The Impact of Academic Aspect Quality on Student Disengagement in Higher Education

Abstract

Higher education (HE) in developing countries like Vietnam has shown to be a growing sector even under the worldwide impacts of the COVID-19 pandemic. However, due to the typical characteristics of education in general and HE in particular, there is a crucial issue for HE institutes in balancing growing objectives as a service business and quality assurance objectives including student engagement. The current literature still lacks studies on student disengagement in HE as it focuses on the positive valence of engagement. Hence, this study aims to determine the impact of perceived academic aspect quality (PAAQ) on student disengagement (SD) intention and actual behavior under the lens of extended TRA. The results of our structural equation model reveal the full mediating role of student dissatisfaction in the relationship between PAAQ on SD intention. Once the students have the intention to disengage in learning tasks, they are most likely to act on it. Perceived disengaged behavior cost is robustly proved as a key driver of SD behavior intention and actual disengagement behavior.

1. Introduction

The COVID-19 pandemic has brought about a crisis in all aspects of society all over the world and attracted the attention of numerous scholars. In the socio-economic sector, developing countries are shown to be highly vulnerable to this pandemic and it severely affected work and the workforce [1,2]. However, HE seems not to have suffered much from this catastrophic global event from the aspect of recruiting new students. In developing countries such as China and Vietnam, degrees are highly promoted and considered a must to ensure the possibility to secure future employment. The number of Asian students studying abroad dramatically increased over time for high-ranked international degrees to be recognized [3,4]. Though, the worldwide lockdown since 2020 has hampered this big wave of overseas students from emerging countries. That accordingly even brought the chance for the home-based public and private higher education institutes (HEI) in opening more “hot” majors and courses, establishing international cohort programs which previous has been booming since the marketization policies and the implementation of the financial autonomy mechanism of public universities. However, it also raises the issue of balancing the marketing and branding objectives of HEI as a business and the quality assurance objectives of an educator. It means that as a business coping with increasing competition, HEIs aim at recruiting more students, gaining their satisfaction, keeping them positively engaged with the HEI by enrolling in further higher education programs, giving favorable WOMs and giving recommendations about the HEI, and supporting the HEI as alumni [5,6,7,8,9,10,11]. As educators, they should aim at student engagement in studying tasks for quality assurance, higher retention, and prevention of dropping out (e.g., [12,13,14]).

In that circumstance, teaching staff in HEI are multitasking with teaching, developing new programs and courses, researching and other support services, and administrative jobs [15]. They are considered as service-provider staff and have to assure their “customer” satisfaction. Most HEIs, in Vietnam for instance, have now conducted surveys on students’ assessment of teaching quality for the purpose of gaining national and international accreditation certificates. As an educator, teaching staff have to meet the student’s more and more demanding academic needs as with the popularity of the internet, students can now easily access a variety of online worldwide academic sources. Many of those are free of charge. As the result, it is almost obvious that when students pay for the costly services to HEIs it encompasses high involvement of students and leads to their initiative in making a post-purchase assessment [16]. How to get the students engaged in the studying tasks as well as making them satisfied has become an even more challenging job for teachers than it already was [17,18]. As a result, this research is looking for the answer to the question of what if students are dissatisfied with academic aspect quality. Does it relate to the disengagement behavior of the students? How?

2. Literature Review and Theoretical Framework

2.1. Student Engagement (SE) vs. Student Disengagement (SD)

Student engagement is defined as the “time and effort students devote to educationally purposeful activities” [19]. The current literature on higher education shows that the range of definitions for student engagement converges to emphasize three interrelated aspects of student engagement: cognitive, behavioral, and affective [20]. Later on, as noted by Kuh et al. [21], student engagement (SE) represents two critical features. The first is the amount of time and effort students put into their studies and other educationally purposeful activities. The second component is how the institution deploys its resources and organizes the curriculum, other learning opportunities, and support services to induce the student to participate in activities that lead to the experiences and desired outcomes such as persistence, satisfaction, learning, and graduation. From this perspective, by identifying the second component of SE, literature on SE also viewed students as “customers” rather than solely “products” and pays attention to their satisfaction with the HEI’s offerings, though still frequently used the term “student engagement” instead of “customer engagement”.

