https://dinastipub.
org/DIJEFA Vol.
No.
4, 2025
DOI: https://doi.
org/10.
38035/dijefa.
https://creativecommons.
org/licenses/by/4.
The Influence of Education Service Quality on School Selection Interest with School Image as a Mediating Variable at SMA X Bandung Susanti Setiawaty1*.
Bobby Wiryawan Saputra2 Sekolah Tinggi Ilmu Ekonomi Harapan Bangsa.
Bandung.
Indonesia, mm-24171@students.
Sekolah Tinggi Ilmu Ekonomi Harapan Bangsa.
Bandung.
Indonesia, bobby@ithb.
Corresponding Author: mm-24171@students.
Abstract: This study aims to analyze the effect of educational service quality on school choice interest with school image as a mediating variable at SMA X in Bandung City.
The quality of education services is measured based on the dimensions of reliability, responsiveness, assurance, empathy, and tangibles.
School image is assessed through personality, reputation, value, and corporate identity.
School choice interest includes attention, interest, need, enjoyment, and motivation.
A quantitative approach was applied using a survey method involving 238 respondents consisting of students and parents.
Data analysis was conducted using Partial Least Squares Structural Equation Modeling (PLS-SEM).
The results show that:
educational service quality has a positive and significant effect on school image.
educational service quality has a positive and significant effect on school choice interest.
school image has a positive and significant effect on school choice interest.
school image significantly mediates the relationship between educational service quality and school choice interest.
The findings highlight the importance of enhancing educational service quality to build a strong school image and attract student and parent interest in choosing a school.
Keywords: Educational Service Quality.
School Image.
School Choice Interest.
PLS-SEM.
INTRODUCTION
Education is a lifelong part of human life experienced to shape, direct, and enhance the quality of human resources to align with societal aspirations.
Generally, individuals go through formal education school starting from primary, secondary, and up to higher education.
Beyond developing academic intelligence, formal education also functions to build character and skills necessary for daily life (Tilaar, 2.
An article by Siti Afifiyah .
indicates there are 10 skills essential needed by individuals in the workplace, as shown in Figure 1.
3540 | P a g e https://dinastipub.
org/DIJEFA Vol.
No.
4, 2025
Figure 1.
Ten Skills in the Workplace These skills include critical thinking, innovation, problem-solving, technology usage, and various other skills that support success in the modern era.
In this regard.
Senior High Schools (SMA) play a crucial role in developing academic intelligence, character, and the necessary skills to prepare students for higher education and the professional world.
Therefore, selecting a good quality high school (SMA) is an important aspect for both students and parents.
In the school selection process, various factors can influence the decision of students and parents, including the quality of educational services and the school's image (Kotler & Fox.
Research shows that schools with superior educational service quality tend to have higher student and parent satisfaction, which ultimately impacts the decision to choose a school (Putra & Rahmawati, 2.
A good quality educational services encompass effective counseling services, good communication, and complete facilities (Dunggio, 2.
Effective counseling services can assist students with both academic and psychological aspects.
The support provided by counseling helps students resolve personal issues, enchance learning motivation, and provide guidance in making future academic and career choices (Prayitno.
Research by Sari et al.
indicates that schools providing good counseling services have higher levels of student psychological well-being, which impacts their comfort in learning and performing.
Good communication between the school and parents plays a significant role in supporting student's academic and non-academic development.
A harmonious relationship between both parties can increase parental involvement in their children's education, create a more conducive learning environment, and help the school understand student's needs and potential more deeply (Epstein, 2.
According to research conducted by Kraft and Dougherty .
, good communication between schools and parents contributes to improving student's academic achievement, strengthen learning motivation, and reducing negative behavior in the school environment.
Furthermore, digital technology increasingly facilitates interaction between schools and parents through various communication platforms, such as email, school applications, and social media, making information delivery easier and increasing active parental involvement (Olmstead, 2.
With open and effective communication, schools and parents can collaborate to create a better learning experience for students and build a more supportive and collaborative educational environment.
According to research conducted by Yulianto .
, facilities such as well-equipped laboratories, libraries, comfortable classrooms, and adequate sports facilities contribute significantly to student's learning motivation and academic achievement.
