Influence of self-efficacy and metacognition on malaysian pre-university students’ chemistry academic motivation: the moderating role of gender and locality

Abstract

This study explores the relationships among self-efficacy, metacognition, and academic motivation in chemistry, focusing on the moderating roles of gender and locality among Malaysian pre-university students. Using a quantitative approach, data were collected from 556 students and analysed through partial least square-structural equation modelling (PLS-SEM). The findings indicate that both self-efficacy and metacognition significantly predict academic motivation. Specifically, metacognition had a stronger influence on academic motivation (β = 0.412) than self-efficacy (β = 0.288). Gender significantly moderated the relationship between self-efficacy and academic motivation, with female students showing a stronger link between self-efficacy and motivation than male students (β = −0.07). However, locality did not significantly moderate the relationships between self-efficacy, metacognition, and academic motivation. The combination of self-efficacy and metacognition accounted for 42.3% of the variance in academic motivation (R² = 0.423). These findings provide valuable insights into the factors that drive academic motivation in chemistry education. They suggest that educators emphasise self-efficacy, especially among female students, and integrate metacognitive strategies into the curriculum to enhance pre-university student motivation. Future research should explore the effects of educational interventions targeting self-efficacy and metacognition on academic motivation in chemistry.

---

1. Introduction

Academic motivation is crucial in the educational process, directly influencing students' engagement levels (Dierendonck et al., 2021) and affecting their learning performance (Cayubit, 2022; Naim and Karpudewan, 2024). Numerous studies indicate its significance, highlighting its strong predictive power for students' performance in chemistry courses (Ferrell et al., 2016; Pratt et al., 2023). For instance, in a study involving 449 undergraduates across 18 institutions in the United States, Pratt et al. (2023) found a positive relationship between the students' intrinsic motivation and their academic performance in foundational-level inorganic chemistry courses. Pratt et al. (2023) affirmed the findings of Liu et al. (2017), involving 238 university students in a first-semester general chemistry course and confirmed the correlation between behaviour propelled by internal rewards (intrinsic motivation) and academic achievement in chemistry. Ardura et al.'s, (2021) study involving secondary school students in Spain revealed that the level of motivation impacts students' decisions regarding their continued engagement with chemistry in subsequent educational phases or opting to abandon the subject.

Self-efficacy and metacognition predominate among the various factors influencing academic motivation in chemistry. Higher levels of self-efficacy among students are associated with increased motivation to participate in academic pursuits, persevere through challenges, and set ambitious goals for their academic endeavours (Syskowski and Kunina-Habenicht, 2023). Metacognition, the awareness and regulation of one's cognitive processes, influences how students plan, monitor, and evaluate their learning strategies and motivates them to learn chemistry as they are better at managing challenges, preventing burnout, and experiencing a state of flow in their chemistry learning (Guo et al., 2022). These claims (i.e., self-efficacy and metacognition influence academic motivation) are particularly true for pre-university students (Zhou et al., 2020). Pre-university programmes are conducted to equip students with the necessary skills and knowledge to transition to university higher education successfully (Allen et al., 2022). Students enrolled in these programmes face greater academic demands than in secondary school. They are also expected to take more responsibility for their learning.

Beyond the effects of self-efficacy and metacognition on academic motivation, gender also influences motivation to learn chemistry (Salta and Koulougliotis, 2020; Ardura et al., 2021), self-efficacy (Sunny et al., 2017), and metacognition (Abdelrahman, 2020). While the effects of gender on these factors are well-documented individually, there is a scarcity of research on how gender collectively moderates the relationships between self-efficacy and academic motivation, as well as between metacognition and academic motivation. Similarly, locality, defined as rural or urban, also influences academic motivation (Saw and Agger, 2021), self-efficacy (Astalini et al., 2020), and metacognition (Bakkaloglu, 2020). The effects of locality on these factors are usually documented in isolation.

Strong evidence on the influence of metacognition and self-efficacy on academic motivation, as well as the moderating role of gender and locality in these relationships, has been explicitly presented in the aforementioned studies. However, a common limitation in these studies is that the influence of each variable is examined in isolation. In reality, these variables do not exist independently—students' metacognition and self-efficacy develop simultaneously, and they also belong to different gender groups and localities. While there is extensive literature in chemistry education analysing these variables separately, research exploring their complex interconnections, including moderating effects that reflect real-world settings, remains limited. This study aims to bridge this gap by collectively examining the effects of self-efficacy and metacognition on chemistry academic motivation among pre-university students, while also investigating how gender and locality moderate these relationships.