Student engagement where existing students are viewed as focal actors is considered a buzzword in higher education. SE has been increasingly researched, theorized, and debated with growing evidence of its critical role in achievement and learning [12,13,14,22,23,24,25]. We can retrieve four major research perspectives on SE: the behavioral perspective, which emphasizes student behavior and institutional practice; the psychological perspective, which defines engagement as an individual psychosocial process; the socio-cultural perspective, which emphasizes the importance of the socio-political context; finally, the holistic perspective, which takes a broad view of engagement. Studies from all four perspectives focus on identifying antecedents, dimensions, measurement, and outcomes of engagement of the student in learning tasks to promote positive SE.

Regarding non-positive SE, disengagement of learners, in general, is considered one of the biggest challenges for teachers in the classroom every day and draws the attention of numerous scholars [17,26,27]. In their published book, Fredricks et al. reveal the common fact that learners from elementary, secondary, and high school often view academics as boring and having little meaning in their lives. As a result, researchers in the education sector have also paid much attention to this topic. Their purpose is to understand why the learners disengage in learning tasks to prevent it and find solutions for higher student retention and reduce the rate of dropping out.

However, in higher education, the students are over 16 years old, and in some Asian countries like Vietnam, first-year students in HE are 18-year-old adults, so theoretically they are capable of making their own decision in their education path for their future. As a result, another conceptual framework is required for student disengagement in higher education. Chipchase et al. [18] argued that like engagement, disengagement is likely to not be a “steady state” characteristic of students. Disengagement, whether ongoing or intermittent, obvious or subtle, may result in the student dropping out of units or a course of study. Accordingly, that would lead to the accumulating of debt or achieving lower grades with poorer employment prospects [28]. For HEIs, this may result in loss of income and, if the problem is sufficiently large, in negative HEI reputational effects. Chipchase et al. [18] reviewed literature in SD and categorized indicators of student academic disengagement into eleven target areas. Though the identified indicators are limited to indicators of academic disengagement measured in the first-year cohort, they are also relevant and measurable to all cohorts and will also identify students for whom disengagement may occur later in their program of study (2017). Further research to complete the theoretical framework of SD in HE as well as empirical research to examine the framework are still in need and waiting for the attention of researchers.

2.2. Academics Aspect Quality and Student Dissatisfaction

As the quality of services provided and student satisfaction are essential for the survival of an HEI in the educational market [29], many studies have dealt with the theme of service quality in HEIs, such as the adaptations of the SERVQUAL [30] and SERVPERF [31] scales, as well as the development of new instruments such as HEdPERF [32], the HEDQUAL [33] scale, and other scales developed by individual authors.

For Abdullah [24], the generality of the SERVQUAL and SERVPERF scales is still hazy when they are replicated to evaluate perceived HEI quality, as, even with studies on service quality, there remain unresolved questions, mainly in regards to the most proper measurement instrument for evaluating each type of service. The use of the most appropriate measurement tool would help HEI managers in assessing provided service quality properly, thus enabling them in using the results to improve the service performance [34].

Thus, Abdullah [24] created a new measurement scale named HEdPERF, which was based on the SERVPERF scale, which considers the specific determinants of service quality in higher education such as non-academic aspects, academic aspects, reputation, access, program issues, and understanding. Icli and Anil [33] assess HEdPERF to be the most developed scale in the literature to measure service quality in higher education. Among these determinants, researchers in SE have paid great attention to and proved the important impact of “academic aspects”, “educators”, and “teaching staff” on student engagement in learning activities in traditional as well as online environments [22,25,35,36,37] and students satisfaction and advocacy to the school [6,8,9,10,11,38,39]. Furthermore, it is observed the teaching staff dilemma in the changing HE context with increasingly fierce competition among domestic and international-related HEIs, and even within an HEI. The teaching staff in HEI are multitasking staff in an education service business to serve the “customer” for HEI’s objectives in competition and growth, and also respectful educators for their students. Hence, the urging question is how do the students perceive academic aspect quality which is under the teaching staff’s responsibility, and are they dissatisfied at any level? Does their perception of academic aspect quality relate to their disengagement in learning tasks?