Additionally, a study by Uline & TschannenAaMoran .
shows that schools with adequate facilities tend to create a more conducive learning environment, enhance social interaction, and strengthen student's skills in various aspects.
Thus, investing in the procurement and maintenance of quality school 3541 | P a g e https://dinastipub.
org/DIJEFA Vol.
No.
4, 2025
facilities should be a priority for educational institutions to ensure student success in facing the challenges of the modern world.
In addition to the quality of educational services, the school's image is a crucial factor influencing the interest of students and parents in choosing a school.
A school's image is formed from its academic and non-academic reputation, discipline, and good relations between the school and the community (Supriyadi, 2.
Schools with a positive image more easily gain trust from prospective students and parents.
Factors such as academic achievement, level of discipline, and participation in various national and international competitions can enhance the school's reputation (Nugroho, 2.
Therefore, building and maintaining a good school image is an important strategy to increase the interest of students and parents in choosing a school.
Sari .
in her research revealed that consumer decisions in purchasing a product are influenced by various factors, such as product quality, packaging design, brand, seller service, product availability, product size, and sales time.
Before making a purchase, consumers tend to consider these aspects to ensure their satisfaction.
The same principle applies to school selection by students and parents.
The interest in attending an institution depends not only on the quality of education offered but also on other factors such as the quality of educational services .
ounseling services, available facilities, communication effectivity between the school and parent.
and the school's image.
If there is no interest in a particular school, students and parents will certainly not choose that school as a place to continue their education.
Therefore, this research aims to analyze the influence of educational service quality on the decision to choose SMA X in Bandung City, with school image as a mediating variable.
In the context of school competition, understanding the factors that influence selection interest is essential for schools to enhance competitiveness and attract more students.
Therefore, the quality of educational services and the school's image are two key factors believed to have a significant influence on the interest of students and parents in choosing SMA X in Bandung.
Conceptual Framework Figure 2.
Research Model Research Hypotheses H1: There is a significant influence of educational service quality on school choice interest.
3542 | P a g e https://dinastipub.
org/DIJEFA Vol.
No.
4, 2025
H2: There is a significant influence of school image on school choice interest.
H3: There is a significant influence of educational service quality on school image.
H4: There is a significant influence of educational service quality on school choice interest with school image as a mediating variable.
METHOD
Research Method The research method refers to the systematic approach carried out by researchers through specific stages, aiming to solve problems or obtain answers to research questions (Prastowo & Sandra, 2.
In this study, the researcher applied a verificative method because there are several variables whose relationships will be analyzed, and it aims to present data systematically and factually regarding facts and relationships between variables.
According to Djollong .
, quantitative research emphasizes the use of quantitative data in data collection methods in the field with the aim of testing established hypotheses.
Therefore, this research is categorized as quantitative research with a verificative approach.
Research Object This research was conducted at SMA X Bandung, which is the location of the research population, through a series of processes, starting from phenomenon or problem discovery, thesis preparation, research instrument creation, data collection, data processing, to reporting of results, carried out from January to June 2025.
Research Population A population is a group of people, institutions, events, or other subjects to be described or generalized (Vogt & Johnson, 2.
According to KBBI, a population is a group of people, objects, and others that serve as a source for sampling or a collection that meets certain criteria related to the research problem.
According to Roflin.
Liberty, and Pariyana .
, a population is the subject or object to be studied.
From this, it can be concluded that a population is a group that can be described because it shares one or more simillarities.
The population in this study consisted of 582 individuals, comprising 291 students and 291 parents of students at SMA X in Bandung City.
Sampling Technique A sample is a part of the population that becomes the object of research and is used as a data source, where the sample is considered capable of representing the entire population, both in terms of number and characteristics (Asrulla.
Jailani, & Jeka.
According to Hutami .
, a sample is a group of individuals selected from a population and serves as a representation of all members of the population.
A sample is a portion of the population chosen as the object of research and is considered capable of representing the entire population, both in terms of quantity and characteristics.
This study uses a sample determined using the Slovin formula and the Lemeshow formula as follows:
ycA ycu= .
cA y yce 2 )) ycu= .