2. Background of study

2.1. Academic motivation

Academic motivation is essential for student engagement and success in educational settings. It is a multifaceted concept explored through multiple theories, each providing distinctive insights into the factors prompting students' learning behaviours (Urhahne and Wijnia, 2023). The most prominent academic motivation theories include attribution theory (Graham, 2020), expectancy-value theory (Eccles and Wigfield, 2020), social cognitive theory (Schunk and DiBenedetto, 2020), achievement goal theory (Urdan and Kaplan, 2020), and self-determination theory (Ryan and Deci, 2020). Although these theories do not offer a unified concept of academic motivation, they collectively provide a comprehensive framework for describing, explaining, and predicting learning behaviours' direction, initiation, intensity, and persistence (Urhahne and Wijnia, 2023). The current study adopts Ryan and Deci's (2020) self-determination theory, which is extensively applied in chemistry education as a model to explore and comprehend students’ academic motivation (McAlpin et al., 2023). Elford and colleagues (2022) explain that self-determination theory is particularly suitable for fostering motivation in chemistry education because the three psychological needs – autonomy, competence, and relatedness – are echoed by the theory that is deeply embraced for designing effective instructional strategies. This theory identifies two (2) global types of motivation: intrinsic and extrinsic (Ryan and Deci, 2020). Intrinsic motivation in chemistry education pertains to students’ internal drive to learn chemistry for its inherent satisfaction without external incentives, such as a student's spontaneous interest in understanding chemical reactions and solving complex problems (Meydan, 2021; Pratt et al., 2023). Consequently, this type of motivation is often associated with promoting deep learning in chemistry (Liu, 2021). Moreover, according to Ryan and Deci (2020), autonomy support in the classroom, mediated by intrinsic motivation, encourages students to be self-driven and feel a sense of ownership over their learning – leading to deeper learning as they develop the ability to form connections between new and existing knowledge.

On the other hand, extrinsic motivation includes the desire to engage in chemistry class activities for external rewards or outcomes, such as grades, praise, or recognition (Meydan, 2021; Pratt et al., 2023). Elford et al. (2022) state that while extrinsic motivation can encourage students to complete tasks and achieve scholastic targets, it may not lead to a deep understanding or long-term interest in chemistry. This is because extrinsic motivation focuses on external incentives or pressures rather than genuine interest or personal satisfaction. Nevertheless, according to self-determination theory, extrinsic motivation ranges along a continuum from integrated regulation to external regulation, with the former being closely aligned with intrinsic motivation as it is more self-determined (Ryan and Deci, 2020). Apart from intrinsic and extrinsic motivation, Ryan and Deci (2020) assert that amotivation must be considered in understanding human motivation fully. Amotivation refers to a state in which individuals lack the motivation to act due to a perceived absence of competence or control over the outcomes of their actions. Unlike intrinsic motivation, which arises from genuine interest, or extrinsic motivation, driven by external rewards, amotivated individuals experience a disconnect between their efforts and potential results. In an educational setting, students may experience amotivation when they feel that their actions will not lead to success, resulting in disengagement and a lack of persistence in their academic pursuits.

2.2. Self-efficacy and academic motivation

Self-efficacy is essential in comprehending how students are engaging with academic tasks. According to Bandura's (1986) social cognitive theory, self-efficacy is a pivotal factor, representing individuals' beliefs in their ability to organise and execute actions required to attain specific goals successfully. These beliefs are shaped by four key sources: mastery experiences, vicarious experiences, verbal persuasion, and emotional/physiological states. While Bandura's (1986) conceptualisation of self-efficacy emphasises the importance of task-oriented self-belief, other definitions offer valuable perspectives. For example, general self-efficacy highlights the individual's confidence in managing multiple tasks and coping with obstacles across different domains, whereas specific self-efficacy is limited to a particular task (Luszczynska et al., 2005). Contextual self-efficacy considers how a specific context or situation affects an individual's confidence in their ability to execute a particular task (Rieder et al., 2021). Moreno et al. (2021) state that self-efficacy is domain-specific, i.e., confidence in one area (e.g., biology) may not translate to another (e.g., chemistry).

