Young Learners’ Perceptions of How Their Teachers, Peers, and Parents Impact Their EFL Motivation and Task Engagement: A SEM Study

Abstract

This study examined how young learners’ (YLs’) views of teachers’, peers’, and parents’ roles influence their motivation and task engagement in learning English, as well as their parents’ perspectives on their children’s motivation and task engagement, using a quantitative cross-sectional design. The study draws on self-determination theory, social cognitive theory, YLs’ language learning motivation, and task engagement. Data from surveys were analyzed using structural equation modeling with IBM SPSS AMOS Version 26. The sample included 598 YLs and their parents from Indonesian elementary schools. The model examines direct and indirect effects of four independent variables (YLs’ and parents’ perceptions of teachers’, peers’, and parents’ roles) on a dependent variable (task engagement) with a mediator (motivation). The model fit was adequate (Chi-Square = 67.452, CMIN/DF = 3.895, RMSEA = 0.07, RMR = 0.014, CFI = 0.934, and TLI = 0.976). Both children’s and parents’ perceptions positively influenced children’s motivation and engagement. Motivation significantly influenced task engagement and mediated the impact of children’s and parents’ views on it. The findings recommend engaging parents, encouraging peer collaboration, and training teachers to build a supportive environment for young English learners.

1. Introduction

Motivation and task engagement are important factors in young learners’ (YLs’) additional language learning. Teachers, peers, and parents are social factors (Kos, 2023; Qureshi et al., 2023) that play prominent roles in YLs’ motivation and engagement (Mihaljević Djigunović & Nikolov, 2019; Nikolov & Mihaljević Djigunović, 2019). In the Indonesian context, characterized by collective cultural values that prioritize family participation and social harmony (Deci & Ryan, 2000), comprehending the impact of these relationships on motivation and engagement is crucial for devising successful educational strategies. Earlier studies have shown that positive relationships with teachers and peers can enhance engagement in classroom activities (Gan, 2021; Li et al., 2024). Parental behaviors, including encouragement and active involvement in educational activities, can boost children’s motivation (Choi et al., 2024; Tanaka & Takeuchi, 2024).

Despite these findings, empirical evidence remains sparse regarding how these factors impact YLs’ task engagement. This study aims to address this gap by using structural equation modeling (SEM) to explore the complex interrelationships among the roles teachers, peers, and parents play in children’s motivation and task engagement in Indonesian primary schools. The study seeks to offer new insights for parents, teachers and policymakers into how fostering language learning environments can enhance student motivation and encourage persistent task engagement.

This research examines the relationships among YLs’ perceptions of the roles their teachers, parents, and peers play in their motivation to learn English in primary schools in Indonesia. Additionally, we investigate the association between parents’ perceptions of their role in shaping their children’s motivated behavior. More specifically, we attempt to determine the extent to which YLs’ motivation serves as a predictor of their task engagement in English learning. We examine whether the perceptions of YLs and their parents exert a direct influence on the children’s engagement with English tasks. Finally, we investigate YLs’ motivation as a mediator between their own and their parents’ perceptions of task engagement in learning English.

This study seeks to answer seven research questions:

• How are YLs’ perceptions of their teachers’, parents’, and peers’ roles directly associated with their motivation to learn English?
• How are parents’ perceptions of their role directly associated with YLs’ motivation to learn English?
• How does YLs’ motivation to learn English predict their engagement with English language learning tasks?
• To what extent do YLs’ perceptions of their teachers’, parents’, and peers’ roles directly predict their task engagement?
• To what extent do parents’ perceptions of their role directly predict YLs’ task engagement?
• How does YLs’ motivation to learn English mediate the relationships between their perceptions of their teachers’, parents’, and peers’ roles and their task engagement?
• How does YLs’ motivation to learn English mediate the relationships between parents’ perceptions of their role and their task engagement?

2. Literature Review

2.1. Theoretical Framework

This study relies on two interconnected theoretical frameworks to clarify the influence of teachers, peers, and parents on young students’ motivation and task engagement: self-determination theory (SDT) and social cognitive theory. SDT highlights the importance of satisfying three essential psychological needs, relatedness, competence, and autonomy, to boost motivation and engagement (Ryan & Deci, 2020). In primary education, the encouragement and support of teachers and parents, along with the acceptance of peers, fulfill children’s need for forming relatedness. Encouragement through challenges, guidance, and feedback boosts their sense of competence. When children view their learning activities as self-driven and valued by important figures in their lives, their engagement and commitment increase, and they become gradually autonomous over time. This framework underscores how environments created by educators, family, and friends can either enhance or diminish student motivation.

SDT encompasses a range of regulation types; however, in this study we distinguish between intrinsic and extrinsic motivation (Ryan & Deci, 2020) in a way that is both deliberate and theoretically justified. Primarily, given the participants’ developmental stage, simplifying the distinction was essential for maintaining clarity and ensuring age-appropriate conceptualization.

Additionally, the study was conducted in a specific Indonesian context characterized by a collectivist cultural character (Hofstede, 2011). However, within SDT, autonomy is not synonymous with individualism, but it concerns the sense of volition (Ryan & Deci, 2020) which can exist alongside collectivist values. Therefore, we assume that autonomy-supportive techniques are compatible with interdependent cultural contexts. Although a more detailed cultural analysis could broaden the discussion, the aim of the current study was to highlight the influence of teachers, peers, and parents on motivation, rather than to offer a cross-cultural reinterpretation of SDT.