2.3. Theory of Reasoned Action (TRA) and Perceived Behavior Cost

The theory of reasoned action (TRA) [40] proposed that behavioral intentions could be explained by “attitudes” towards a particular behavior and “subjective norms”. As explained by the TRA, individuals’ behavioral intention will increase if their attitudes toward the behavior become more favorable. Subjective norms denote the perceived social pressure to perform a behavior or not. Accordingly, individuals’ intention to perform a certain behavior will increase if their subjective norms toward that behavior become more favorable. Attitude is regarded as a direct predictor of behavioral intention, which will shape individuals’ intentions toward specific issues. Social norms denote the social pressure that individuals perceive to have to engage in a behavior, mostly based on beliefs about expectations of relevant reference groups related to the behavior. Therefore, they perceive themselves to be expected to behave in a certain manner.

In addition to some addressed typical characteristics of the education sector to date:

The main task of students in school is to study.

Teachers are highly respected in many societies, especially in Asian countries, and thereby by students [41].

The popularity of the internet and social networking sites and social media has brought about a change from traditional to online and blended learning environments that may cause benefits and barriers for both teacher and student [35,42], and the ease for students in connecting with each other, and access free online knowledge sources.

Then, the authors apply the TRA as a lens for examining the impact of attitudinal factors on student disengagement intention and behavior in HEI. Furthermore, student Engagement in learning tasks in blended learning requires exploiting resources like a computer, mobile phone connected to wifi for online learning, as well as knowledge and skills in studying online. Meanwhile, disengagement is the opposite of engagement which means NOT contributing to class discussion, NOT participating in discussions online or face to face, etc. Individuals are supposed to be typically motivated by weighting expected costs and benefits of alternatives [43]. Then the authors extend a factor of perceived disengaged behavioral costs of the disengagement behaviors to the theoretical framework of this research. The theoretical framework is shown in Figure 1 below.

Supposing that in the role of the learner, the students may still think disengagement behavior is not good. However, when perceiving low academic teaching quality, instead of complaining directly to the teachers due to their respect for them [41], they may choose to passively engage or disengage in learning activities as observing the same response from other students. Students may think that even if they disengage in learning tasks to some extent, they still pass the course or get good grades due to the teaching staff’s hesitation to fail the students or give low grades as it somehow reflects teaching quality [44]. Under huge workload and seeing that HE students are adults, teaching staff also may find disengagement behavior of students is normal.

Accordingly, based on the above literature review and arguments, this research will test the hypotheses below:

Hypothesis 1 (H1). 

Perceived academic aspect quality negatively relates to student dissatisfaction.

Hypothesis 2 (H2). 

Perceived academic aspect quality negatively impacts student disengagement behavior intention.

Hypothesis 3 (H3). 

Student dissatisfaction positively mediates the relationship between perceived academic aspect quality and SD behavior intention.

Hypothesis 4 (H4). 

Students’ attitude toward SD behavior negatively influences their intention to have disengagement behavior.

Hypothesis 5 (H5). 

Students’ subjective norm toward SD Behavior positively influences their intention to have SD behavior.

Hypothesis 6a (H6a). 

Students’ perceived disengaged behavior cost toward SD behavior positively influences their intention to have SD behavior.

Hypothesis 6b (H6b). 

Students’ perceived disengaged behavior cost toward SD behavior positively influences their actual SD behavior.

Hypothesis 7 (H7). 

