Using self-determination and expectancy theory to evaluate hybrid learning

Abstract

This research uses self-determination theory and expectancy theory to investigate how Indonesian students responded to hybrid learning in the post-COVID-19 pandemic. This research specifically examines the impact of instructional support and peer support on academic performance, as well as the role of learning satisfaction as a mediator. Using structural equation modelling, the findings indicate that instructional support has a significant impact on learning satisfaction and academic performance. Indeed, peer support has a significant impact on learning satisfaction and academic performance. Furthermore, learning satisfaction acts as a significant mediator of these indirect effects of these variables on supporting academic performance among Indonesian students. This study also provides implications and contributions of using hybrid learning in promoting academic learning in the post-COVID-19 pandemic.

Keywords:

• Academic performance
• instructional support
• peer support
• learning satisfaction

Reviewing Editor:

• Sammy King Fai Hui, Curriculum & Instruction, The Education University of Hong Kong, Hong Kong

Subjects:

• Business, Management and Accounting
• Higher Education
• Open & Distance Education and eLearning
• Teachers & Teacher Education

Introduction

The post-COVID-19 pandemic has gradually returned to face-to-face teaching and learning. The university education system needs to accommodate this mandatory change using hybrid learning. Some prior studies documented that hybrid learning plays an essential role in education after the pandemic subsides, especially in higher education institutions, as it offers a way to combine the benefits of face-to-face teaching with the flexibility of online learning (Nambiar, 2020; Sulaiman et al., 2023). However, several studies noted some challenges in integrating face-to-face and online learning. For instance, some studies Mazzara et al., 2022; Raes, 2022) remarked that hybrid learning has the potential to make it difficult for educators to identify students’ problems that have consequences for providing appropriate learning support. Furthermore, hybrid learning relies on technological adoption; thus, some technical issues, such as internet connectivity problems and device compatibility issues, often disrupt both students and instructors (Li et al., 2021).

This challenge has led to many online courses provided to students that do not represent excellent online learning. Thus, further exploration of hybrid learning needs to be focused on the satisfaction assessment of students (Xiao et al., 2020). There are not many studies investigating students’ perceptions of using hybrid learning, primarily regarding students’ satisfaction with hybrid education in the post-COVID-19 period. Assessing hybrid learning by evaluating student satisfaction and achievement is essential either for students and their institutions or for the business industry that is the potential recruiter (El-Hilali et al., 2015). Students’ satisfaction is an integral part of university evaluation, as students become one of their stakeholders (Botek, 2013). Furthermore, the assessment is beneficial for improving the quality of learning that will affect the overall quality of life and can help teachers reflect on their teaching and empower them (Shek et al., 2022).

To explain this relationship, self-determination theory (SDT) can be involved, as it deals with the fulfillment of basic human needs consisting of autonomy, competence, and relatedness, which can create instrumental motivation and relatedness (Deci & Ryan, 2012; Ryan & Deci, 2000a). The level of satisfaction of basic needs determines the level of motivation obtained from controlled motivation to autonomous motivation (Hansen et al., 2021). From an empirical point of view, instrumental motivation, which is the inherent satisfaction of a person, produces positive behavior such as persistence, which in turn results in better performance (Reeve & Jang, 2006; Ryan & Deci, 2000a). Later, Van Nuland et al. (2012) revealed that only a few studies have captured the overall hierarchical SDT model, where it is still unclear how exactly the components in the model are related (e.g. full or partial mediation) and what the strengths and directions of the path are in the SDT model.

A preliminary study by Ryan and Deci (2020b) stated that basic psychological support can increase intrinsic motivation and internalization, which results in higher achievement. In connection with this, it becomes interesting when learning satisfaction, which is an instrumental form of motivation, is used as a mediation in this research to uncover the hierarchical SDT model, as stated by Van Nuland et al. (2012). Assessment also shows its important role in the review of expectancy-value theory (EVT). EVT modeled that a person’s achievement of performance is a function of the expected expectation of success and subjective task value (Schrader & Helmke, 2015). A prior study by Wang and Xue (2022) noted that further investigation into the change in the value of tasks in EVT construction is still needed, given that they have a prominent role in questioning the interrelationship with the expectation of success and influencing motivation and learning.