Thus, Bandura's (1986) definition of self-efficacy is applied in this study as it highlights the key factors – mastery experiences, vicarious experiences, verbal persuasion, and emotional/physiological states – essential for successful engagement in chemistry learning. Mastery experiences are the most powerful influence on students' self-efficacy in chemistry, stemming from their direct successes or failures in tasks such as solving chemical equations, conducting experiments, or mastering complex concepts. Success in these tasks strengthens their self-efficacy, while repeated failures can diminish it. Vicarious experiences occur when students observe their peers successfully completing chemistry-related tasks, such as performing lab experiments or explaining complex theories. This can enhance their belief in their abilities through modeling. Verbal persuasion, including encouragement and feedback from teachers, lab partners, and classmates, boosts students’ confidence in tackling chemistry challenges. Finally, students’ self-efficacy in chemistry is influenced by their emotional and physiological states, such as stress or anxiety during exams or lab work. Effectively managing these responses can enhance their confidence and engagement in chemistry learning.

Several studies have identified self-efficacy as a key factor in academic motivation. Abdolrezapour and colleagues (2023) observed a positive relationship between self-efficacy and academic motivation in the context of online education. Additionally, Shofiah et al. (2023) found that, among Indonesian undergraduates, self-efficacy mediated the relationship between academic motivation and achievement during online learning. While the broader academic literature provides substantial evidence of the link between self-efficacy and academic motivation, this connection has not been extensively explored in chemistry education settings. However, illuminating insights emerge from a study conducted by Alci (2015) involving general chemistry undergraduates. Alci's (2015) data analysis revealed significant positive correlations between self-efficacy and extrinsic and intrinsic motivation. A correlation between self-efficacy and intrinsic and extrinsic motivations is also evident in studies conducted using the Science Motivation Questionnaire II (Glynn et al., 2011) and the Chemistry Motivation Questionnaire II (Salta and Koulougliotis, 2015). Another study involving pre-university students in a different US institution highlighted that self-efficacy is interconnected with other motivational constructs, such as personal and situational interest (Ferrell et al., 2016). Therefore, this suggests hypothesis H1: there is a significant positive effect of self-efficacy on academic motivation in chemistry among pre-university students.

2.3. Metacognition and academic motivation

The term metacognition was first coined by Flavell (1979). According to Flavell (1979), metacognition involves monitoring one's memory, comprehension, and other cognitive activities – an idea succinctly captured as thinking about one's thinking (Janeck et al., 2003; Hong et al., 2015). Metacognition is a diverse concept encompassing various definitions and applications across different contexts. For instance, in psychology, metacognition is an umbrella term that comprises a range of processes, from isolated actions (e.g., introspective accuracy) to more integrated activities like self-assessment (Pinkham, 2019). Brown (1987) defined metacognition as skills that can be nurtured and enhanced over time, specifically in education settings. Other researchers, such as Schraw and Moshman (1995), offered a definition extended from Flavell's (1979) in which metacognition is comprised of declarative knowledge (i.e., knowing what) and procedural knowledge (i.e., knowing how) in metacognitive processes. Flavell's (1979) definition of metacognition is applied in this study due to its holistic integration of knowledge and regulation of cognition aspects, which are essential in chemistry education (Lavi et al., 2019). Knowledge of cognition refers to what individuals know about their cognition (Flavell, 1979). This aspect of metacognition includes three (3) metacognitive knowledge types: declarative, procedural, and conditional (Schraw, 2001). The other metacognition aspect, the regulation of cognition, denotes the actions that aid individuals in regulating learning and problem-solving (Flavell, 1979). It comprises three (3) metacognitive regulations: planning, monitoring, and evaluation (Schraw, 2001).

Metacognition and academic motivation exhibit a positive correlation. Educational activities focusing on metacognition are linked to increased student motivation (Saxena, 2020). The study conducted in Canada with high school students revealed that they exhibited a higher level of metacognition in their learning processes, establishing a noteworthy positive correlation with their academic motivation (Landine and Stewart, 1998). Similarly, Abdelrahman (2020) conducted research in the United Arab Emirates (UAE) among undergraduates, emphasising a significant and positive association between students' metacognition and academic motivation. Furthermore, the findings underscored robust correlations: one between metacognitive knowledge awareness and academic intrinsic motivation and another between metacognitive regulation awareness and intrinsic motivation (Abdelrahman, 2020). Sen's (2016) study explored the impact of metacognitive levels on the motivation of Turkish chemistry preservice teachers. The investigation delved into the correlation between metacognition and academic motivation among the participants. Results revealed a significant difference in the mean intrinsic motivation score among high-intermediate and high-low groups, with no notable distinction in the intermediate-low category (Sen, 2016). The claims above lead to hypothesis H2: there is a significant positive effect of metacognition on academic motivation in chemistry among pre-university students.

