THE INDIRECT INFLUENCE OF MOUTH-TO-MOUTH MARKETING ON THE CHOICE CRITERION OF PUBLIC EDUCATION INSTITUTIONS

Several studies show a positive influence from word of mouth (WOM) marketing on the choice criterion of an Education Institution (EI). However, these researches did not investigate the indirect effect from WOM marketing on such choices and are also influenced by self-selection problems, which could result in biased estimators for the effect of WOM marketing on choice criterion for an EI. This study aims to verify if the suggestion by students’ Education Institution through WOM marketing influences indirectly their choice criterion. A theoretical model was to mitigate endogenous features related to these criteria and also minimize auto selection problems. The research – with 1220 answers - was applied in a public IE that promotes many education modalities. The results show there is no evidence that WOM marketing systematically has straight effect on students’ choice criterion by an IE. However, it is concerned that WOM marketing can enhance influence on some variables that interfere on students’ choice criterion by an IE . DOI: 10.18226/23190639.v7n2.04 Hildo Anselmo Galter Dalmonech, Fundacao Instituto Capixaba de Pesquisas em Contabilidade, Economia e Financas (FUCAPE). Brasil. E-mail: hildo.anselmo@ifes.edu.br Danilo Soares Monte-mor*, Fundacao Instituto Capixaba de Pesquisas em Contabilidade, Economia e Financas (FUCAPE). Brasil. E-mail: danilo@fucape.br Submetido: Junho 2018 Aceito: Abril 2019 *Contato para Correspondencia


Introduction
Companies that produce consumer goods have intensified their efforts mainly on service development and delivery purposing to maintain a competitive advantage (Jacob & Ulaga, 2008;Eggert, Hogreve, Ulaga, & Muenkhoff, 2014;Benedettini, Neely, & Swink, 2015). On the same way, educational institutions [EI] aim to progress and remain in the market through the efficiency of their processes, seriousness and high performance (Juliatto, 2013).
In educational market, enrollments are decreasing, fact that can be caused, in parts, by both increased competition and changes on age pyramid (Taneguti, 2013). This way, educational managers need to seek strategies to recruit students as a survival way, becoming the differential of success of an EI in a demanding and competitive scenario (Paiva, Costa, Barbosa, & Neto, 2014). In this context, WOM marketing appears as a differential for IE, becoming a force in which the personal influence from opinion leaders can demand for their products (Arndt, 1967;Santiago, Arruda, Dantas, & Oliveira, 2014).
Among the works that deal with the influence of WOM marketing on selection criteria are Arndt (1967), Bickart and Schindler (2002), Silva, Wassally, Silva and Santos (2008), Libai et al., (2010), Berger (2014), Santiago et al. (2014) among others. Overall, most of the results points to a positive influence of WOM marketing on choice criterion. However, from the theoretical point of view, such researches do not investigate the indirect effect of WOM marketing. These results provide insights to verify the ability of WOM marketing to intensify the effect of other variables on choice criterion. Furthermore, about econometric point of view, the used designers do not allow control by certain aspects, including unobservable characteristics related to WOM marketing which influence the choice criterion.
Such designers, therefore, are influenced by a self-selection problem, implying in biased estimators for WOM marketing effect on choice criterion by an EI. This work ambition is to verify if the indication of an EI by its students through WOM marketing indirectly influences their selection criteria. That is, if WOM marketing influences other variables, interfering in their choice criterion. For this, a model with different research design from the usual one in literature was proposed, counting with a control group (new students that were not exposed to that had no indication), and consequently mitigating endogenous aspects about EI choice criteria, minimizing self-selection problems.
This study provides knowledge sprawling about WOM marketing, systematizing the identified variables in literature that motivate the choice for an EI in order to eliminate the In practical terms, it is aimed that this research provides support for the EI managers be able to rely on the results for future decision-taking, with respect to projects and public policies related to the students' enrollment. In addition, the competition growth for vacancies and the quality improvement of selected students show potential ways to reduce school dropout (Sousa, 2014).

Influencing factors in the selection criteria for an EI
To understand what lead the student to choose an EI, it was taken as a basis the literature which explains the students' attraction, once on factors that have influence on their choice criteria. These factors are presented in Table 1: Petruzzellis and Romanazzi (2010), Walsh et al. (2018).