Furthermore, in connection with SDT, this theory provides support for the possible role of satisfaction mediation less described by SDT. The urgency of empirical studies on the position of satisfaction in EVT construction also clarifies the objective of exploring the relationship between learning support, learning satisfaction, and academic performance in hybrid learning post-COVID-19. This study provides some contributions. First, it can guide educators in designing hybrid learning experiences that align with SDT principles and expectancy theory which can further create effective learning environments. In addition, it can gain insights from applying these theories can inform the development of interventions and support systems that enhance students’ motivation, engagement, and performance in hybrid learning contexts.

The rest of the paper is provided below. The following section reviews the relevant literature, accompanied by the method and materials. Findings are provided and discussed in the next sections. The conclusion, implications, and future suggestions are provided in the last section.

Literature review

Learning support: instructional and peer support

Instructional and peer support are social engagement in instrumental form. Instructional support is defined as instructional guidance, dialogue, and course structure provided by the lecturer to the student, while peer support is the support of peers in the learning environment (Lee et al., 2011). Instructive guidance consists of answering students’ questions, correcting their misunderstandings, providing clear instructions, raising relevant resources, and promoting constructive feedback on their tasks and performance. As support, peers can help other students by answering questions, encouraging and motivating each other, and forming learning groups for the course. In this study, the definition of instructional support and peer support refers to the provision of job-related information and feedback, also called instrumental support (Reblin & Uchino, 2008). As instrumental, instructional, and peer support become resources to perform work effectively, one is more likely to experience a more satisfying work environment (McVeigh et al., 2019). Both conditions of such high support encourage (1) the exchange of advice, feedback, and guidance to perform clinical tasks and (2) the development of a strong emotional climate of empathy, trust, and mutual reinforcement that thus improves work performance, which is reflected in positive work evaluations, realized in the form of high job satisfaction (Orgambídez et al., 2022).

Learning satisfaction and academic performance

Several references have defined satisfaction in learning. For instance, Martin and Bolliger (2022) mentioned that student satisfaction is the fulfillment of student needs and the perception of satisfaction with the factors of learners, instructors, courses, programs, and organizations related to the learning environment. Furthermore, students’ satisfaction can be formed into several educational supports, such as facilities, the effectiveness of lecturers, and libraries (Kanwar & Sanjeeva, 2022). The scope of the definition of student satisfaction is quite broad, consisting of fifteen dimensions (Gruber et al., 2010). Among them are aspects of the atmosphere between students, courses, libraries, and how relevant teaching is to practice. In short, student satisfaction is related to college experience and values perceived and received while attending educational institutions (Magolda & Astin, 1993). Students’ satisfaction can be expressed as their feelings and attitudes toward the lectures (Lee et al., 2011; Topala & Tomozii, 2014).

Academic performance is also essential to build self-determination theory (SDT) hierarchically and understand the expectancy-value theory (EVT) model. Academic performance can also be equated with students’ achievement. It refers to the statement that work performance is a stage of achievement defined as the completion of work carried out by an individual within the organization. Some prior studies pointed out that academic performance is the result achieved by students in terms of academic experience (El-Hilali et al., 2015; Rajabalee & Santally, 2021). Performance in employment is defined as the level of productivity of an individual employee relative to his or her peers and work-related results (Trivellas et al., 2015). There are also those who define performance as a series of activities that contribute positively to the organization not as a result of an activity (Nemteanu & Dabija, 2021). Based on the description, the academic performance of this research is the result of hybrid learning.

Learning support on learning satisfaction

Self-determination theory (SDT) is widely used in educational research (Ryan & Deci, 2000a) and can explain student involvement from a need satisfaction perspective (Xia et al., 2022). This theory argues that adolescent relationships with teachers and peers provide information about their competence to succeed, how they organize their efforts, and their motivational orientation toward mastering or achieving performance goals (Diaconu-Gherasim et al., 2019). When external, social-interpersonal, and internal circumstances facilitate or support the fulfillment of individual needs for autonomy and competence, then intrinsic motivation increases (Legault, 2016). Further, Ryan and Deci (2000b) explained that instrumental motivation means the condition in which an individual performs a behavior for the inherent interest or satisfaction they find in the task. In sum, the support of teachers and peers in learning reflects support for the fulfillment of the need for competence, which then affects the instrumental motivation of learning satisfaction.