2 )) ycu = 237,07 OO 238 ycyyceycuycyycoyce 3543 | P a g e https://dinastipub.
org/DIJEFA Vol.
No.
4, 2025
This study uses proportional random sampling technique as the sampling method, which is the random selection of samples based on the proportion of the population size without regard to existing strata (Riyanto, 2.
The sample was obtained from two groups, namely 119 students and 119 parents of students, as both are considered capable of providing data relevant to the research focus.
Data Collection Method Data collection technique refers to the steps or methods used to obtain the data needed in a study.
In this study, data was collected through questionnaires, which is a data collection technique carried out by submitting a number of written questions or statements to respondents (Rahman, 2.
Each question item will be assessed using a scoring system to determine the level or weight in the study.
The determination of these weights refers to a commonly used model, namely a Likert scale with five answer options: strongly disagree (STS), disagree (TS), neutral (N), agree (S), and strongly agree (SS).
Data Processing Technique Validity is a measurement of the accuracy of a test instrument in performing its measuring If a test instrument performs its measuring function accurately and according to its purpose, then its reliability or validity is high (Ramadhan.
Siroj, & Afgani, 2.
validity test is very important to ensure that the instruments used can measure what they intend to measure correctly.
In this study, the validity tests used are convergent validity .
uter loading test and average variance extracted tes.
and discriminant validity .
eterotrait-monotrait ratio and Cross Loadin.
using the Smart PLS software.
Outer Loading Test The outer loading test aims to see the extent to which indicators correlate with the measured construct.
The outer loading test provides a more detailed picture because it examines the extent to which each indicator has a higher correlation with its original construct than with other constructs (Rynkky & Cho, 2.
A good outer loading value is usually > 0.
If the loading value is below 0.
4, it is recommended to consider removing the indicator (Hair et al.
, 2.
Average Variance Extracted (AVE) Test According to Hair et al.
Average Variance Extracted (AVE) functions to examine the convergent validity of constructs within the measurement model.
AVE
calculation is based on the average squared outer loading of indicators that contain the A higher AVE value means that the construct can explain most of the variance of the indicators it uses.
A good AVE value is usually > 0.
Heterotrait-Monotrait Ratio (HTMT) Test The heterotrait-monotrait ratio test aims to show whether different constructs are indeed separate and unique from each other.
This helps ensure that the indicators of one construct do not overlap with other constructs.
HTMT is a recommended method for testing discriminant validity in SEM models, especially due to its higher sensitivity compared to other methods such as the Fornell-Larcker Criterion.
An HTMT value below the threshold of 0.
90 indicates that the tested constructs have good discriminant validity (Henseler.
Ringle, & Sastedt.
Reliability can be defined as the level of confidence or consistency of a measurement result (Ramadhan.
Siroj, & Afgani, 2.
In this study, the reliability test used is Cronbach's Alpha with the help of Smart PLS software, which aims to measure the internal consistency or reliability of a series of survey items.
A higher Cronbach's Alpha value indicates consistent responses to a series of questions.
A variable is considered reliable if Cronbach's Alpha and Composite Reliability exceed 0.
3544 | P a g e https://dinastipub.
org/DIJEFA Vol.
No.
4, 2025
Multicollinearity test is a statistical technique used to detect the presence of multicollinearity among predictor variables.
The multicollinearity test is performed to determine whether there is a high linear relationship between predictor variables that can lead to biased data.
A good model should not have correlations in the predictor variables (OAoBrien, 2.
In this study, the multicollinearity test is based on the tolerance and VIF (Variance Inflation Facto.
values using Smart PLS software, with a limit below or equal Hypothesis Testing Significance Test of Path Coefficient The significance test of path coefficient is a statistical analysis that functions to test the influence between variables, where a positive value indicates a direct relationship and a negative value indicates an inverse relationship (Marliana, 2.
The significance test of path coefficient is used to test direct effects .
ypotheses 1 to .
and indirect effects .
The basis for testing the results is a confidence level of 95% or a significance level of 5% (= 0.
, where the P-value < 0.
