use in education has been limited to higher education [16,20] and the researcher was unable to find any studies where it was used in K-12 contexts. Unfortunately, the use in HE is not suitable for K-12 contexts. In a university setting, the student receiving the service also decides whether it fits their expectations. In K-12 education, parents determine if the service meets their expectations though they may have only received some of the services directly and some of the services indirectly. As such, there is a need for a different measurement instrument that addresses this gap.

Expectation Disconfirmation eory. Service quality and satisfaction is a measure between what is expected and what
is experienced and is commonly referred to as expectation disconfirmation theory [21] and previously known as expectation confirmation theory [22,23]. Expectation disconfirmation theory can be applied to a product or a service. A person with low expectations may be satisfied with the same level of experience that a person with high expectations would not. As such, expectation disconfirmation theory informs perceived service quality. Customers are satisfied with a service experience if it meets or exceeds their expectations [8]. In the same way, parents will be satisfied with their service experience from the school if it meets or exceeds their expectations.

Satisfaction.
e satisfaction of both students and parents is often closely monitored by schools using both internal and external procedures. School-related nonacademic and academic aspects are frequently included in the multifaceted concept of satisfaction. Parents who are not happy will complain and spread negative messages, and if things do not change, they will withdraw their children from the school and look for another one.
In the research, satisfaction is informed by the expectancy disconfirmation theory [22][23][24] as the "consequence of the difference between the expected and perceived performance" [25]. Consumer satisfaction is a unique type of customer attitude used in the service sector that takes into account how much a customer likes or dislikes a service after using it [26].
Research is also interested in how satisfaction affects customer loyalty. Customer loyalty can be exhibited in repurchasing, continued use, or positive word of mouth. Jain et al. were not able to establish a link between satisfaction and WOM or brand loyalty. While satisfaction was not sufficient by itself in generating WOM, building customerbrand relationships on social media did have a significant and positive impact on brand trust, brand loyalty, and WOM for the brand [27].

Word of Mouth.
Word of mouth (WOM) occurs when a consumer expresses their individual experiences with the company to other consumers. Previous research has shown that WOM has a direct effect on a consumer's expectation and on a consumer's perceived benefit which leads to a decision to purchase. WOM is generated by the consumer after purchasing and thereby influences other potential customers. In this way, the consumer is the producer of the WOM [27]. WOM is often a behaviour associated with customer loyalty. Like customer loyalty, parent loyalty can be defined as parents who give positive word of mouth, recommends the school to others, and encourages others to use the school service [28].
A special kind of WOM includes the liking, commenting, and sharing of posts on social media [29]. e effect of WOM is directly related to the strength of the tie between the author and receiver [30]. Prospective parents put a lot of faith in the WOM of their friends who are already enrolled in the institution. and protecting their children while also encouraging their children to be involved in community service. Academic achievement of the children of Generation Y parents is increasing and parents are likely to have higher expectations of their children's teachers and schools. [31].

Research Framework and Hypotheses.
In addition to developing a reliable instrument to measure service quality, the current study developed the following research model to investigate the effect of service quality on satisfaction and word of mouth where service quality is considered an independent variable and satisfaction and word of mouth are considered dependent variables. e following hypotheses are proposed: H1. A reliable and contextually valid instrument can be developed to measure service quality of parents in international schools.
H2. Service quality has significant and positive impact on satisfaction.
H3. Service quality has significant and positive impact on parent word of mouth. e following conceptual framework provided a model to examine the relationships between service quality and satisfaction and service quality and word of mouth as seen in Figure 1.

e Contextualized Scale Items.
Previously developed for private healthcare by Lam [32] and operationalized by Cham [33], the 23 scale items were selected based on the reliability evidenced through its Cronbach Alpha value of 0.839 [33]. e researcher converted the scale items for the K-12 private international school context. e conversion was verified through an IOC test with 3 experts in the K-12 education sector. e experts were asked to determine the suitability of the construct to measure the given variable. Items that did not achieve a majority approval (>0.6) were revised based on the expert feedback and resubmitted during the second round. Results of the IOC are shown in Table 1. e 23 items of Cham (2016) were converted to 27 items as shown in Table 2 [33]. Industry experts specifically separated the identified subject in the original scale items from "staff" to "teachers" and "office and support staff" in the revised scale items to more closely reflect the different roles in a school.