Social cognitive theory emphasizes the significance of observational learning, self-efficacy, and the reciprocal interactions between individuals and their environment (Bandura, 1986). Children tend to adopt behaviors and attitudes they observe in their teachers, parents and peers. For example, when a teacher confidently uses English in daily routines, YLs can also use English naturally. Watching enthusiastic classmates can boost learners’ confidence and motivation. Similarly, parents’ positive attitudes toward education can influence their children’s beliefs and practices. This framework shows how societal influences and personal factors shape YLs’ behavior, motivation, and engagement with tasks.

Children’s motivation is dynamic and closely linked to contextual factors (Mihaljević Djigunović & Nikolov, 2019). YLs’ motivated learning behavior interacts with their teachers’ motivation and develops based on their perceptions of their parents’, teachers’, peers’ motives, beliefs, and expectations, and other internal and external classroom factors. Moreover, the older the learner, the less influence teachers and parents have on their motivation, but the greater the influence of their peers is (Mihaljević Djigunović & Nikolov, 2019). These points are considered in the next section.

2.2. The Role of Social Factors in YLs’ Motivation and Engagement

Significant others, including teachers, peers, and parents, play a crucial role in YLs’ motivation to learn an additional language and to engagement with specific tasks (Mihaljević Djigunović & Nikolov, 2019). Learners’ attitudes and perceptions of their educators impact their level of motivation (Hiver & Al-Hoorie, 2020). A few studies have been conducted on learners’ perceptions of their teachers’ and their peers’ motivation (Tseng, 2021; Wallace & Leong, 2020), and task engagement (Lu et al., 2022). Overall, YLs’ perceptions of their teachers are positively associated with their level of motivation (Wallace & Leong, 2020). More specifically, learners’ negative perceptions of their teachers may have detrimental impacts on their motivation, as demonstrated by instances where a student expressed unfavorable opinions regarding their teacher’s literacy practices, consequently influencing their motivation to learn the English language (Tseng, 2021). Moreover, learners’ perceptions of their educators can affect their cognitive and emotional task engagement in the learning process (Lu et al., 2022). The level of involvement perceived by learners about their teachers significantly influences their engagement mediated by motivation (Núñez et al., 2019). Teachers serve crucial role models in the English learning process. Thus, teachers’ roles in the literature include being role models and engagement facilitators, and scaffolding learning. Consequently, learners’ motivation to learn English is influenced by their interactions with teachers (Liu & Chiang, 2019).

Peers also play a central role in shaping children’s motivation and task engagement. In accordance with social learning theories, observing peers exhibiting enthusiasm and perseverance can inspire similar behaviors in fellow learners (Bandura, 1986). Interaction among peers can enhance not only motivation and confidence, but it can also foster engagement in authentic communication (Bui & Dao, 2023). Additionally, it has been shown that the way YLs perceive their classmates can influence their motivation to learn English (Wallace & Leong, 2020). Not only in adult learners, but also in YLs’ classrooms, peer support can have a significant impact on learners’ motivation (Solhi, 2024) and their emotional and behavioral engagement (Ansong et al., 2017). Furthermore, some peers can serve as role models, competitors, collaborators, and help scaffold one another’s learning (Mihaljević Djigunović & Nikolov, 2019).

Parental involvement is recognized as an essential factor influencing children’s academic motivation and task engagement. It includes an array of activities, such as assisting with homework (Núñez et al., 2019), engaging in communication with teachers (Fitriani & Anam, 2024), and taking part in school-related events (Fan & Williams, 2010; Rivera & Li, 2019). Research consistently shows that when parents are actively involved, children tend to have positive attitudes toward learning and develop higher levels of persistence (Rivera & Li, 2019). Parents’ support predicts students’ task engagement (Ansong et al., 2017). Students’ perceptions of parental involvement in their homework significantly impacts their engagement with assigned tasks mediated by motivation (Núñez et al., 2019).

Parental encouragement and support can reinforce children’s motivation and task engagement. YLs’ perception of their parents’ expectations, such as to achieve a specified level in their exams, impacts their motivation. For example, Wallace and Leong (2020) found that a cultural focus on extrinsic incentives for learning with instrumental objectives may be instilled at an early age and nurtured through parental influence. Furthermore, respecting children’s learning styles, providing educational information, and offering emotional support can also enhance student motivation, whereas parental pressure for high academic outcomes may diminish intrinsic motivation (Choi et al., 2024). Multiple studies found that parents’ involvement in their children’s study of English can significantly boost their motivation (e.g., Butler, 2015; Nie & Mavrou, 2025; Sumanti & Muljani, 2021).

In summary, parental roles in children’s English learning include their ongoing involvement (Núñez et al., 2019) and expectations (Butler, 2015; Butler & Le, 2018; Wallace & Leong, 2020), and ensuring children’s access to resources, to enrichment activities (Butler, 2015; Butler & Le, 2018), and to extracurricular opportunities (Torrecilla & Hernández-Castilla, 2020).

2.3. Motivation and Task Engagement

Frameworks interpret motivation and engagement distinctively. Motivation is the force initiating, directing, and sustaining goal-oriented behaviors (Dörnyei & Ottó, 1998), whereas engagement involves the active participation reflected in behavioral, emotional, cognitive, and social aspects of learning (Philp & Duchesne, 2016). Based on SDT, motivation can be intrinsic, deriving from interest and enjoyment, or extrinsic, driven by external rewards or pressures (Ryan & Deci, 2020). YLs’ motives are sometimes driven by their parents’ expectations (e.g., they want YLs to get better grades in English). This aligns with the ought-to-self in the second language (L2) motivational self-system model (Dörnyei, 2005) which also includes rewards or positive feedback from parents and teachers. Moreover, YLs’ motivation is dynamic, it changes all the time, and it is closely linked to contextual factors (Mihaljević Djigunović & Nikolov, 2019).