In hybrid learning, instructor support and guidance positively correlate to students’ satisfaction (Allen et al., 2002) and play an important role in learning satisfaction (Bolliger & Martindale, 2004). A well-designed course structure by a lecturer will increase student satisfaction with learning (Almaiah & Alyoussef, 2019). Teacher instructional support, by providing learning effectively and influencing students to do better (Gopal et al., 2021), indicates that the learning process or teacher support has led to student satisfaction (Ladyshewsky, 2013). In addition, the instructional guidance of Lee et al. (2011) can be construed as an activity to understand the needs of the learners. Furthermore, Kauffman (2015) called it to ensure student satisfaction. Besides, in the case of an employee, someone who has greater social support is more likely to increase their job satisfaction (Yuh & Choi, 2017).

A previous study by Martin and Bolliger (2022) found that instructor facilitation became a factor influencing online learning satisfaction. Furthermore, Lipka et al. (2019) explained that academic support increases learning ability, which in turn leads to student satisfaction. The faculty’s expertise and the classroom environment create a positive impact on college student satisfaction (Butt & Rehman, 2010). The researchers also revealed that teachers’ teaching behaviors that contribute to student satisfaction include interaction and relationships with students (Hill et al., 2003), helpfulness (DeShields et al., 2005), and feedback (Hill et al., 2003). In the study, the role of the instructor can also be analogous to that of the leader in the company, who can provide support to employees (Yuh & Choi, 2017). The study found that the support of the director had a positive influence on employee satisfaction. Thus, the first set of hypotheses is provided below.

H1a: Instructional support influences learning satisfaction

H1b: Peer support influences learning satisfaction

Learning support on academic performance

The basic relationship between learning support and academic performance is social exchange theory (SET). This theory defines how social interaction is determined by the benefits obtained from the exchange of services (Jahan & Kim, 2021). Good performance will be difficult without social support, but with social support, employees can work harder and better (Wulan, 2019). In detail, Cropanzano and Mitchell (2005) showed that there are four relationships in SET, including: Cell 1 (Match—Social Transaction in a Social Relationship), Cell 2 (Mismatch Economic Transactions in a Social Relationship), Cell 3 (Mismatch—Social Transaction in an Economic Relationship), and Cell 4 (Match—Economic Transaction in an Economic Relationship). Learning support and the academic performance of the student are included in cell 2, i.e. economic transactions in a social relationship. Similarly, Ahmad et al. (2022) found that a father who gives money to his son and does not ask for details belongs to the relationship category in cell 2. Learning support from both lecturers and peers received by students was exchanged for higher performance.

The prior explanation has revealed why learning support can also be seen as social support. Next, there is empirical evidence of the impact of social support on performance. Research results show that supervisory support affects employee performance (Msuya & Kumar, 2022). Another study by Saleem et al. (2022) also revealed that social support affects employee performance. Research in the field of education found that the support of teachers, peers, and parents positively influenced the remedial success of science students in the university placement test in Nigeria during the COVID-19 pandemic (Okwuduba et al., 2022). Furthermore, Ghaith (2002) found that teacher academic support alone had a positive influence on academic performance, while teacher personal support, peer academic backing, and peer personal support had no influence. Indeed, Jelas et al. (2016) showed that peer support directly influenced academic performance, but parent and teacher support did not. Although there are differences in findings about the influence of learning support on academic performance, the two conceptualists agree that learning support has a positive influence on academic performance. The next set of hypotheses in this research is provided below.

H2a: Instructional support influences academic performance

H2b: Peer support influences academic performance.

Learning satisfaction on academic performance

The basic theory for the relationship between learning satisfaction and academic performance is the expectancy-value theory (EVT). The theory suggests that people choose to perform because they are motivated by a combination of people’s expectations for success and the value of subjective tasks in a particular domain (Leaper, 2011). It further explains that the value of the task has four components, including: attainment value (e.g. importance of doing well), intrinsic value (e.g. personal enjoyment), utility value (e.g. perceived usefulness for future goals), and cost (e.g. competition with other goals and her colleagues). People who do not do their duty well are possible because they lower the duty to protect their own self-esteem (Rosenzweig et al., 2019). Indeed, Eccles and Wigfield (2020) described the influence of interest-enjoyment value on performance in the EVT model. In short, in the context of this research, according to EVT, student academic performance is driven by intrinsic value, that is, learning satisfaction.