Students’ intention to have disengagement behavior will positively influence actual SD behavior.

3. Methodology

3.1. Measurements

Measurements in this study are adopted from established instruments in the literature. Specifically, academic aspect quality items are adapted from the HedPerf model [32]; student disengagement items are employed and developed from Bergdahl et al. [27]. Three attitudinal constructs are adapted and developed from the theory of planned behavior [43,45]. Apart from items associated with the studied constructs, we also solicit demographic and basic information from respondents regarding gender, year at university, affiliation, and homeland.

Prior to the official distribution of the survey questionnaire, we conducted two pilot tests. The first one is for the purpose of face validity. Thus, two experts were asked to read and provide feedback in terms of terminology. Second, 50 students were asked to answer the survey and provide feedback also in terms of terminology. We adjusted the survey questionnaire used for official distribution based on this feedback.

3.2. Research Setting and Data Collection

Three higher education institutions located in Hanoi, Vietnam were selected to participate in this study. First, we ask for student affairs divisions of these three universities for their approval to conduct this study. Subsequently, we were allowed to access the three Facebook groups gathering students from these three institutions. Next, we posted the link directing to the online questionnaire in these three groups asking student members of these three groups to answer the questionnaire. Data collection was undertaken between February 2022 and April 2022. Eventually, we obtain 462 validated answers.

4. Research Results

4.1. Demographic and Basic Information of Respondents

Table 1. Demographic and basic information of respondents.

Characteristic	Respondents
	Frequency (n = 462)	%
Gender
Did not disclose	5	1.1
Male	89	19.3
Female	368	79.7
Year
First year	155	33.5
Second year	163	35.3
Third year	93	20.1
Fourth year	51	11.0
Affiliation
VNU University of Business and Economics	25	5.4
National Economics University	167	36.1
VNU International School	261	56.5
Not to disclosed	9	1.9
Homeland
Hanoi	189	40.9
Other cities/provinces	273	59.1

represents the demographic and basic information of our surveyed respondents. Specifically, among 462 respondents, 89 are males (19.3%), 368 are female (79.7%), and the other 5 respondents (1.1%) did not want to disclose their gender. Regarding students’ years at university, 155 are in their first year (33.5%). The respective figures for the second, third, and fourth years are 163 (35.3%), 93 (20.1%), and 51 (11%). Regarding affiliation, 25 (5.4%) are from VNU University of Business and Economics, 167 (36.1%) are students of National Economics University, and 261 (56.5%) are from VNU International school. In terms of homeland, 189 (40.9%) are from Hanoi and 59.1% are from other cities or provinces.

4.2. Measurement Validation

In order to verify the validation of measurement, we employ confirmatory factor analysis (CFA), using AMOS 24.0. As shown in

Table 2. Results of multiple fit indices.

Index	Result	Acceptable Level
Chi-square	1375.070	-
Degree of freedom	489	-
Chi-square/Degree of freedom	2.812	<5
TLI	0.932	>0.9
IFI	0.941	>0.8
NFI	0.911	>0.9
RMSEA	0.063	<0.08
CFI	0.941	>0.9

, all results of multiple fit indices are satisfactory with relevant indices such as CFI, NFI, IFI, TLI, RMSEA, and Chi-square/degree of freedom being higher or lower than the respective acceptable levels. Next, factor loadings are examined with all items’ factor loadings lower than 0.7 being removed. As shown in

Table 3. Results of factor loading for confirmatory factor analysis.