Previous studies examining determinants of task engagement have highlighted how teachers, parents, peers, and motivation shape it (Dörnyei & Ushioda, 2011). Constructive relationships with educators and peers are significantly associated with cultivating autonomous motivation, including personal goals, and enhanced task engagement, both of which are critical for academic achievement. In contrast, parental relationships, although more indirect, are also influential in fostering motivation and task engagement within educational settings (Collie et al., 2016). For example, YLs who had contentious interactions with their parents and teachers showed decreased perseverance in tasks, whereas those with relationships marked by closeness and support experienced increased motivation and enhanced engagement with tasks (Silinskas & Kikas, 2022). Understanding these factors within particular cultural settings, such as Indonesia, is crucial for scaffolding the process of English language learning.

In sum, significant positive correlations were found between task engagement and motivation in studies involving children and adolescents learning EFL (Mihaljević Djigunović & Nikolov, 2019; Zhang & Crawford, 2024). Both extrinsic and intrinsic motivational factors influence YLs’ EFL learning motivation and task engagement during face-to-face English instructional sessions (Jiao, 2024). As has been discussed, multiple variables interact and impact outcomes. These are examined in a proposed model, presented in the next section.

3. Method

3.1. Research Design

This research used a quantitative cross-sectional design to investigate the relationships among factors, related to EFL teachers, parents, and peers, impacting YLs’ motivation and task engagement in Indonesian primary schools. Data was gathered at one specific moment to reflect YLs’, their teachers’, and parents’ perceptions (Creswell & Creswell, 2018). We used structural equation modeling (SEM) to evaluate the proposed model, allowing for simultaneous examination of both direct and indirect interactions among latent variables (Kline, 2023). These included YLs’ perceptions of their teachers, peers, and parents’ views as well as their parents’ perception of their role (independent variables), plus YLs’ motivation (mediator variable) and task engagement (dependent variable).

3.2. Participants

Participants included 598 fifth graders and their 598 parents at 15 public and private primary schools in Indonesia. They were chosen through a random sampling technique. We included one of the YLs’ parents; it was up to them to decide if the mother or father participated in the study. The description of participants is presented in

Table 1. Description of participants.

	Variables	Descriptor	N	Percentage (%)
YLs	Gender	Male	303	50.67
		Female	295	49.33
		Total	598	100
Parents	Gender	Male	528	88.29
		Female	70	11.71
		Total	598	100
	Parents’ education	Senior high	261	43.65
		Diploma degree	41	6.85
		Bachelor’s degree	215	35.96
		Master’s degree	64	10.70
		Doctoral degree	17	2.84
	Parents’ occupation	Housewife	254	42.47
		Laborer	41	6.85
		Trade/Entrepreneur	58	9.70
		Private employee	61	10.20
		Public employee	184	30.77
	Parents’ income	IDR 1–5 million	374	62.54
		IDR 6–10 million	94	15.72
		IDR 11–15 million	76	12.71
		IDR > 15 million	54	9.03

Note: parents’ education is the highest level of formal education that parents achieved; parents’ occupation is the the job, profession that a parent does to earn a living; parents’ income is the total monthly earnings that parents receive from their job.

.

3.3. Instruments

We used two instruments for data collection. The YLs’ questionnaire (42 items) measured five latent variables: children’s perception of their teachers as role models (3 items), engagement facilitators (3 items), scaffolders (3 items); YLs’ perception of their peers as competitors (3 items) and collaborators (3 items), and their perception of parents. The latter concerned parental involvement/expectation (4 items), access to resources (4 items). Additional items comprised YLs’ motivation including intrinsic motivation (3 items), extrinsic motivation (3 items), and finally, task engagement. This part of the instrument included: behavioral (4 items), emotional (3 items), cognitive (3 items), and social (3 items).

The parents’ questionnaire comprised 19 items tapping parents’ perception of their roles across five dimensions: parental involvement (4 items), parental expectation (5 items), access to resources (5 items), enrichment activities (3 items), and extracurricular activities (2 items) (see

Table 2. Instrument description.

Instrument	Laten Variable	Observed Variable	No of Items	Total
YLs’ questionnaire	YLs’ perception of teachers (SPT)	Role model (RM)	1, 2, 3	3
		Engagement facilitator (EF)	4, 5, 6	3
		Scaffolding (S)	7, 8, 9	3
	YLs’ perception of peers (SPPr)	Competitor (C)	10, 11, 12	3
		Collaborator/Scaffolding (C/S)	13, 14, 15	3
	YLs’ perception of parents (SPPt)	Parental involvement and expectation (PI/PE)	16, 17, 18, 19	4
		Access to resources (AR)	20, 21, 22, 23	4
	Motivation (M)	Intrinsic (I)	24, 25, 26	3
		Extrinsic (E)	27, 28, 29	3
	Task engagement (TE)	Behavioral (BE)	30, 31, 32, 33	4
		Emotional (EE)	34, 35, 36	3
		Cognitive (CE)	37, 38, 39	3
		Social (SE)	40, 41, 42	3
			Total	42
Parents’ questionnaire	Parents’ perception of their role (PPR)	Parental involvement (PI)	1, 2, 3, 4	4
		Parental expectation (PE)	5, 6, 7, 8, 9	5
		Access to resources (PAR)	10, 11, 12, 13, 14	5
		Enrichment activities (EA)	15, 16, 17	3
		Extracurricular activities (ExA)	18, 19	2
			Total	19

).