Some empirical evidence suggests the influence of learning satisfaction on academic performance. For example, some studies (e.g. El-Hilali et al., 2015; Lee & Lee, 2008) adopted the theory of consumer behavior in the field of marketing. Furthermore, Rajabalee and Santally (2021) also showed a positive relationship between learning satisfaction and student’s performance in learning using online modules. In the field of employment, based on expectancy theory, it is proven that there is a positive influence of job satisfaction on performance (Eliyana & Ma’arif, 2019; Latifah et al., 2023). Such empirical evidence is similar to Trivellas et al. (2015), who mentioned that satisfaction with employee career success can affect job performance. Increased job satisfaction among employees will motivate them to achieve better results, to plan their work more thoroughly, and to become more efficient in the fulfillment of tasks (Nemteanu & Dabija, 2021). Other empirical evidence from scholars (e.g. Blahopoulou et al., 2022; Husein & Hanifah, 2019; Inayat & Jahanzeb Khan, 2021) explained this relationship. The next hypothesis is presented as follows:

H3: Learning satisfaction influences academic performance

Learning satisfaction as mediator

Learning satisfaction is potentially a mediator between learning support and academic performance. It is based on the hierarchical SDT model of Van Nuland et al., (2012) and the EVT model (Eccles & Wigfield, 2020; Rosenzweig et al., 2019). Based on the SDT hierarchies, the support learning received by students during hybrid learning shows fulfillment of their needs, so that their instructional motivation in the form of satisfaction increases (Legault, 2016). Furthermore, Ryan and Deci (2000b) explained that the subsequent satisfaction that you have generates positive behavior that, in turn, results in better performance (Reeve & Jang, 2006; Ryan & Deci, 2020a; Van Nuland et al., 2012). On the EVT side, student learning perceptions during hybrid learning from lecturers and peers form their attitude toward the experience received during learning. The attitude of students toward the experience they receive affects their value, that is, their satisfaction, and ultimately also affects their performance. Thus, the last set of hypotheses is presented as follows.

H4a: Learning satisfaction mediates instructional support and academic performance

H4b: Learning satisfaction mediates peer support and academic performance

Methodology

Research design

This study involved cross-sectional research design using a questionnaire on the impact of learning support and learning satisfaction on academic performance of business education students in Indonesia. In addition, the research design exists also to analyze the role of learning satisfaction mediation in the model shown by Figure 1. Variables in the study design are categorized as exogenous variables consisting of teacher support and peer support, mediator variables learning satisfaction, and endogenic variables academic performance.

Figure 1. Research design.

Data collection and participants

The data were collected from business education students who have experience following hybrid learning and are spread across five public universities in Indonesia using a simple random sampling technique. Simple random sampling enables respondents to have the same chance of being selected. The respondents in this study were business students, at least in their second semester of study, with the consideration that they have sufficient experience regarding hybrid learning. A self-administered survey with online questionnaires was administered to gather the data between February and May 2023. This survey distributed approximately 450 questionnaires, and 441 responses were returned. Each participant required approximately 15 minutes to complete the questionnaires. After removing incomplete and missing data from respondents, 423 final datasets were used for analysis.

Permission for the survey was granted by each university, and the respondents were asked to provide information voluntarily. The ethical clearance for this study was granted by the Ethical Committee of Universitas Negeri Malang. With regard to respondents’ characteristics, 72.34% were female students, and 27.66% were male students (see Table 1). Additionally, based on campus domicile, 87.47% of respondents ruled on Java Island, and 12.53% ruled on Sumatra Island. The datasets also showed that there were 36.91 respondents who were students in the sixth semester, 41.14% of respondents were students in the fourth semester, and 4.30% were students in the second semester. The respondents involved in this survey were capable of using learning applications for educational purposes.

Table 1. Characteristics of respondents.