2.4. Moderating role of gender

Gender differences in academic motivation for chemistry have been extensively investigated, revealing that the extent and nature of these differences can vary. Salsabila and Huda (2023) report significant differences in chemistry learning motivation based on gender among high school students in Indonesia, suggesting that academic motivation levels in chemistry are influenced by gender. Male students are generally more intrinsically motivated in chemistry (which correlates with better academic performance) than their female counterparts (Meydan, 2021). This is potentially due to societal expectations and gender roles that influence their engagement in chemistry (Salsabila and Huda, 2023). For example, female students tend to feel like outsiders in the classroom because they perceive chemistry as a subject more suited for boys (Cousins and Mills, 2015). In contrast, a study by Ardura et al. (2023) on Spanish high school students found that female students exhibited higher intrinsic motivation than their male counterparts when choosing chemistry. Male students, on the other hand, were more influenced by external motivators such as career prospects (Ardura et al., 2023). This aligns with a study by Salta and Koulougliotis (2020), which suggests that female students might be more intrinsically motivated to learn chemistry than male students. However, the study also suggests that the impact of gender on motivation to learn chemistry is likely low and indirect (Salta and Koulougliotis, 2020). Similarly, another study conducted by Salta and Koulougliotis (2022) among undergraduates in Greece found that females and males exhibited overall similar motivations for learning chemistry, indicating that gender has a negligible effect on academic motivation in chemistry.

The research on how gender influences chemistry self-efficacy presents mixed findings. Moreno and colleagues (2021) report that all undergraduate groups experienced increased chemistry self-efficacy after a semester, with both male and female students showing similar growth over time. However, another study conducted among 300 chemistry pre-service teachers in Indonesia found that female participants were significantly more self-efficacious than their male counterparts (Wahyudiati et al., 2019). Similarly, a study by Whitcomb et al. (2020) reported that female engineering students outperformed male participants in chemistry courses, and their self-efficacy aligns with their performance. However, over time, a non-significant gender gap in chemistry self-efficacy emerged, with male engineering students having slightly higher self-efficacy in several areas (Whitcomb et al., 2020). This echoes a study that found male undergraduates at a US institution believed more strongly in their ability to succeed in chemistry than their female peers, which led to higher test anxiety among female undergraduates (Sunny et al., 2017). Thus, these findings indicate gender differences in chemistry self-efficacy (Sunny et al., 2017; Wahyudiati et al., 2019; Whitcomb et al., 2020). The literature points to factors contributing to gender differences in chemistry self-efficacy, such as self-regulatory strategies (Seyhan, 2016) and anxiety and psychological factors (Amaliyah et al., 2021). The effects of gender on academic motivation and self-efficacy, as explained earlier, explain the formulation of hypothesis H3: gender moderates the relationship between self-efficacy and academic motivation.

Research on the influence of gender on metacognition in chemistry is limited, and the findings that do exist present mixed results. Jumani et al. (2010) studied 900 students in Pakistan, and the findings indicated similar metacognitive awareness levels between males and females. This aligns with a recent study by Ahmed and colleagues (2019) involving 125 Indonesian high school students. The male and female students were equally competent in the three (3) metacognitive knowledge domains, i.e., declarative knowledge, procedural knowledge, and conditional knowledge, specifically in the chemical bonding topic (Ahmed et al., 2019). Therefore, these studies indicate that metacognition levels in chemistry are not affected by gender, especially in the context of high school students. However, a study conducted among undergraduates revealed differences in metacognition between males and females. A qualitative study reported by Muteti et al. (2023) showed that female general chemistry students at one institution in the United States used metacognitive strategies less than males before a 50-metacognition instruction, but the metacognitive equity gap was reduced over the semester of data collection following the metacognition lesson. Similarly, in a related science field (i.e., biology), research revealed that female university undergraduates in Indonesia might be more inclined toward metacognition than their male counterparts, potentially contributing to females demonstrating superior problem-solving skills in science (Adiansyah et al., 2021). Merchán Garzón et al. (2020) suggest that the factors contributing to the differences in metacognition levels between males and females may include social, educational, or psychological factors that influence how both genders approach learning and problem-solving in chemistry. The effects of gender on academic motivation and metacognition, as explained earlier, explain the formulation of hypothesis H4: gender moderates the relationship between metacognition and academic motivation.