These two instruments were developed based on the literature related to the roles of teachers, peers, and parents (e.g., Fan & Williams, 2010; Mihaljević Djigunović & Nikolov, 2019; Pinter, 2017), language learning motivation (Mihaljević Djigunović & Nikolov, 2019), intrinsic and extrinsic motivation (Ryan & Deci, 2020), and task engagement (Philp & Duchesne, 2016). They were piloted with 371 and 579 students and 270 parents of fifth graders out of this sample, confirming their validity and reliability. The model evaluating students’ perceptions demonstrated a good fit, indicated by indices (CFI = 0.996, TLI = 0.980, SRMR = 0.014, RMSEA = 0.049). Similarly, the motivation model exhibited a good fit, as evidenced by indices (CFI = 0.994, TLI = 0.989, SRMR = 0.021, RMSEA = 0.042). Moreover, the task engagement model had a good fit, it was characterized by fit indices (CFI = 0.944, TLI = 0.926, SRMR = 0.039, RMSEA = 0.079). The parents’ questionnaire indicated that the model fit was adequate (CFI = 0.945, TLI = 0.934, SRMR = 0.045, RMSEA = 0.059). The average variance extracted (AVE) values of the questionnaires exceeding 0.50 satisfy the requirements for convergent validity. Additionally, the composite reliability (CR) values exceeding 0.70 affirm the internal consistency reliability. The two questionnaires can be seen in Appendix A and Appendix B.

3.4. Data Collection Procedure

First, ethical approval was acquired from the IRB Doctoral School of Education, University of Szeged (number 24/2023). Then, signed consent forms were obtained from parents before they filled out the surveys online via Google Forms from April to June 2025. The online questionnaires were distributed to participants with the help of English teachers using WhatsApp version 2.25.22.83. All participants were informed that their involvement was voluntary, and their information was confidential.

3.5. Data Analysis

We used SEM for data analysis with IBM SPSS AMOS Version 26. The model was specified with six latent variables representing YLs’ perceptions of teachers’, peers’, and parents’ roles and their parents’ perceptions of their role, YLs’ motivation, and task engagement. All were measured through their respective indicators. Model evaluation involved assessing the measurement model for reliability and validity, including CR, AVE, and discriminant validity. Then we examined the structural model by analyzing path coefficients, and tested mediating effects to explore the complex relationships within the model; finally, we calculated squared multiple correlations (R²) for endogenous variables (motivation and task engagement).

3.6. Proposed Model

We hypothesized that the proposed model showed the relationships among YLs’ perceptions of their teachers’, parents’, and peers’ roles, their parents’ perceptions of their role, YLs’ motivation to learn English, and their task engagement. In this model, we estimated all proposed paths, including direct and indirect associations, simultaneously to capture the potential influence of each variable on YLs’ motivation and task engagement. The proposed model incorporated correlations among the exogenous latent constructs to account for shared variance across perceptions of social support (see Figure 1). Model estimation was conducted using the maximum likelihood method (Kline, 2023) in AMOS (Arbuckle, 2017). We assessed the overall fit of the proposed model using multiple fit indices to determine whether the specified paths adequately represented the observed data prior to any model specification.

4. Results

4.1. Prerequisite Analysis

Prior to executing the primary data analysis, we conducted a normality test and a multicollinearity test (Hair et al., 2018; Kline, 2023). An examination of normality reveals that, on a univariate level, each measurement indicator within this study adheres to a normal distribution, as indicated by the critical ratio value being less than 2.58 or exceeding −2.58. Furthermore, from a multivariate perspective, the critical ratio value of 1.947, which is below the threshold of 2.58, further confirms that the data conforms to a normal distribution (Hair et al., 2018; Kline, 2023).

Inner model multicollinearity can be seen based on the correlation value between exogenous variables.

Table 3. Correlation matrix for inner model.

	SPPt	SPPr	PPR	SPT	M	TE
SPPt	1.000
SPPr	0.596	1.000
PPR	0.208	0.238	1.000
SPT	0.517	0.577	0.209	1.000
M	0.602	0.689	0.578	0.632	1.000
TE	0.406	0.407	0.297	0.321	0.672	1.000

presents the correlations among the latent constructs. Motivation (M) shows a strong correlation with YLs’ perception of peers (SPPr) (0.689), YLs’ perception of teachers (SPT) (632), YLs’ perception of parents (SPPt) (0.602), and parents’ perception of their role (PPR) (0.578), highlighting its pivotal position in the model. Task engagement (TE) correlates strongly with motivation (M) (r = 0.672) and maintains moderate connections with YLs’ perception of teachers (SPT) (0.321), parents’ perception of their role (PPR) (0.297), YLs’ views of parents (SPPt) (0.406), and YLs’ perception of peers (SPPr) (0.407). A strong correlation exists between YLs’ views of peers (SPPr) and YLs’ perception of teachers (SPT) (0.577), and a substantial relationship is also observed between children’s perception of parents (SPPt) and of peers (SPPr) (r = 0.596). In contrast, parents’ perception of their role (PPR) exhibits relatively weaker correlations with YLs’ perception of teachers (SPT) (r = 0.209), parents (SPPt) (0.208), and peers (SPPr) (0.238).

Overall, these results indicate that motivation is the most interconnected construct, whereas parents’ self-perception shows a weaker alignment with YLs’ perspectives on their educational setting. The results showed that no correlation between exogenous latent constructs is greater than 0.9 or lower than −0.9; therefore, in this model, there is no multicollinearity at the latent construct level or inner model level (Kline, 2023).