Items	Mean	SD	Factor Loading
Perceived Academic Aspect Quality (PAAQ)
PAAQ1	5.97	1.211	0.799
PAAQ2	5.73	1.268	0.901
PAAQ3	5.34	1.465	0.779
PAAQ4	5.54	1.315	0.876
PAAQ5	5.76	1.266	0.913
PAAQ6	5.76	1.261	0.908
PAAQ7	5.16	1.544	0.711
PAAQ8	5.33	1.425	0.808
PAAQ9	5.98	1.187	0.826
Dissatisfaction (DISSAT)
DISSAT1	3.22	1.976	0.997
DISSAT2	3.16	2.036	0.943
DISSAT3	3.41	2.173	0.834
Student Disengagement Behavior Intention (DBI)
DBI2	2.04	1.540	0.893
DBI3	1.89	1.511	0.957
Attitude Toward Disengagement Behavior (ATDB)
ATDB1	4.82	1.852	0.755
ATDB2	5.33	1.865	0.917
ATDB4	5.55	1.827	0.897
ATDB5	5.42	1.739	0.651
Actual Student Disengagement Behavior (DB)
DB1	2.47	1.746	0.696
DB2	1.97	1.527	0.943
DB3	1.86	1.484	0.952
Subjective Norm to Disengagement Behavior (SNDB)
SNDB1	3.56	1.829	0.872
SNDB2	3.77	1.851	0.917
Perceived Disengagement Behavior cost (PDBC)
PDBC1	3.45	1.870	0.753
PDBC2	3.01	1.857	0.789
PDBC3	2.90	1.884	0.866
PDBC4	3.12	1.949	0.89
PDBC5	2.88	1.910	0.907

, only items with factor loadings higher than 0.7 remained. The results of factor loadings partly ensure the convergent validity of our empirical data. In order to further confirm the convergent validity, composite reliability (CR), and average variance extracted (AVE) are also accessed. As shown in

Table 4. Convergent and discriminant validity.

	CR	AVE	PAAQ	SAT	DBI	ATDB	ATNE	DB	SNDB	PDBC	SNEB
PAAQ	0.955	0.702	0.838
SAT	0.948	0.860	−0.103 *	0.927
DBI	0.922	0.856	−0.133 **	0.345 ***	0.925
ATDB	0.884	0.660	0.344 ***	−0.089 †	−0.128 *	0.812
ATNE	0.916	0.785	−0.120 *	0.268 ***	0.400 ***	0.047	0.886
DB	0.903	0.760	−0.160 **	0.348 ***	0.858 ***	−0.085 †	0.431 ***	0.872
SNDB	0.889	0.801	−0.020	0.278 ***	0.348 ***	0.256 ***	0.398 ***	0.332 ***	0.895
PDBC	0.924	0.711	−0.036	0.350 ***	0.492 ***	0.050	0.479 ***	0.502 ***	0.606 ***	0.843
SNEB	0.902	0.754	⁂	0.211	0.373	0.088	0.868	0.391	0.358	0.436	0.869

Note: † p > 0.050; * p < 0.050, ** p < 0.010, *** p < 0.001, ⁂ Correlation is not specified in this model.

, all CR and AVE values are higher than their respective acceptable levels (0.7 for CR and 0.5 for AVE). Apart from convergent validity, discriminant validity may also be a problem in CFA. To address this issue, we compare the values of AVE with the square of correlation coefficients. As shown in Table 4, all AVE values are higher than the respective squared correlation coefficient, we may conclude that discriminant validity is not a problem in this study.

4.3. Results of Structure Equation Model

Results of the structural equation model (SEM) are shown in

Table 5. Results of structural equation model.

	Beta Coefficient	p Value	Hypothesis
Dependent variable: DISSAT
PAAQ	−0.105	0.027	H1 supported
R2	11%
Dependent variable: DBI
PAAQ	−0.041	0.329	H2 not Supported
DISSAT	0.172	***	H3 supported
SNDB	0.08	0.19	H5 not Supported
ATDB	−0.151	0.001	H4 supported
PDBC	0.31	***	H6a supported
R2	29.1%
Dependent variable: DB
DBI	0.805	***	H7 supported
PDBC	0.105	0.002	H6b supported
R2	73.5%