05 or t-statistic > 1.
Effect Size Test .
-square.
The effect size test .
-square.
is a statistic test used to measure the strength of the relationship between variables in a structural model.
The effect size .
test is used to complement the significance test .
ath coefficien.
because a significant effect is not necessarily large.
In this study, the test was performed using Smart PLS software.
The limitation of the Effect Size .
test is that it can only test the magnitude of the direct effect .
ypotheses 1 to .
The interpretation of f2 values is shown in Table 1.
Table 1.
Interpretation of f2 Values Coefficient Interval Effect < 0.
Small Medium > 0.
Large Mediating Effect Size Test .
The mediating effect size test .
is a statistic used to test the magnitude of the mediating variable's influence at the structural level (Yamin, 2.
The mediating effect size test .
is used to complement the significance test .
ath coefficien.
because a significant effect is not necessarily large.
In this study, the test was performed using Smart PLS software.
The Mediating Effect Size test .
complements the effect size test .
to examine the magnitude of the indirect effect .
The interpretation of upsilon v values is shown in Table 2 (Yamin, 2.
Table 2.
Interpretation of Upsilon v Values Coefficient Interval Effect < 0.
Small Medium > 0.
Large RESULTS AND DISCUSSION Respondent Description by Gender There were 125 female respondents .
%) and 113 male respondents .
%).
This indicates that more female respondents participated in this study compared to male Figure 3 below shows the respondent data based on gender.
3545 | P a g e https://dinastipub.
org/DIJEFA Vol.
No.
4, 2025
Male Female Figure 3.
Distribution of Respondent Data by Gender Respondent Description by Age There were 119 respondents .
%) in the 15-18 age group, 23 respondents .
%) in the 35-45 age group, 79 respondents .
%) in the 45-55 age group, and 17 respondents .
%) in the 55-65 age group.
The 15-18 age group was the most dominant, accounting for 50%.
This dominance is due to the primary focus of questionnaire distribution being directed towards Meanwhile, the oldest age group, 55-65 years old, only represented about 7%.
Figure 4 below shows the respondent data based on age.
15-18 age group 35-45 age group 45-55 age group 55-65 age group Figure 4.
Distribution of Respondent Data by Age Respondent Description by Occupation There were 119 respondents .
%) who were students, 63 respondents .
%) were entrepreneurs, 38 respondents .
%) were private employees, and 18 respondents .
%) were The composition of respondents was dominated by students, at 50% or 119 people.
Figure 5 below shows the respondent data based on occupation.
Students Entrepreneurs Private Employees Housewives Figure 5.
Distribution of Respondent Data by Occupation 3546 | P a g e https://dinastipub.
org/DIJEFA Vol.
No.
4, 2025
PLS Model Program Scheme Figure 6.
PLS Model Program Scheme Validity Test Outer Loading Test The outer loading test aims to assess the extent to which indicators correlate with the measured construct.
Table 3 below shows the outer loading results for each item.
Table 3.
Convergent Validity Test Results with Outer Loading Test
Latent Variable Manifest Variable Outer Loading Significance KLP1
0,704
Valid
KLP2
0,763
Valid
KLP3
0,705
Valid
KLP4
0,723
Valid
KLP5
0,743
Valid
KLP6
0,753
Valid
KLP7
0,727
Valid
KLP8
0,728
Valid
KLP9
0,576
Sufficient KLP10
0,601
Sufficient KLP11
0,768
Valid
KLP12
0,700
Valid
KLP13
0,751
Valid
Education Service KLP14
0,746
Valid
Quality KLP15
0,598
Sufficient KLP16
0,608
Sufficient KLP17
0,814
Valid
KLP18
0,771
Valid
KLP19
0,763
Valid
KLP20
0,766
Valid
KLP21
0,621
Sufficient KLP22
0,738
Valid
KLP23
0,738
Valid
KLP24
0,773
Valid
KLP25
0,737
Valid
KLP26
0,606
Sufficient KLP27
0,682
Sufficient 3547 | P a g e https://dinastipub.
org/DIJEFA Latent Variable School Image School Selection Interest Vol.
No.