Outer model multicollinearity can be seen based on the correlation value between indicators, especially between exogenous variable indicators.

Table 4. Correlation matrix for outer model.

	SE	CE	EE	BE	E	I	PIPE	AR	C	CS	ExA	EA	PI	PE	PAR	RM	EF	S
SE	1.00
CE	0.534	1.00
EE	0.505	0.653	1.00
BE	0.528	0.678	0.546	1.00
E	0.462	0.498	0.475	0.493	1.00
I	0.593	0.640	0.510	0.534	0.782	1.00
PIPE	0.490	0.529	0.504	0.524	0.441	0.567	1.00
AR	0.521	0.562	0.536	0.556	0.469	0.602	0.528	1.00
C	0.405	0.438	0.417	0.433	0.250	0.321	0.414	0.440	1.00
CS	0.363	0.392	0.373	0.388	0.224	0.288	0.371	0.394	0.384	1.00
ExA	0.112	0.121	0.115	0.120	0.068	0.088	0.109	0.116	0.116	0.104	1.00
EA	0.117	0.127	0.121	0.125	0.072	0.092	0.114	0.121	0.121	0.109	0.575	1.00
PI	0.105	0.113	0.108	0.112	0.064	0.082	0.102	0.108	0.109	0.097	0.515	0.539	1.00
PE	0.068	0.074	0.070	0.073	0.042	0.054	0.067	0.071	0.071	0.063	0.336	0.352	0.314	1.00
PAR	0.109	0.118	0.113	0.117	0.067	0.086	0.106	0.113	0.113	0.101	0.536	0.562	0.502	0.328	1.00
RM	0.386	0.417	0.397	0.412	0.302	0.388	0.397	0.421	0.432	0.387	0.095	0.099	0.089	0.058	0.092	1.00
EF	0.411	0.443	0.423	0.439	0.321	0.413	0.422	0.448	0.460	0.412	0.101	0.106	0.094	0.062	0.098	0.400	1.00
S	0.373	0.402	0.383	0.398	0.292	0.375	0.383	0.407	0.418	0.374	0.091	0.096	0.086	0.056	0.089	0.363	0.386	1.00

presents the correlations among the indicator levels. For example, motivation indicators, extrinsic (E) and intrinsic (I), had a strong correlation (0.782). Parents’ perception of their role indicator, such as extracurricular activities (ExA) had a very weak correlation with most indicators. However, the results indicated that no correlation between indicators is higher than 0.9 or lower than −0.9; therefore, there is no multicollinearity at the indicator level or outer model level (Kline, 2023). This implies that the indicators were adequately distinct, and none had high correlation to cause estimation or identification issues within the SEM.

4.2. Measurement Model Assessment

In the assessment of the measurement model, we evaluated systematically the validity and reliability using factor loadings, composite reliability and average variance extracted. Furthermore, we examined the model for potential Heywood cases (Byrne, 2013; Kline, 2023) and analyzed the correlations among latent variables to evaluate discriminant validity.

Table 5. The AVE and CR values of constructs.

Constructs	AVE	CR
SPT	0.518	0.763
SPPr	0.595	0.746
SPPt	0.558	0.716
PPR	0.516	0.842
M	0.572	0.728
TE	0.545	0.861

presents the AVE and CR values for each construct, signifying that the measurement model demonstrates acceptable convergent validity and internal consistency reliability. All constructs exhibit AVE values above the suggested cutoff of 0.50, indicating that over 50% of the variance in the indicators is accounted for by the latent constructs rather than error (Fornell & Larcker, 1981). For instance, the YLs’ perception of peers (SPPr) AVE value stood at 0.595, indicating that a significant 59.5% of the variance in the observed variables was explained by the latent construct, thus demonstrating strong convergent validity. Conversely, the parents’ perception of their role (PPR) showed a lower AVE value of 0.516, suggesting that while it maintains adequate convergent validity, its indicators represent the construct with slightly less intensity compared to constructs associated with higher AVE values. Furthermore, all CR values exceed the minimum recommended cutoff of 0.70, indicating adequate reliability and internal consistency (Hair et al., 2018).

In summary, these outcomes confirm that YLs’ perceptions of their teachers’ role (SPT), parents’ role (SPPt), peers’ role (SPPr), as well as their parents’ perceptions of their role (PPR), and motivation (M), and task engagement (TE) were assessed both reliably and validly.

Table 6. Variances.

	Estimate	S.E.	c.r.	p
SPT	0.074	0.011	7.009	<0.001
PPR	0.146	0.016	9.337	<0.001
SPPr	0.087	0.013	6.823	<0.001
SPPt	0.199	0.021	9.658	<0.001
eM	0.106	0.050	2.101	0.036
eY	0.030	0.007	4.598	<0.001
eS	0.137	0.009	14.625	<0.001
eEF	0.150	0.011	13.693	<0.001
eRM	0.199	0.013	15.240	<0.001
ePAR	0.133	0.010	13.157	<0.001
ePE	0.168	0.010	16.258	<0.001
ePI	0.131	0.009	13.946	<0.001
eEA	0.178	0.015	11.747	<0.001
eExA	0.178	0.014	12.917	<0.001
eCS	0.166	0.012	14.398	<0.001
eC	0.132	0.011	12.514	<0.001
eAR	0.156	0.013	12.087	<0.001
ePIPE	0.126	0.009	13.556	<0.001
eI	0.132	0.011	12.514	<0.001
erE	0.162	0.012	13.739	<0.001
eBE	0.086	0.006	13.730	<0.001
eEE	0.118	0.008	14.433	<0.001
eCE	0.102	0.008	13.485	<0.001
eSE	0.115	0.008	14.763	<0.001

provides information about variances. We found no indicators with latent constructs that have negative variance values; therefore, there is no Heywood case problem (Kline, 2023). The error terms (e.g., eRM = 0.199, ePE = 0.168, eAR = 0.156) are well within acceptable ranges, while the critical ratios remain high (c.r. > 1.96), supporting the estimates’ stability. Consequently, although certain predictors, such as SPPt (0.199) and PPR (0.146), exhibit notably significant contributions, the model contains no improper estimates and is statistically robust, free from the concerns of Heywood cases.