Sample Size and Method.
Given the framework of 3 latent variables with 37 observed variables, an anticipated effect size 0.2, and a probability level of 0.05, it was calculated Mobile Information Systems that a minimum sample size of 296 would be required to determine effect. As such, a goal of 400 responses was set and exceeded. In this study, nonprobability sampling was used. e school granted access to 1937 parent emails and contact information. e researcher was also able to collect contact details of alumni's parents who would fit the Y population.
e researcher was able to use judgement sampling by emailing all parents whose children were currently enrolled in the school. Upon completing the questionnaire, respondents were asked to forward the questionnaire to another potential participant including alumni parents whose contacts were not available at the school. e use of snowball sampling further increased response success.
A pilot test using the modified scale items and a 5-point Likert scale was administered by emailing 99 current parents and resulted in 33 valid responses. e collected data was used to test the reliability of the modified scale items. e service quality construct achieved a Cronbach Alpha value of 0.956 which indicates that the items are internally consistent and reliably indicate the service quality construct. e resulting prototype instrument was then distributed to the parents at e International School of Macao by email and they were asked to complete the survey in English or Chinese. ere were 475 total responses by Generation Y parents. Of the 475 valid responses, 74.3% (353) were female, 24.0% (114) were male, and 1.7% (8) preferred not to state their gender.

Demographics.
Of the 475 responses, 422 were complete and valid and 74.9% (316) were female, 23.2% (98) were male, and 1.9% (8) preferred not to state their gender. Most respondents were married or living with a partner (90.8% (383)), 5.7% (24) were single or divorced, and 3.5% (15) preferred not to state their marital status. e following descriptive statistics and assessment of normality are presented in Table 3. While the values for skewness and kurtosis between -2 and +2 are considered acceptable to prove normal univariate distribution [34], only SAT4 exceeds the -2, +2 limit. However, Kline suggests that when using a large sample population procedure, such as SEM, one could reject the null hypothesis (of consistency with the normal distribution) and adopt a more descriptive approach to assessing normality. As such, the results are considered acceptable [35].

Confirmatory Factor Analysis.
To evaluate the convergent and discriminant validity of the constructs and to determine the model fit, confirmatory factor analysis (CFA) was utilized. To test convergent validity, the following four criteria should be met as suggested by Hair et al. [36]; namely, the construct reliability (Cronbach Alpha value) should be greater than 0.7; the explained variance (AVE) within each construct should be larger than 0.5; the standardized factor loading of each observed variable to the latent construct should be at least 0.60; and the composite reliability (CR) should be at least 0.70. As seen in Table 4, all four criteria were met. All the observed variables had a factor loading greater than the recommended minimum of 0.6. e Cronbach Alpha values are well above the minimum of 0.7.
e AVE values for all constructs exceed the minimum of 0.5 and the composite reliability (CR) values meet the minimum of 0.7. ese results demonstrate that all the constructs in this study achieved the acceptable level of convergent validity. Two observed variables from service quality, namely, SQ1 and SQ2, were removed to get model fit.
Discriminant validity of this study was assessed using the Heterotrait-Monotrait Ratio of Correlations (HTMT) method proposed by Hensler et al. [37]. e statistical variances between the constructs are measured using discriminant validity. By comparing the correlations of indicators across constructs to the correlations of indicators within a concept, discriminant validity may be evaluated. e model has discriminant validity if the correlation ratio is below 0.9 [37]. e HTMT plugin by Gaskin was used in AMOS to determine the HTMT values [38]. As seen in Table 5, the correlation ratios between all variables are below 0.9; thus discriminant validity for this study was achieved. Since convergent and discriminant validity were proved, construct validity was established. Based on the suggestion by Hair et al. [36] the following criteria were used to determine model fit including chisquare fit statistics over degrees of freedom (CMIN/DF), goodness of fit index (GFI), adjusted goodness of fit index (AGFI), normed fit index (NFI), comparative fit index (CFI), Tucker-Lewis index (TLI), and root mean square error of approximation (RMSEA). A research model with a good fit meets the following criteria: CMIN/DF greater than 5; GFI greater than 0.85; AGFI, NFI, CFI, and TLI greater than 0.8; and REMSEA less than 0.08. As seen in Table 6, the current research model exceeded every criterion and achieved acceptable model fit.