Category	Description	Frequency	Percentage
Gender	Female	306	72.34%
	Male	117	27.66%
Grade	Semester 6	189	36.91%
	Semester 4	212	41.41%
	Semester 2	22	4.30%
Location	Sumatra Island	53	12.53%
	Java Island	370	87.47%
Skills in using learning application	Highly unskillful	1	0.24%
	Unskillful	1	0.24%
	Average	54	12.77%
	Skilled	238	56.26%
	Highly skilled	129	30.50%

Measurement and data analysis

Measurement items were provided on the basis of preliminary well-established papers and relevant literature. The adoption of the instrument was provided by contextualizing the reference instrument with the aim of surveying hybrid learning and converting the language into Indonesian. In detail, instructional support was measured using five items (e.g. during hybrid learning, the lecturer provides clear instructions for tasks and quizzes), which was adopted from Cocquyt et al. (2019). Furthermore, peer support was calculated using five items (e.g. during hybrid learning, there was an opportunity to study with my classmates) adopted from Lee et al. (2011). Later, learning satisfaction was estimated using instruments adopted from Lee et al. (2011), consisting of five items (e.g. I feel like I have gained a positive learning experience with hybrid learning). Lastly, students’ achievement was measured using five items (e.g. I expect a good score in hybrid learning) adopted from Lee and Lee (2008).

The process of adoption of instruments is carried out through discussion among researchers to evaluate the suitability of the format and words in each question item. All questionnaire items were measured using a five-point Likert scale, where 1 represents ‘strongly disagree’ and 5 represents ‘strongly agree’. Furthermore, the data were analyzed using PLS-SEM with Smart-PLS 3 to evaluate the measurement model and structural model to identify critical success factors and relationships among the hypotheses. In detail, first, a measurement model (outer model) was performed to test the validity and reliability of the model. Second, we provided a structural model (inner model) to estimate the internal consistency. Lastly, we conducted SEM to test the hypotheses (with direct and indirect effects) of the research model.

Results

Measurement model

Data processing using SEM-PLS techniques with Smart-PLS version 3. According to Hair et al. (2014), validity and reliability values are the stages of evaluation of a measurement model to provide an empirical measure of the relationship between indicator and construction (measurement model). First, the outer loading value of all items presented in Table 2 is higher than 0.5, and the average variance extracted (AVE) is upper than 0.5, remarking that these construction items have convergent validity (Hair et al., 2014). We removed PS1, PS2, and PS5 from PS construction and removed LS1 from LS construction because it has an outer loading of less than 0.5. Second, discriminatory validity is based on cross-loading values and square root values of AVE (√AVE). The structural items studied in the table also have discriminatory validity, where the outer loading value of a structure or variable is greater than cross-loading with another structure and (√AVE) is higher than its highest correlation with other structures (Hair et al., 2014). Third, Cronbach’s alpha value > 0.6 and composite reliability (CR) > 0.6 mean that existing structures have structural reliability (Hair et al., 2014). Thus, the research instrument is stated to be bold and reliable.

Table 2. Validity and construct reliability.

Construct	Item	Factor Loading	AVE	Cronbach’s Alpha	CR	Conclusion
AP	AP1	0.803	0.664	0.831	0.888	Reliable and Valid
	AP2	0.803
	AP3	0.858
	AP4	0.794
IS	IS1	0.798	0.647	0.863	0.901	Reliable and Valid
	IS2	0.809
	IS3	0.858
	IS4	0.808
	IS5	0.745
LS	LS1	0.789	0.635	0.809	0.874	Reliable and Valid
	LS2	0.799
	LS3	0.802
	LS4	0.796
PS	PS1	0.842	0.648	0.863	0.902	Reliable and Valid
	PS2	0.748
	PS3	0.864
	PS4	0.793
	PS5	0.772

Note. AP = academic performance, IS = instructional support, PS = peer support, LS = learning satisfaction, CR = construct reliability, AVE = average variance extracted.

The square root of the AVE of each construct should be greater than the correlation with any other construct to meet discriminant validity. As informed in Table 3, the model has good discriminant validity considering each loading value of a latent variable is greater than the other.

Table 3. Discriminant validity.

	AP	IS	LS	PS
Academic Performance (AP	0.815
Instructional Support (IS)	0.632	0.804
Learning Satisfaction (LS)	0.752	0.679	0.797
Peer support (PS)	0.686	0.790	0.704	0.805

In addition, we also performed fit measures before processing the structural model. In this paper, we involved Standardized Root Mean Squared Residual (SRMR), exact fit criteria like d_ULS, d_G, Chi-Square, and NFI. Table 4 shows that the values of SRMR (< 0.08), d_ULS and d_G (p-value > 0.05), and NFI close to 1, indicating that the model has met the criteria of model fit. Thus, we can further process the path analysis.