Table 7. Correlations among latent variables.

	Estimate	p
PPR <--> SPPt	0.288	0.004
SPT <--> SPPt	0.517	<0.001
SPT <--> PPR	0.289	0.009
PPR <--> SPPr	0.338	0.004
SPT <--> SPPr	0.577	<0.001
SPPr <--> SPPt	0.596	<0.001

presents the correlations among latent variables. They showed significant positive relationships, indicating meaningful associations between parents’ perceptions (PPR), and YLs’ perceptions of parents (SPPt), teachers (SPT), and peers (SPPr). For example, students’ perceptions of their teachers’ (SPT) and peers’ role (SPPr) were strongly correlated (r = 0.577, p < 0.001), supporting the interconnected influence of these social factors on YLs’ motivation and engagement.

4.3. Goodness of Fit

The model exhibited an overall satisfactory fit, as evidenced by a Chi-Square value of 67.452 (with a p-value indicative of fit) and a CMIN/DF of 3.895, both of which reside within the acceptable range (Schumacker & Lomax, 2010). The RMSEA was reported at 0.07 and the RMR at 0.014, both indices suggesting an adequate model fit. Furthermore, the goodness-of-fit indices supported this inference, with CFI = 0.934 and TLI = 0.976, each exceeding the commonly recommended thresholds (Hu & Bentler, 1999), thus indicating that the model effectively represents the observed data. In other words, the model fit the data very well.

4.4. Path Analysis

4.4.1. Direct Effect

YLs’ motivation (M) is significantly influenced by various factors, according to the direct effect results. Specifically, YLs’ perceptions of their teachers’ (SPT), parents’ (SPPt), and peers’ roles (SPPr), and parents’ perceptions of their roles (PPR) positively impact YLs motivation to learn English (M). Among these, YLs’ perceptions of their parents’ roles (SPPt) exhibits the strongest impact (0.676), and parents’ perceptions of their roles (PPR) has the lowest impact (see

Table 8. Direct effect.

Direct Effect (DE)	Estimate	p
M <--- SPT	0.509	<0.001
M <--- SPPr	0.583	<0.001
M <--- SPPt	0.676	<0.001
M <--- PPR	0.346	0.004
TE <--- SPT	0.246	0.009
TE <--- PPR	0.297	0.009
TE <--- SPPr	0.284	0.009
TE <--- SPPt	0.352	<0.001
TE <--- M	0.696	<0.001

). Furthermore, task engagement (TE) is also strongly affected by these variables. Positive associations are observed between task engagement (TE) and YLs’ perceptions of teachers’ (SPT), peers’ (SPPr), and parents’ roles (SPPt), and parents’ perceptions of their roles (PPR). YLs’ perceptions of parents’ role (SPPt) had the strongest impact on task engagement (TE) among others (0.352), indicating a moderate impact, and YLs’ perceptions of teachers’ roles (SPT) had the weakest impact on task engagement (TE) (0.246), indicating a small to moderate impact (Hair et al., 2018; Kline, 2023). Importantly, motivation (M) is a strong predictor of task engagement (TE) (0.696) (Kline, 2023), indicating that YLs’ motivation is vital in task engagement. All noted effects are statistically significant and indicate strong interconnections among these variables as p < 0.001, indicating significant, p = 0.004, and p = 0.009, very significant.

4.4.2. Indirect Effect

The results relating to indirect effects demonstrate that YLs’ perceptions of their teachers’ (SPT), parents’ (SPPt), and peers’ role (SPPr), and their parents’ perceptions of their role (PPR) significantly predict YLs’ task engagement (TE) through the mediation of their English learning motivation (M). As seen in

Table 9. Indirect effect.

Indirect Effect (ID)	Estimate	p
TE <--- M <--- SPT	0.354	0.004
TE <--- M <--- SPPr	0.406	0.001
TE <--- M <--- SPPt	0.471	0.001
TE <--- M <--- PPR	0.240	0.009

, the most significant and strong indirect effect originates from SPPt influencing task engagement (0.471), followed sequentially by YLs’ perceptions of peers’ role (SPPr) (0.406), YLs’ perceptions of their teachers’ role (SPT) (0.354, p = 0.004), and parents’ perceptions of their role (PPR) (0.240). Each pathway corresponds with statistically significant p-values less than 0.01. They indicate robust mediation. These perceptions predominantly exert a positive influence on task engagement by enhancing YLs’ motivation to learn English.

4.4.3. Total Effect

Table 10. Total effect of the model.