4, 2025
Manifest Variable KLP28
KLP29
KLP30
KLP31
KLP32
KLP33
KLP34
KLP35
KLP36
KLP37
KLP38
KLP39
KLP40
CS1
CS2
CS3
CS4
CS5
CS6
CS7
CS8
CS9
CS10
CS11
CS12
CS13
CS14
MMS1
MMS2
MMS3
MMS4
MMS5
MMS6
MMS7
MMS8
MMS9
MMS10
MMS11
MMS12
MMS13
MMS14
MMS15
MMS16
MMS17
MMS18
MMS19
Outer Loading 0,743 0,740 0,638 0,668 0,778 0,789 0,801 0,840 0,779 0,697 0,809 0,768 0,729 0,824 0,842 0,869 0,853 0,848 0,871 0,854 0,846 0,803 0,855 0,836 0,816 0,808 0,816 0,780 0,786 0,798
0,854
0,856
0,718
0,841
0,855
0,904
0,919
0,897
0,875
0,870
0,878
0,839
0,836
0,484
0,614
0,602
Significance Valid Valid Sufficient Sufficient Valid Valid Valid Valid Valid Sufficient Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Sufficient Sufficient Sufficient From the data in Table 3, 60 items have an outer loading value above 0.
70, indicating that these indicators are valid and contribute strongly to the measured construct.
Meanwhile, 13 items have outer loading values between 0.
40 and 0.
Although their values are below the ideal standard, these items can still be retained because the Average Variance Extracted (AVE) and overall construct reliability are still within high limits, making them acceptable as conveyed by Hair et al .
3548 | P a g e https://dinastipub.
org/DIJEFA Vol.
No.
4, 2025
Average variance extracted (AVE) Test Based on Table 4, the Average Variance Extracted (AVE) values for each construct meet the minimum criteria of above 0.
This indicates that the proportion of variance successfully explained by the indicators in each construct is high, thus the convergent validity in this model can be categorized as very good.
Table 4.
Convergent Validity Test Results with AVE Test
Latent Variable AVE Value Education Service Quality 0,529 School Image
0,704
School Selection Interest
0,653
Discriminant Validity Test Based on Table 5, the discriminant validity values measured using the HeterotraitMonotrait Ratio (HTMT) test show results below the threshold of 0.
Thus, all constructs in this model are declared discriminantly valid.
This means that each construct in the model truly measures a different concept from each other, and there is no overlap in meaning between constructs.
Table 5.
Discriminant Validity Test Results using HTMT Test School Image (Z) Education Service Quality (X) 0,342 Education Service Quality (X) 0,845 0,374 School Selection Interest (Y) Reliability Test In this study, the reliability test used was Cronbach's Alpha with the help of Smart PLS software, aiming to measure the internal consistency or reliability of a series of survey items.
A higher Cronbach's Alpha value indicates a consistent response to a series of questions.
Table 6 below shows the reliability test results and is declared very reliable because yu > 0,8.
Table 6.
Reliability Test Results using Cronbach's Alpha Cronbach's Alpha Composite Reliability 0,968 0,971 School Image (Z) 0,977 0,978 Education Service Quality (X) 0,969 0,972 School Selection Interest (Y) Criteria Very Reliable Very Reliable Very Reliable Structural Model Test Table 7 below shows that there is no multicollinearity between variables because the VIF value is <5.
Table 7.
Multicollinearity Test Results based on VIF values School Image (Z) School Selection Interest (Y) School Image (Z) 1,166 Education Service Quality (X) 1,000 1,135 Path Coefficient Significance Test The results of the path coefficient significance test to examine the direct effect .
ypotheses 1 to .
are shown in Table 8 below.
3549 | P a g e https://dinastipub.
org/DIJEFA Vol.
No.
4, 2025
Table 8.
Path Coefficient Significance Test Results for Direct Effect
Original Sample Standard
T Statistics Sample Mean
Deviation School Image (Z) -> School
0,794
0,796
0,032
24,559
Selection Interest (Y) Education Service Quality (X) -> 0,341 0,345 0,052 6,603 School Image (Z) Education Service Quality (X) -> 0,109 0,112 0,038 2,871 School Selection Interest (Y) P Values 0,000 0,000 0,004 The results of the path coefficient significance test to examine the indirect effect .
are shown in Table 9 below.