Table 4. Model fit measures.

	Model Saturated	Model Estimated
SRMR	0.077	0.077
d_ULS	1.376	1.376
d_G	0.517	0.517
Chi-Square	1252.108	1252.108
NFI	0.775	0.775

Structural model and hypothesis testing

Structural model evaluation is preceded by testing VIF values that indicate the existence of structural multicollinearity (Hair et al., 2014). VIF values (see Table 5) show that the VIF value is less than 5, so the structural model has no collinearity problems. As shown in Table 5, the researchers subsequently evaluated models that included examination of levels R², Q², and the statistical significance and relevance of the path coefficient (Hair et al., 2014). Determinant coefficient R², which measures the strength or contribution of exogenous variables to the endogenous variable is 0.639. As for the recommended R² value is more than 0.1 (Hair et al., 2014). Next, the Q² value shows the predictive relevance of endogenous construction or the accuracy value of the predictive model PLS-SEM (Msuya & Kumar, 2022; Ringle et al., 2017). Q2 on the structural model of this study is 0.663, exceeding the value of .0. As a result, the model has predictive significance (Hair et al., 2014).

Table 5. Statistical significance and relevance of the path coefficient.

Construct	AP	IS	LS	PS
Academic Performance (AP)
Instructional Support (IS)	0.069		0.326
Learning Satisfaction (LS)	0.516
Peer support (PS)	0.269		0.450

The next evaluation of the structural model (see Table 6 and Figure 2) is to test the hypothesis to see the relevance of the relationship between variables. H1a (IS → LS) has a p-value of 0.000 (β = 0.326, p-value < 0.05), meaning IS influenced on LS, H1b (PS → LS) has p-value of 0.000 (β = 0.450, p-value < 0.05), meaning PS influencing on LS. The results of the H1a, and H1b, tests show that H2a, H2b, and H3c are accepted. H2a (IS → AP) has a p-value of 0.214 (β = 0.169, p-value < 0.05), meaning IS having no influence on AP, H2b (PS → AP) has p-value of 0,000 (β = 0.269, p-value <0.05), meaning PS has an influence upon AP, indicating H2a and H2b are accepted. The next test is on H3, where the p-value is 0.000 (β = 0.516, p-value < 0.05), which means LS affects the AP and H3 is accepted. Finally, the mediation test by LS on H4a and H4b. The p-value for H4a (IS → LS → AP) is 0.000 (β = 0.168, p-value < 0.05), which means LS mediates IS against AP, H4-b (PS → LS →AP) 0.000 (β = 0.232, p-value < 0.05), which means LS mediates PS against AP.

Figure 2. Structural model.

Table 6. Hypothesis testing.

Hypothesis	Path	Path Coefficient	p-value	Decision
H1a	IS → LS	0.326	0.000	Accepted
H1b	PS → LS	0.450	0.000	Accepted
H2a	IS → AP	0.169	0.004	Accepted
H2b	PS → AP	0.269	0.000	Accepted
H3	LS → AP	0.516	0.000	Accepted
H4a	IS → LS → AP	0.168	0.000	Accepted
H4b	PS → LS → AP	0.232	0.000	Accepted
R^2		0.617
Q^2		0.541

Note. AP = academic performance, IS = instructional support, PS = peer support, LS = learning satisfaction.

Discussion

Testing has shown that instructional support has no impact on learning satisfaction or academic performance. This means whether or not the instructional guidance, dialogue, and course structure given by the lecturer to the student during hybrid learning do not affect changes in learning satisfaction or academic performance. This can happen because the proportion of online learning in hybrid learning is greater than in offline learning. The results are consistent with the statement of respondent representatives from each university that, after the pandemic, they are more likely to be engaged in online learning than offline learning. This condition can make it difficult for the lecturer to know the student’s problems, including providing appropriate learning support when it becomes important (Mazzara et al., 2022). Improving the student’s relationship with their teacher will not result in a gain in achievement, but that relationship must be close, positive, and supportive (Rimm-Kaufman & Sandilos, 2010). With these findings, instructional support is not a resource for students to perform their work effectively, thus enabling them to experience a more satisfactory working environment, as stated by McVeigh et al. (2019).