WOM marketing
The kind of marketing in which students on their own speak to others encouraging or not their purchase or use of products and services from an EI. Brooks (1957), Arndt (1967), Bickart and Schindler (2002), Silva et al. (2008) The studies on WOM marketing have been already widespread; however, some answers have not been clarified yet. The results so far point out that WOM marketing is frequent and important for consumer's behavior, what may influence other consumers to choose the product or service (Berger, 2014;Pereira et al., 2015;Voyer, & Ranaweera, 2015;Dalmonech et al., 2016;Herold et al., 2016;Wien, & Olsen, 2017;Le, 2018). In Brazil, for example, it can be verified that the behavior of WOM marketing action from students influences on the decision by an EI (Santiago et al., 2014), it also indicates services provided by a Public EI and the behavior of the WOM marketing is directly linked to students' satisfaction (Dalmonech et al., 2016). And there is a larger influence for a Particular EI due to publicity made by its students and ex-students than investments made with external marketing (Silva et al., 2008).
These results suggest that there is a relationship between WOM marketing and EI's selection criteria, but it is also known that WOM marketing, besides all disclosure and indication, can also inform and resolve doubts about some variables that influence the student's choice of an EI. Forward this scenario, the following hypotheses were formulated: H1a: The WOM marketing has a positive influence on students' choice for an EI.
The EIs need to look for ways to increase students' perceptions of quality to, this way, recommend the EI and continue their studies in the future, and even get on other levels of the EI; guaranteeing to the new students that they will also receive a quality service (Petruzzellis, & Romanazzi, 2010;Dlačić et al., 2014;Jalilvand et al., 2017). This can also Página | 80 lead to a higher number of recruited students, since quality is also an attribute of attracting students to an EI (Mainardes, 2007;Meneghelli, 2011).
The quality of an EI can be considered as a reflection of its infrastructure quality and employees' quality as well, and considering the first one with a more significant impact on student learning, becoming important for its development in such a way that it has been the main target of public policies that is increasingly committed to make them get to a better quality (Cavalcante, 2014;Hemsley-Brown, & Oplatka, 2015). Among the general factors influencing the criteria of choosing an EI, those related to personal reasons are listed as influential elements, and the choice is made through their own preferences (Yamamoto, 2006). One of the expectations of students before choosing an EI to study is its location, in order to choose a closer one to their home (Meneghelli, 2011). On the other hand, the research by Reis and Freitas (2014) shows that the factors related to personal reasons that attract students to EI such as proximity to home, influence of friends and family and safety on campus were not relevant in the student's decision-taking process.
People are increasingly looking for a quality EI and with a good reputation in the market, as a proof of this, there is the research of Perfeito et al. (2004) who discovered that the price of tuition presents little influence on the students' opinion, In the process of choosing an EI. However, students are increasingly charging from the EI for their return on investment (Petruzzellis, & Romanazzi, 2010;Walsh et al., 2015).
The student search for an EI that can insert him or her into the job market, by those EIs should not simply be knowledge providers; it is necessary to understand that the nowadays student due to the economic conditioning, is visibly worried about its insertion in the job market. (Valerio, & Pizzinatto, 2003;Mainardes, 2007;Petruzzellis, & Romanazzi, 2010;Walsh et al., 2018).
EIs should build partnerships with companies that are directly involved in the business field in which their courses are offered, ensuring that their students have greater employment opportunities. In other words, EIs should encourage career preparation programs Página | 81 (Petruzzellis, & Romanazzi, 2010;Walsh et al., 2018).
Generally, the future consumers search, by the simple observation of other consumers, identifying how that service is provided, thus defining it, in order to choose it or not (Moore et al., 2005). In this way they choose services by watching what happens in the environment. When observing the involvement of other people can then be influenced by the quality of the activities and align with their behavior, opting or not for the service. (Lehn, 2006).
In summary, the literature provide evidence on several factors that influence students' choice criterion by an EI. These factors, however, can be influenced by the perception that students build when subject to WOM. This expectation leads to the next hypothesis: H1b: The WOM marketing intensifies the influence of factors that have effect on students' choice criterion by an EI.