Table 9.
Path Coefficient Significance Test Results for Indirect Effect Original Sample Standard T Statistics Sample Mean Deviation School Image (Z) -> School Selection Interest (Y) Education Service Quality (X) -> School Image (Z) Education Service Quality (X) -> 0,271 0,275 0,043 6,371 School Selection Interest (Y) P Values 0,000 Based on these provisions, the hypothesis testing results for direct and indirect effects can be seen in Table 10 below.
Hypothesis Hypothesis 1 Hypothesis 2 Hypothesis 3 Hypothesis 4 Table 10.
Hypothesis Testing Results Path t value From Via X (KLP) Y (MMS) 2,871 1,96 Z (CS) Y (MMS) 24,559 1,96 X (KLP) Z (CS) 6,603 1,96 X (KLP) Z (CS) Y (MMS) 6,371 1,96 0,004 0,000 0,000 0,000 Decision Ho Rejected Ho Rejected Ho Rejected Ho Rejected Hypothesis 1 (H.
: Assesses the effect of Educational Service Quality on School Selection Interest.
Based on the test results, educational service quality shows a significant direct effect on school selection interest.
This confirms that good educational services directly influence prospective students' preferences in choosing a school.
Hypothesis 2 (H.
: Assesses the effect of School Image on School Selection Interest.
The test results show that school image has a significant direct effect on school selection This finding indicates that a positive perception of school image encourages prospective students/parents to choose that school.
Hypothesis 3 (H.
: Assesses the effect of Educational Service Quality on School Image.
The results show that educational service quality also has a positive and significant effect on school image, indicated by a positive path coefficient and a t-statistic value above the critical value (> 1.
96 for = 0.
This indicates that an improvement in service quality will enhance the school's image.
Hypothesis 4 (H.
: Assesses the effect of Educational Service Quality on School Selection Interest with School Image as a Mediating Variable.
These results indicate that school image mediates the relationship between educational service quality and school selection interest, meaning that a positive perception of service quality will form a good school image, and that positive image will ultimately encourage an increase in interest in choosing the school.
3550 | P a g e https://dinastipub.
org/DIJEFA Vol.
No.
4, 2025
Effect Size .
Test Table 11.
fA Test Results for Direct Effect School Image (Z) School Selection Interest (Y) School Image (Z) Education Service Quality (X) 1,724 0,033 0,132 Table 11 shows the effect size .
A) values for each relationship between constructs as The effect of Educational Service Quality on School Selection Interest shows an fA value 033, which falls into the small category.
This indicates that the direct contribution of educational service quality to school selection interest is quite limited.
The effect of School Image on School Selection Interest has an fA value of 1.
724, which is a very large effect and far exceeds the large category threshold.
This indicates that school image is the most influential construct on school selection interest.
The effect of Educational Service Quality on School Image shows an fA value of 0.
which is still in the small to moderate category.
Nevertheless, this finding still shows that educational service quality contributes to forming a school's image, although not Mediation Effect Size .
Test Based on the calculations in Table 12, the role of school image in mediating the indirect effect of educational service quality on school selection interest at the structural level is categorized as a moderate effect.
Table 12.
Upsilon v Test Results Effect Educational service quality -> school image -> school selection interest Upsilon .
Statistic .
= 0,073 Description Influence towards Discussion Hypothesis 1: The Effect of Educational Service Quality on School Selection Interest The analysis results show that educational service quality has a direct and significant effect on school selection interest with a t-statistic value of 2.
871 (>1.
and a p-value of 0.
(<0.
However, the effect size .
A) value of 0.
033 indicates that the magnitude of this effect is relatively small.
This finding is consistent with research by Erinawati & Syafarudin .
, which confirms that educational service quality has a positive and significant effect on the decision to choose a school.
This means that the better the perceived service .
uch as facilities, teacher responsiveness, administrative reliability, and empath.
, the greater the likelihood of students .
nd parent.
choosing that school.