The research also found that peer support has an impact on learning and academic performance. What needs to be noted is that the negative path coefficient indicates the negative influence of peer support on learning satisfaction and academic performance. By looking at the peer-support construction indicators, we suspect that the presence of peers who are wondering about hybrid learning is actually judged to interfere with their understanding of the material submitted by the lecturer or group of presenters. There is a negative effect of peer support on this achievement, as found by Ghaith (2002) in the study of English as a Foreign Language (EFL) in Turkey. Groups that are constantly hostile, mostly not highlighting academics, will create frustration and have a negative impact on their drive to learn and academic achievement (Filade et al., 2019). Students are teenagers who need to be encouraged to choose peers and work to fight the negative impact of their friendships (Peza, 2015).

The findings about learning support for learning satisfaction and academic performance contradicted previous research. Some studies (e.g. Akhtar & Nazarudin, 2020; Orgambídez et al., 2022; Yuh & Choi, 2017) have previously identified the impact of social support on employee satisfaction and employee performance. Someone with greater social support is more likely to increase their job satisfaction (Yuh & Choi, 2017) and work better (Wulan, 2019). Previous research revealed that academic support and lecturer expertise had a positive impact on student satisfaction and their achievement (Lipka et al., 2019; Okwuduba et al., 2022). In online learning, instructional support has been found to influence student learning satisfaction (Almaiah & Alyoussef, 2019; Bolliger & Martindale, 2004).

Nevertheless, this study supports some prior papers (e.g. Gunzenhauser et al., 2021; Sawitri et al., 2018) in explaining this relationship. In more detail, Gunzenhauser et al. (2021) found that primary school teachers’ support for online learning had no impact on student reading, arithmetic, or math word problems. Furthermore, Sawitri et al. (2018) remarked that superior support had no influence on employee performance, and this was presumed due to the complexity of the remuneration system, which could not be well explained by the superior. The findings also reinforce research recommendations (Imran et al., 2023) about the importance of exploring why college and postgraduate students choose face-to-face learning over hybrid or online learning. During a pandemic, online courses offered may not represent excellent online learning, but rather serve as a temporary solution (Crew & Märtins, 2023).

The next discussion shifted to the influence of learning satisfaction on academic performance. The satisfaction experienced by students during hybrid learning affects their academic performance. Increased student learning satisfaction will motivate them to achieve better results, to plan their work more thoroughly, and to become more efficient in fulfilling tasks (Nemteanu & Dabija, 2021). These findings support some studies (e.g. El-Hilali et al., 2015; Lee & Lee, 2008; Rajabalee & Santally, 2021) showing that there is a positive relationship between learning satisfaction and student performance. In the field of employment, based on the expectancy theory, some studies (e.g. Blahopoulou et al., 2022; Davar & RanjuBala, 2012; Eliyana & Ma’arif, 2019; Husein & Hanifah, 2019; Inayat & Jahanzeb Khan, 2021; Latifah et al., 2023) remarked that employee satisfaction can affect job performance.

The latest findings show the role of learning satisfaction mediation in the relationship between learning support and academic performance. Learning satisfaction does not mediate the relationship of instructional support with academic performance. This means that learning satisfaction cannot affect academic performance, even though support learning has changed up or down). On the contrary, learning satisfaction mediates peer support for academic performance. In other words, learning satisfaction can affect academic performance as there are changes in peer support. The path coefficient on mediation path H4b is smaller than the path coefficient in H2b. This explains that the mediation effect maximizes the negative impact of peer support, thereby enhancing student academic performance.

The final discussion should be directed at the context of the SDT and EVT. On the H4a track, this research was unable to show that the fulfillment of the need for competence caused the student’s instrumental motivation in the form of increased satisfaction, thus affecting their academic performance. On the contrary, the H4b track showed that the degradation of the need for competence through peer support can be minimized by learning satisfaction and ultimately affect academic performance. In other words, the presumption of mediation according to the SDT hierarchies, as expected by Van Nuland et al. (2012), is proved in line H4b. In the EVT approach, the findings above H2a show that task values are not able to mediate between student affective memories of their achievements, but task values can partially mediate the relationship of a student’s affective memory (in the form of peer support) to academic performance as modeled by the previous studies (Eccles & Wigfield, 2020; Rosenzweig et al., 2019).