Paths	Predictor	DE on TE	ID via M	Total Effect
SPT → TE; SPT → M → TE	SPT	0.246	0.354	0.600
SPPr → TE; SPPr → M → TE	SPPr	0.284	0.406	0.690
SPPt → TE; SPPt → M → TE	SPPt	0.352	0.471	0.823
PPR → TE; PPR → M → TE	PPR	0.297	0.240	0.537

illustrates the total effects presented in the model. For example, YLs’ perceptions of their teachers’ role (SPT) has a small to moderate direct effect on task engagement (TE) (0.246) and a moderate indirect effect via motivation (0.354), resulting in a large total effect (0.600), as 0.600 > 0.50 is considered large (Hair et al., 2018; Kline, 2023). This outcome means that when the SPT increases by one standard deviation, task engagement (TE) is expected to increase by 0.600 standard deviation. The results demonstrate that each predictor, YLs’ perceptions of teachers’ (SPT), peers’ (SPPr), and parents’ roles (SPPt), and parents’ perceptions of their role (PPR), exerts a significant influence on task engagement both directly and indirectly through motivation as a mediator, thereby underscoring the critical role of motivation in the pathway to task engagement.

4.5. Squared Multiple Correlations (R²)

Table 11. R² values for endogenous constructs.

	Estimate	p
M	0.697	<0.001
TE	0.826	<0.001

illustrates the squared multiple correlations (R²) in the model, highlighting that a significant portion of the variance in the endogenous variables is explained by the predictors. For YLs’ motivation (M), the R² value is 0.697 (p < 0.001), showing that around 69.7% of the variance in YLs’ motivation is explained by its predictors: YLs’ perceptions of teachers’ (SPT), peers’ (SPPr), and parents’ role (SPPt), and parents’ perceptions of their role (PPR). Task engagement (TE) has an even higher R² value of 0.826 (p < 0.001), signifying that the model accounts for 82.6% of the variance in task engagement. Task engagement (TE) is directly influenced by YLs’ perceptions of teachers’ (SPT), peers’ (SPPr), and parents’ roles (SPPt), and parents’ perceptions of their role (PPR) and indirectly influenced through motivation (M). These high R² values indicate that the model has strong explanatory power both in motivation and in task engagement (Kline, 2023).

4.6. Final Model

Figure 2 illustrates the final SEM model of the relationships among YLs’ perceptions of their teachers’, parents’, and peers’ roles, their parents’ perceptions of their role, YLs’ motivation to learn English, and their task engagement at 15 Indonesian primary schools. The SEM technique facilitates the concurrent examination of both direct effects of YLs’ and parents’ perception of roles in their children’s motivation and task engagement, as well as indirect effects of YLs’ and parents’ perceptions of their impact on task engagements via YLs’ motivation, thus offering a holistic insight of the manner in which these social perceptions impact YLs’ motivation and task engagement in English language education. The final model was evaluated through the utilization of various fit indices to ensure its precise representation of the dataset.

In summary, both YLs’ and parents’ perceptions influenced the children’s motivation and task engagement. YLs’ motivation influenced their task engagement and their perceptions influenced task engagement via their motivation. Moreover, parents’ perceptions also influenced task engagement via YLs’ motivation.

5. Discussion

Based on the results of the analysis, the model fit was satisfactory: Chi-Square = 67.452, CMIN/DF = 3.895, RMSEA = 0.07, and RMR = 0.014. CFI = 0.934 and TLI = 0.976 exceeded recommended thresholds, confirming the model’s effectiveness. Therefore, this model can be used to examine children’s and parents’ perceptions of motivation and task engagement, especially at the elementary school level. According to this model, task engagement is influenced by children’s and parents’ perceptions and YLs’ motivation. The effect of the YLs’ and parents’ perceptions on task engagement is also mediated by motivation. Enhancing task engagement involves more than just addressing motivation. It is crucial to account for how students and parents view their roles, as these perspectives significantly impact both motivation and task engagement.

All constructs have AVE values above 0.50, showing that indicators’ variance is mainly due to the latent construct (Fornell & Larcker, 1981). CR values exceed 0.70, indicating reliability and consistency (Hair et al., 2018). These outcomes validate the reliable and valid measurement of YLs’ perceptions of their teachers’, parents’, and peers’ roles, as well as parents’ self-perception of their own roles, and YLs’ motivation and task engagement.

Regarding research question 1, YLs’ perceptions of their teachers’, parents’, and peers’ roles all positively impact motivation. This is in line with the framework Mihaljević Djigunović and Nikolov’s (2019) proposed for YLs’ language learning motivation. YLs’ perception of their teachers and peers had a significant and strong impact on their motivation (0.509 and 0.583, respectively) (Hair et al., 2018; Kline, 2023). These results also align with the study conducted by Wallace and Leong (2020) that showed that YLs’ perceptions of their teachers are positively associated with their levels of motivation. Moreover, YLs’ classmates also influence their motivation to learn English (Wallace & Leong, 2020). This result is in harmony with findings that peer support has a significant influence on learners’ motivation (Solhi, 2024) and the interaction among peers enhances learners’ motivation (Bui & Dao, 2023).

Furthermore, YLs’ perceptions of their parents’ role had the strongest impact on their motivation (0.676), suggesting that YLs’ motivation to learn English significantly increases when they experience high levels of parental support and involvement. YLs’ perception of their teachers and peers impact also had a significant and strong impact on their motivation (0.509 and 0.583, respectively) (Hair et al., 2018; Kline, 2023). This result supports the claim made by Mihaljević Djigunović and Nikolov’s (2019) that as learners age, their motivation is less impacted by teachers and more importantly shaped by their peers.

As far as research question 2 is concerned, parents’ perceptions of their roles positively impact YLs’ EFL learning motivation. Parents’ perceptions of their roles had a moderate impact on YLs’ motivation; however, this is the lowest among the relationships (0.346) (Hair et al., 2018; Kline, 2023). Again, this further supports the point (Mihaljević Djigunović & Nikolov, 2019) assertion peers’ impact increases as learners age. The result is also in line with a previous study on parental involvement in children’s learning. The more parents are concerned with how their children progress, the more likely they are to develop positive attitudes toward learning and higher levels of persistence (Rivera & Li, 2019).