Methodologies
To verify the influence of indication of an EI by its students through WOM marketing on choice criterion by an EI, a descriptive and cross-sectional quantitative methodology was adopted (Marconi, & Lakatos, 2010). To provide this, it was chosen a primary data collection, through the application of a structured and self-fillable questionnaire, available printed and filled in person in the classroom on September and October of 2015, by incoming students in 2015.
Representing the choice criterion for an EI, it was defined as a study field students who joined in 2015 at Education, Science and Technology Federal Institute in Espírito Santo (IFES) that has 21 campuses and 01 rectory and "is a government agency from Federal Department linked to the Ministry of Education, which promotes education in different teaching modalities" (Lei nº 11.892, 2008). Its goal was to include as students from different contexts as possible to reach several teaching modalities in the same EI. In addition, it was considered the environment in which the EI was not directly geared towards the market with many dissemination formats and in various kinds of media to win the competition.
The research instrument was adapted from Mainardes' study (2007), with some adjustments to fit to the study target. Some of the statements from Mainardes' study (2007) were grouped and ordered with different nomenclatures. Another adaptation was to make it possible for the student to express what really led him / her to choose that EI. Finally, it was To mitigate the endogenous aspects of selection criterion, it was developed a model presented in Figure 1. This model had been be used to test the Hypotheses H1a and H1b. In this case, the questionary was applied to a group of students that were exposed to WOM and to another group that were not. At the first part of the questionnaire, there was a question: "Did I have any contact with another IFES student before deciding to study at IFES?" It differentiated two groups, treatment and control. The students coming from the treatment group are those who had contact with EI students before the choice, that is, they had an indication of EI students and this way it could be used to measure the effect of word of mouth marketing on the criterion of choice. And students from the control group are those who did not have contact with EI students before the choice, that is, they had no indication of EI students.
In the second part of the questionnaire, three affirmations were created for the variable criteria of choice, followed by a seven-level Likert scale (ranging from [1] strongly disagree and [7] strongly agree), thus, the variable final note was formed by the average per respondent.  However, these presented results have an endogenous relationship between the choice criterion and WOM marketing, and they may also imply in biased estimators to approximate their effect. Table 2 shows results of each variable, eliminating this endogenous relation. To make it possible, it was used the model II, in which the hypothesis H1b is also tested.
At  Source: Applied research data. ***, ** and * represent significant coefficients at 1%, 5% and 10% respectively. Treat: Dummy Variable = 1 the group that received indication from another student; 0 the group that received no indication from another student; At technical education modality, with 95% confidence, it can be inferred that in the control group there is a relation between the variables "EI Quality" (-0.512) and "Infrastructure" (0.579) with the student's choice criteriafor the EI, and they were used when the EI was chosen to study. This quality variable negative coefficient suggests the student becomes less judgmental when perceives the EI as high quality place. In this case, it is possible that EI meets the student's expectations, distinguishing it from the market by its set of properties, attributes and teaching conditions. In the same way, in the treatment group regarding to the quality of the EI, the result remained, rejecting the hypothesis H1b.
About the infrastructure, it can be noticed that the student thought to decide and became more judgmental about choices, showing that this variable influenced the choice criterion foran EI, what can be explained by the fact that students who look for this teaching modality already have a higher maturity degree, in which case turns infrastructure to an important goal to their professional qualification, affecting their choice criteria. In the treatment group regarding the infrastructure, it continues influencing the choice criteria, but with less intensity, what can be explained by the student whose WOM marketing influence encouraged studyingin a EI because of its infrastructure, so that the student heals doubts about the EI and also about this variable, showing that the effect of WOM marketing intensify the influence of the infrastructure on EI's selection criteria, not rejecting the hypothesis H1b.
In the higher education modality, in relation to the control group, it can be inferred, at 95% confidence that there is a relation between the variables "Consumer to consumer" (-0,456) and "Price" (0.293), and with 90% of confidence between "EI Quality" (0,429) and "Brand / Image" (0.417) with EI student selection criteria.