The complexity of these research findings has important implications for future practitioners (teachers, lecturers, and colleagues) and researchers. The findings on the relationship between learning support, learning satisfaction, and academic performance need attention from educational practitioners and researchers. They need to explore how pedagogical approaches have shifted and adapted to online and hybrid learning, including applying the principles of e-pedagogy (Simuth & Sarmany-Schuller, 2012) at the university level. Teachers and students both need guidance on how to interact effectively in online courses and other environments (Blaine, 2019). A lecturer in a hybrid learning environment becomes a facilitator, mediator, mentor, or coach (Klimova & Kacetl, 2015). Exploring appropriate forms of support is also needed, as stated by Ang et al. (2021), that hybrid learning requires greater instructional support. For example, lecturers need to frequently ask oral questions and provide training in class throughout lessons, as well as establish rules for class participation for students (Li et al., 2023). College management practitioners need to realize learning satisfaction, as this can improve their student’s academic performance as well as be mediators in peer support relationships with academic performance. Future researchers will also need to conduct longitudinal studies to establish the link between learning support, learning satisfaction, and academic performance, given that the research is only exploring it with cross-sectional data approaches.

Implications

The results of the research have implications for practitioners and academics alike. The results showed that learning satisfaction partially mediates peer support with academic performance, but learning satisfaction does not mediate instructional support with academic performance. It suggests that there is indeed a possible role for mediation by internal motivation in the form of a chapter on the relationship between the fulfillment of basic human needs and performance or achievement. The results also showed that affective memories in the form of peer support influenced task value and learning satisfaction. The output of learning satisfaction testing on academic performance is not different from the results of previous research studies, while other paths are quite demanding attention because of their uniqueness. To that end, educators are advised to enhance the application of e-pedagogy principles at the university level, develop guidance to conduct effective interactions in online courses, and explore forms of support suitable for hybrid learning. University management practitioners need to realize student learning satisfaction, while future researchers also need to undertake longitudinal studies to ensure the path and relationship of learning support, learning satisfaction, and academic performance in the case of hybrid learning and other forms of learning.

Conclusion

The discussion of the results indicates that instructional support and peer support have a significant influence on learning satisfaction and academic performance. This study also shows that learning satisfaction plays a prominent role in mediating the relationship between instructional support and academic performance, as well as peer support and academic performance. This indicates that the development of these hierarchies (Van Nuland et al., 2012) requires further study of the possibility of internal motivation, i.e. satisfaction being the mediation between the fulfillment of basic human needs and performance or achievement. On the other hand, the development of the expectancy-value theory model requires further investigation of changes in the value of tasks. Learning satisfaction tested its role as a mediator between learning support and academic performance, where learning support, from the perspective of self-determination theory, is assumed to be the fulfillment of human needs for competence. In addition, from the perspective of the expectancy-value theory, this study places learning satisfaction as a form of task value whose existence is influenced by affective memories, that is, learning support received by students.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Wening Patmi Rahayu

Wening Patmi Rahayu is a professor of entrepreneurship education and SMEs at the Faculty of Economics and Business, Universitas Negeri Malang, Indonesia. She is concerned with research regarding educational studies, entrepreneurship education, and small-medium enterprises.

Raya Sulistyowati

Raya Sulistyowati is an assistant professor in entrepreneurship education at the Faculty of Economics and Business, Universitas Negeri Surabaya, Indonesia.

Heri Pratikto

Heri Pratikto is a professor at the Faculty of Economics and Business at Universitas Negeri Malang.

Rachmad Hidayat

Rachmad Hidayat is an assistant professor at the Faculty of Economics and Business, Universitas Negeri Malang.

Bagus Shandy Narmaditya

Bagus Shandy Narmaditya is an assistant professor in economic education at the Faculty of Economics and Business, Universitas Negeri Malang, Indonesia.

Zamzani Zainuddin

Zamzani Zainuddin is an associate professor at the Faculty of Education, Universiti Malaya, Malaysia.

Siti Zumroh

Siti Zumroh is a student at the Faculty of Economics and Business, Universitas Negeri Malang, Indonesia.

Rila Ayu Agnes Indarwati

Rila Ayu Agnes Indarwati is a student at the Faculty of Economics and Business, Universitas Negeri Malang, Indonesia.