Chi-square = 1520.786; degree of freedom = 505; Cmin/df = 3.011; normed fit index (NFI) = 0.902; root mean square error of approximation (RMSEA) = 0.066; the Tucker-Lewis coefficient (TLI) = 0.924 and comparative fit index (CFI) = 0.932; PAAQ: perceived academic aspect quality; DISSAT: dissatisfaction; DBI: student disengagement behavior intention; ATDB: attitude toward disengagement behavior; DB: actual student disengagement behavior; SNDB; PDBC: perceived disengagement behavior cost; *** p-value is close to 0.

and Figure 2. Specifically, our empirical analysis computes that 11% of the variance of DISSAT is explained through PAAQ. In the same vein, 29.1% of the variance of DBI is explained through DISSAT, ATDB, and PDBC; meanwhile, 73.5% of the variance of DB is explained through DBI and PDBC.

Regarding path analyses, our empirical analysis reveals six out of eight hypotheses are accepted while the two others are rejected. Specifically, PAAQ appears to have a significant negative impact on DISSAT (β = −0.105 p-value < 0.005). Thus, H1 is accepted. DISSAT and PDBC have significant positive impacts on DBI with β-values and p-values of 0.172 (p < 0.001) and 0.31 (p < 0.001), respectively; whereas ATDB has a significant negative impact on DBI with β-values and p-values of −0.151 and 0.001 (<0.05), respectively. Therefore, H3, H4, and H6a are accepted. The β-values regarding the relationships between PAAQ-DBI and SNDB-DBI are found as higher than 0.05. Thus, H2 and H5 are rejected.

Last but not least, our empirical analysis shows that DBI and PDBC have both significant impacts on DB with the respective values of the relationship between DBI and DB are 0.805 (β value) and <0.001 (p-value) and the respective values of the relationship between PDBC and DB are 0.105 (β-value) and 0.002 (p-value < 0.005). Therefore both H7 and H6b are accepted.

5. Discussion and Conclusions

Student engagement has been considered an important topic in educational practices and by scholars. As the majority of studies focus on positive student engagement, there have been few studies of their disengagement in studying tasks despite their benefits possibly brought to higher education providers. It is necessary to explore the antecedents and test the impact of those factors on student disengagement as it is not obvious how the factor motivating student engagement the most would impact disengagement. This study fills this gap by examining the antecedents of student disengagement of 462 Vietnamese students. The effect of perceived academic aspect quality on student disengagement is tested directly and under the mediation of student dissatisfaction. Furthermore, this study applies the extended theory of reasoned action model including attitude, subjective norms, and perceived disengaged behavior cost to explain student disengagement.

5.1. Theoretical Implications

This empirical study contributes to the literature on student engagement (SE) through a better understanding of the non-positive valence of SE. Our empirical data confirm the negative impact of academic aspect quality on student dissatisfaction. It revealed a full rather than partial mediating role of dissatisfaction on the impact of perceived academic aspect quality on student disengagement behavior intention. This finding is in line with other previous studies within the context of higher education (e.g., [46,47,48]). However, as this study revealed that perceived academic aspect quality is not the most influencing determinant of dissatisfaction and showed no direct impact of that factor on disengagement behavior intention, it suggests that perceived academic aspect quality is not necessarily the best predictor of student disengagement.

This study justifies that the extended TRA model is not fully sufficient to explain student disengagement intention as subjective norms’ effect on SD intention is not supported. This finding is contrary to some previous studies on HE (e.g, [45]) and some previous studies in different settings such as online gambling (e.g, [49]). A plausible explanation for this is that students intend to disengage with learning tasks for other reasons rather than considering what their important people think about that behavior. As in this study, perceived disengaged behavior cost is proved as a key driver of student disengagement behavior intention. Furthermore, research results confirm the negative impact of attitude on disengagement behavior intention which means they believe that their main job in HE is to study.

This study also presents the high score of disengagement behavior intention impact on actual disengagement behavior. It can be interpreted that once the students intend to disengage in learning tasks, they would likely act on it. The perceived disengaged behavior cost is shown to positively stimulate the intention to become actual disengaged behavior.