According to"Consumer to consumer" variable in the control group, the individual shows less judgmental. It is probably due to the behavior between people and environment of an EI, characteristics that makes student safe to choice. These results confirm the Lehn However, at the treatment group in relation to the variable consumer to consumer, the individual becomes more judicious in the choice process for an EI, beginning to reflect more on the subject due to the information contained in this dialogue with the students do not match their perception of the characteristics of the EI that were obtained by their observation. In this case not rejecting the hypothesis H1b, intensifying the effect of this variable. As for the variable "Price", this affects positively the choice criterion for an EI, what can be explained by being a public EI without payment of tuition, making the student think, reflect and be more judicious regarding this variable. One of the possible justifications is that, with the uns Table Brazilian economy, the student try to economize or simply the family income becomes insufficient to enter the private network. These results confirm the findings of Bergamo et al. (2010) that the perception of the price at the time of choosing the course is associated with the family socioeconomic condition, the choice cannot always be made, and the student then evaluates the scholarship options. In the group that had word of mouth marketing, the result remains the rejection of hypothesis H1b.
It can be noted that the variable "Quality" in the control group (without word-of-mouth marketing) was significant with a confidence level of 90% and 95%, respectively, in the groups with and without control of campus, gender, income, and father and mother activity.
By demonstrating that students' reflect on this variable before the decision, that is, this variable influences the choice criteriafor an EI. When the data with the interaction of the dummy Treat were analyzed, it was observed that the variable "Quality" continues to be significant with the same level of 90 and 95% confidence respectively in the groups with and without control, but the word-of-mouth marketing practiced by the students do not need to think about the EI quality anymore, as there is now no doubt about the EI quality, demonstrating that WOM marketing enhances the effect of this variable on EI's rejecting the hypothesis H1b, confirming that WOM marketing practiced by EI students influences the decision of new students. (Berger, 2014;Santiago et al., 2014;Dalmonech et al., 2016).
The "Trademark / Image" variable in the control group (without WOM marketing) was significant only at a 90% confidence level. When analyzing the data with the interaction of the dummy Treat, the variable " Trademark / Image" remains significant, but now with 95% confidence, and demonstrates that the WOM marketing practiced by EI students helps future ones to decide by an EI when observing this variable; they do not need to think about other options or be very judicious in their choice process, because the EI has a favorable reputation in the market. Not rejecting the hypothesis H1b, confirming that the WOM marketing practiced by students of IE enhances the influence of the Trademark / Image in the choice criteria for an EI.
It is also noticed that the results were different among the teaching modalities; this can be justified by the profile of the students of each modality, due to their peculiarities. For example, in the integrated technician to high school the students are young people of low family income looking for knowledge and a profession, already in the higher course the others Página | 89 are yearnings, in which they involve quality, image, educational environment and mainly price (without payment of monthly fees ). The latter certainly related to income.
In general, WOM marketing intensifies the influence of factors that have an effect on EI's student choice criteria. Table 2 presents the results regarding the variables that were intensified by the effect of word of mouth marketing, as discussed in hypothesis H1b. transforms the future student's vision, and also some variables that influence on choice criteria.
Thus, it was verified that this study was relevant for the literature, once it contributed on understanding of WOM marketing for an EI choice criteria, besides revealing the students' behavior faced its indication or not; fact not researched yet, demonstrating that WOM marketing does not directly influence the criterion of choice, but it potentiates other variables that influence the criterion of choice. This way, EIs managers could find themselves on this study to improve their efforts in future decision-making regarding projects and public policies related to the students' enrollment.
However, it is noticed some limitations that conditioned this research conclusions.
First, it did not pretend to search data in all IFES campuses. In the same way, the model did not concern about getting different levels of student's identification with the possible career from the chosen course.
Therefore, it is suggested that future researches try to identify why a student who enters the Technical Course Integrated to the High School at IFES with no indication is not influenced by any of the presented variables. Discover what takes these individuals to enter the IFES course and what weighs in their choice criteria by IFES is important. It is also recommended that future researches could test the model created in a larger sample with features representing the diversity of the researched students, ensuring larger perception about their selection criterion by an EI, including investigating the student's identification levels with a possible career set up by the chosen course. It is also suggested applying the same research in a private EI, once this kind of student may have a different choice criterion than the one who looks for a public EI.