As for research question 3, motivation emerged as a significant influence on task engagement (0.696); this means that the more children are motivated to learn English, the more they tend to engage with tasks in their English classes. This result supports a previous study indicating that motivation influenced how much children engaged in learning English actively (Jiao, 2024).

Regarding research question 4, YLs’ perceptions of their teachers’, parents’, and peers’ roles all positively impacted YLs’ task engagement. YLs’ perception of their parents had the strongest impact among them (0.352), and YLs’ perception of peers’ role had a weak impact (0.284) and teachers’ role exerted the weakest impact (0.246). These results align with previous studies that revealed that parents’ and peers’ support predicts learners’ engagement (Ansong et al., 2017). If children are aware of their parents’ high expectations, positive attitudes, interest, and active involvement in their English learning, these give their actions direction to do more tasks and more often to improve their English. The interaction among peers contributes to fostering students’ engagement (Bui & Dao, 2023).

In response to research question 5, we found that parents’ views of their role positively impacted their children’s task engagement. Parents’ perceptions of their role had a weak impact on task engagement (0.297), indicating a small to moderate impact (Hair et al., 2018; Kline, 2023). Our study highlighted that parental perception of their role is important, but their impact is relatively weak. Overall, although parents’ involvement influences YLs’ motivation (Butler, 2015; Nie & Mavrou, 2025), and parental engagement in children’s English learning contributes significantly to their motivation (Sumanti & Muljani, 2021), these relationships are less impactful.

As for the results concerning research question 6, the results showed that motivation mediated the impact of YLs’ perceptions of teachers’, parents’, and peers’ roles on YLs’ task engagement. YLs’ perceptions of parents’ role had the most significant and moderate indirect effect on task engagement (0.471), followed by YLs’ perceptions of teachers’ role had a significant moderate indirect effect on task engagement (0.345). This aligns with a previous study where motivation mediated the relationship between perceived teacher support and engagement (Yang & Du, 2023). Motivation acts as a mediator in the connection between how students perceive their parents’ and teachers’ involvement in homework and their own engagement with it (Núñez et al., 2019). Each pathway had a significant p-value less than 0.01, signifying strong mediation. These perceptions primarily have a positive effect on task engagement by increasing YLs’ motivation to learn English. The results highlight the essential role of motivation plays in promoting task engagement.

Results related to research question 7 showed that motivation mediated the impact of parents’ perceptions of their role on YLs’ task engagement. Parents’ perceptions of their roles had the lowest and a small to moderate indirect effect on task engagement (0.240), but this interaction indicates strong mediation. Parents’ perceptions primarily have a positive effect on task engagement by increasing their children’s motivation to learn English.

The model’s total effects showed that each predictor, namely YLs’ perceptions of teachers’, peers’, and parents’ roles, and parents’ perceptions of their role, significantly influences task engagement, both directly and indirectly, with motivation as a mediator. As an example, YLs’ perceptions of their teachers have a direct effect on task engagement of 0.246 and an indirect effect through motivation of 0.354, totaling 0.600, indicating a large effect (Hair et al., 2018; Kline, 2023). This result confirmed outcome of a previous study claiming that the perception of teachers influenced engagement via motivation (Yang & Du, 2023).

Motivation has an R² of 0.697 (p < 0.001), meaning 69.7% of motivation is explained by YLs’ perception of their teachers’, parents’, and peers’ roles as well as parents’ views of their roles. Task engagement’s R² is 0.826 (p < 0.001), with 82.6% of task engagement explained by the same predictors, plus an indirect influence through motivation. These R² values show strong explanatory power in motivation and task engagement (Kline, 2023).

6. Conclusions and Implications

We found that YLs’ and parents’ perceptions of various stakeholders’ roles positively impact children’s EFL learning motivation and task engagement. Motivation significantly influenced task engagement and mediated the impact of YLs’ and parents’ perceptions on it.

This study offers insights into the ways in which SEM can analyze complex relationships between social influences (teachers, peers, parents) and psychological outcomes (motivation, task engagement) in the case of young learners of English in Indonesian primary schools. By integrating survey data from 15 schools and validating the model through reliability and validity checks, this research offers a model for future studies on social and motivational factors in English language learning, especially at the primary school level. It also highlights the need to assess both direct and indirect effects to fully understand contextual variables such as school characteristics, teacher and peer support, and family background.

These findings offer strong evidence that teacher, peer, and parent support can significantly boost YLs’ motivation to learn English and to engage with tasks. These results imply what schools should consider implementing strategies like parent involvement, peer collaboration, and teacher training on motivation to foster a supportive environment. The findings help us understand social influences on motivation, showing how context affects young learners’ motivation to learn English and to work on tasks scaffolding learning.

7. Limitation

This study’s focus on fifth-grade students in a specific educational context limits its generalizability. Self-reported surveys may be biased due to social desirability or participants’ self-awareness. The cross-sectional design restricts causal inference about influences on YLs’ motivation and engagement. Although SEM provides a robust analysis, it depends on model fit and observed variables, possibly missing complex motivational processes.

Future research should broaden the study and involve YLs from various grades and educational environments to improve the findings’ generalizability. Longitudinal studies are advised to explore the impact of teachers, peers, and parents on motivation and task engagement over time. Also, using interviews, classroom observations, and teacher reflections could offer deeper triangulation of insights into the dynamics of language learning motivation and task engagement.