5.2. Practical Implications

Our findings have implications for practitioners. First, as this study determined the negative impact of academic aspect quality on student dissatisfaction, we suggest that higher education institutes should keep on improving the quality of academic aspect under the responsibility of teaching staff. However, based on the empirical data of this study, this element only explains 11% of the student dissatisfaction; hence, HEIs should pay higher attention to other elements, for instance, non-academic aspects, access, program issues, and understanding as in Hedperf model [32]. This can be explained by the fact that all universities in the research sample are top universities in Vietnam. Though they have rapidly opened new programs and developed new courses and modules, teaching staff are still basically capable of satisfying the academic needs of students. This element hence is not the best explanation for student disengagement behavior intention.

Second, this study partly confirmed the appropriateness of the extended theory of reasoned action [40] as a predictor of student disengagement behavior intention. The attitude factor is proved to negatively impact disengagement behavior intention. As the respondents of this study are students from top universities and most of them are female, they have good learning attitudes. This is the competitive advantage in quality assurance and reputation that they can leverage.

However, perceived disengaged behavior cost is robustly proved as a key driver of student disengagement behavior intention and actual disengagement behavior with this empirical data from these top universities. It means that, even with good students, if they perceive low costs of disengagement, they still possibly intend to disengage in learning tasks and would likely disengage. As student evaluation of teaching survey (SET) is a popular method of assessing the quality of teaching staff in HE [44], there exists the fact that teaching staff tends to give students high scores for a couple of reasons: (i) high scores represent high teaching quality; (ii) high scores make students satisfied then they would highly assess teaching staff in the SET survey. As a result, it is suggested that HEIs should consider what, when, and how to use student surveys in evaluating teaching staff. For instance, with the advanced technology in digital platforms like learning management systems, HEIs can design the survey at the proper touchpoint of students’ experience in the system to gain the most objective responses. The data collected in the system will be more easily processed and analyzed for accurate findings to improve academic aspect quality. Moreover, the survey should not be around academic aspects only but should cover other elements in HEI service offerings as well.

Another reason aside from the two reasons mentioned above is (iii) high scores in the transcript would favor the students in job-seeking competition over other top schools’ students. This would accordingly promote HEI’s reputation and increase the recruitment capability for any new programs they open. So once the students perceive that even if they disengage in learning tasks they are still able to pass or get high scores as a privilege of top school students, HEIs would hardly maintain or improve the quality of their “product”, graduated students under the evaluation of employers in the labor market. Hence, the students’ survey on their assessment and satisfaction with academic aspects should not be the only and dominant indicator in the evaluation of academic teaching staff. Teaching staff are then free from the pressure of achievement obsession to grade their students reflecting exactly students’ capacity. Accordingly, the disengaged behavior costs are increased and it would help to prevent student disengagement behavior intention.

5.3. Limitations and Future Research

Like many other studies, there might find a number of caveats for further amelioration in this study. First, this study relied on student surveys with subjective scales. Accordingly, other secondary data such as student grade reports, student evaluation of teaching reports, etc., besides primary perceived academic aspect quality data should be used and analyzed in future research on related topics.

Second, another limitation is that only the academic aspect quality of the multidimensional HE service quality is considered in this study. Prospective researchers may explore the impact of other elements of HE service quality on the non-positive engagement of students.

Third, this study only focused on the behavioral component of disengagement. The conceptual model of Naumann et al. [50] suggests that disengagement valance includes cognitive and affective (emotional) components aside from the attitudinal component. Future researchers may conceptualize student disengagement, i.e., incorporating both behavioral and emotional disengagement into one single construct.

Fourth, the sample of this study is from top universities in social-economic majors in Hanoi, Vietnam. A future attempt on this topic might overcome this limitation by including participants from various universities in different majors and trying to compare the different mechanisms leading to the disengagement of students.