Research by Azkiyah et al.
proves that the quality of academic services has a significant effect on the interest of new students with a contribution of 40.
This is in line with the findings of this study that perceptions of service quality influence the tendency of prospective students to choose an educational institution.
Hypothesis 2: The Effect of School Image on School Selection Interest School image is proven to have a very significant effect on school selection interest, with a t-statistic value of 24.
559 and a p-value of 0.
The effect size .
A) value of 1.
724 indicates a very large effect.
Research by Rosha et al.
shows that school image has a positive and significant effect on the decision to choose SD Islam Al-Azhar 32 Padang.
A strong image, reflected by A accreditation, academic reputation, and national partnerships, is proven to be an important determinant in attracting public interest.
Research by Alifiah .
shows that 3551 | P a g e https://dinastipub.
org/DIJEFA Vol.
No.
4, 2025
school image has a significant effect on parents' interest in choosing a school, with a significance value of 0.
A positive image makes it easier for schools to build public trust, thereby encouraging parents' decisions to enroll their children.
Hypothesis 3: The Effect of Educational Service Quality on School Image The results show that educational service quality has a positive and significant effect on school image .
= 6.
p = 0.
The effect size for this relationship is 0.
132, categorized as small to moderate.
Research by Habibah and Bayu .
shows that service quality has a significant effect on school image with a relationship of 0.
This strengthens the notion that public perception of a school is formed through services directly experienced by students.
Similarly, research by Budiarti et al.
shows that service quality has a significant effect on institutional image with a coefficient of 0.
This indicates that an increase in the quality of educational services will strengthen the public's positive perception of the school's Hypothesis 4: The Indirect Effect of Educational Service Quality on School Selection Interest through School Image The indirect effect test shows that the indirect effect of educational service quality on school selection interest through school image is significant .
= 6.
p = 0.
The upsilon v value of 0.
073 indicates that the mediating effect is in the moderate category.
This finding supports the mediation theory by Baron & Kenny .
, which explains that mediation occurs when an intermediary variable bridges the relationship between two other variables.
In this context, school image becomes an important channel in transforming service perception into concrete interest in choosing.
Research by Kusumawati.
Yanamandram, & Perera .
in the context of Indonesian universities shows that school image acts as a mediator between service quality and the decision to choose an educational institution.
CONCLUSION
Based on the results of data analysis and hypothesis testing using the Partial Least Squares (PLS-SEM) method, the following conclusions are obtained:
Educational service quality has a direct and significant effect on school selection interest, but with a relatively small contribution .
A = 0.
This indicates that although the effect is significant, its impact on students'/parents' interest in choosing a school is still limited if based solely on service quality.
School image has the most dominant effect on school selection interest, as evidenced by an fA value of 1.
724, which falls into the very large effect category.
This means that a positive perception of school image is the main factor influencing the decision of students and parents in choosing a school.
Educational service quality also has a significant effect on school image, although its effect is categorized as moderate with an fA of 0.
This confirms that good service can shape the public's positive perception of the school.
School image is proven to mediate the relationship between educational service quality and school selection interest, with an upsilon v value of 0.
073, which falls into the moderate effect category.
This indicates that the indirect impact of educational service quality on school selection interest through school image is stronger than its direct impact.
Based on these conclusions, the researchers provide the following suggestions:
Schools are advised to continuously improve their image, both through promotional media, school achievements, and involvement in social and educational activities.
This is because school image is proven to be a key factor in increasing the interest of prospective students.
3552 | P a g e https://dinastipub.
org/DIJEFA Vol.
No.
4, 2025
Improving the quality of educational services remains important, especially in aspects such as reliability, responsiveness, and empathy towards the needs of students and parents.
Although its direct impact is small, service quality forms the foundation for a positive school image.
Optimizing external communication strategies, so that positive perceptions of the school are not only known by internal parties .
but also reach prospective students and parents outside the school environment.
Social media, testimonials, and public information disclosure activities can be maximally utilized.
Future research is suggested to explore other factors that may influence school selection interest, such as educational costs, school location, excellent programs, or religious values, to broaden the understanding of school selection behavior.
REFERENCES