Structural Equation Model
. A structural equation model (Figure 2) was created to determine the standardized path coefficients of the constructs, given construct validity and an acceptable model fit. e initial model required SAT and WOM to be covaried in order to achieve model fit as    Table 7. us, we can conclude that the model is valid [36]. As can be seen in Figure 2, the standardized direct effects reported by the AMOS software indicate that service quality had 0.76 factor on satisfaction and a 0.60 factor on word of mouth. e results are discussed in the following section.

Discussion
is study's primary goal was to develop a contextually reliable and valid measure instrument for assessing the perceived service quality of parents in international schools. e findings demonstrate that, through IOC and CFA, a reliable instrument can be developed. is study's secondary objective was to use the new instrument to assess the impact service quality had on satisfaction and word of mouth. e impact of these variables and the proposed hypotheses are discussed below and shown in Table 8.

Influence of Service Quality on Satisfaction.
e current study revealed that service quality had a significant and direct impact on parent satisfaction. As such, H2 was supported. Like satisfaction, service quality is a measure of the gap between expected results and actual results. When a parent receives lower than expected service, they perceive it as lower quality. Responding to a parent complaint with good service quality can lead to parent satisfaction. When a parent receives higher than expected service, they perceive it as higher quality. is expectancy confirmation or expectancy disconfirmation applies to satisfaction as well [8]. When a parent's needs are exceeded, they are more likely to be perceived as being satisfied.

Influence of Service Quality on Word of Mouth.
e results of the current study revealed that there is a significant and direct relationship between service quality and parent word of mouth. As such, H3 was supported. When a parent receives greater than expected service at the school, they are more likely to recommend the school to others. It is understandable that when more parents feel that they get higher levels of service at the school, their satisfaction of the school and the likelihood of them recommending the school are increased.

Conclusion
is study reports the development of a contextualized service quality instrument for K-12 parents from one sample of Generation Y parents.
e results demonstrate that modified scale items are reliable and may be used as an appropriate instrument for assessing parents' perception of service quality. e objectives were all answered satisfactorily. e adaptation of the service quality model used in healthcare required industry expert consultation for contextualization and measuring for internal consistency to ensure that the scale items would measure the intended variables. e use of the Item Objective Congruence provided a meaningful context of the scale items for a K-12 international school. Service quality scale items originally referring to staff were split to identify teachers and office and support staff. e development of reliable and contextualized scale items can be used in future research. As suggested by Chatfield and Collins [44] modified scale items can be used reliably as part of a measurement model and further used in the structural equation models (SEM) to identify causal relationships that use service quality as one of the constructs.
School administrators and leaders need a reliable method for measuring the quality of the service that staff were providing in their schools. School administrators and leaders can also benefit by understanding the role that the service that a parent experiences has a direct impact on their satisfaction and on the likelihood to recommend the school. By improving the experience associated with service, parents will have a greater perception of the school quality. e instrument can be used by schools to better evaluate the quality of the service that it provides to parents. Furthermore, the use of the new instrument can be used to survey all the parents in the school and does not need to be limited to a specific subpopulation such as Generation Y.
One limitation is that the developed instrument was only tried out in one international school and specifically limited to one specific population within the larger parent population. Future studies should expand on the population and be implemented within a greater range of international schools. Such development would yield further refining of the instrument.

Data Availability
e data used to support the findings of this study are available from the corresponding author.