2.5. Moderating role of locality

Locality plays a significant role in shaping academic motivation, as students from different geographical backgrounds may encounter distinct challenges and opportunities that impact their learning. Rural areas as regions with lower population density, limited access to infrastructure, and fewer educational resources such as advanced technology and extracurricular programmes. In contrast, urban areas are highly populated regions with greater access to diverse educational opportunities, including well-equipped schools and extracurricular activities. Although research specifically examining how urban and rural settings affect students' academic motivation in chemistry is limited, a study by Zhang and Zhou (2023) in China revealed notable disparities between urban and rural students. The study found that urban students scored significantly higher motivation in learning chemistry than their rural counterparts, though the effect sizes were modest. Insights from related fields also support this finding. For instance, Saleh (2014) reports higher motivation levels in physics among urban students in Malaysia compared to their rural peers. Additionally, Hill et al. (2018) found that American students residing in urban areas are more involved in informal science activities, such as participating in after-school science clubs and visiting museums, which enhanced their motivation and interest in science. These findings suggest that urban schools may offer a better learning experience by providing more extensive educational infrastructure and better access to technology, which helps keep students more engaged in science subjects (Hariyati et al., 2021).

There is scant research on the influence of geographical location (e.g., urban and rural) on students’ chemistry self-efficacy. Nevertheless, the available study's results suggest no significant difference between the urban and rural groups. For instance, a study in Ethiopia reported that rural secondary school students were slightly more self-efficacious in chemistry, but the differences between urban and rural students were not statistically significant (Asfaw, 2022). Findings from general science indicate that locality may play a significant role in determining students’ self-efficacy levels. For example, a study involving 926 junior high schools (i.e., 511 urban students and 415 rural students) in Indonesia revealed that urban students tend to have higher confidence in learning natural sciences than their rural counterparts (Astalini et al., 2020). Similarly, Ibrahim et al. (2019) studied 16-year-old Malaysian students from two (2) urban and two (2) rural secondary schools and found that urban students exhibited more favourable attitudes toward learning physics compared to rural students, which could imply higher self-efficacy. This disparity is attributed to urban students' greater exposure to diverse experiences and role models in science (Hill et al., 2018) and access to more effective pedagogical approaches (Bolshakova et al., 2011). The effects of locality on academic motivation and self-efficacy, as explained earlier, explain the formulation of hypothesis H5: locality moderates the relationship between self-efficacy and academic motivation.

Research on how urban and rural localities influence students’ metacognition in the context of chemistry education is limited. However, evidence suggests that urban students are generally more likely to utilise metacognition than their rural counterparts (Taghieh et al., 2019). For instance, a study conducted in Turkey found that urban primary and secondary school students had significantly higher levels of metacognitive awareness than their rural peers. This difference was attributed to a more active educational environment and greater exposure to stimuli in urban schools (Bakkaloglu, 2020). Studies in fields related to chemistry also provide valuable insights into how geographical location may influence metacognition, particularly in science settings. For example, Alam (2020) examined 840 secondary school students studying biology in India. The findings indicated that urban students demonstrated greater competence in applying metacognitive strategies than their rural peers, contributing to better biology performance among urban students. According to Alam (2020), this difference in metacognition levels may be due to urban students' exposure to better educational environments. Interestingly, a study of metacognition levels in physics among Thai high school students revealed that urban and rural classrooms lacked sufficient metacognitive practices (Pimvichai et al., 2015). Despite urban students having slightly better access to environments conducive to metacognitive development, the impact of geographical location on metacognition was nuanced and complex (Pimvichai et al., 2015). This suggests that the influence of geographical location on students’ metacognition may vary depending on the specific context and subject matter. The effects of locality on academic motivation and metacognition, as explained earlier, explain the formulation of hypothesis H6: locality moderates the relationship between metacognition and academic motivation.

2.6. Contextualisation and conceptualisation of the study

Studies from various countries have demonstrated the influence of self-efficacy and metacognition on academic motivation, and it is anticipated that these connections also exist among Malaysian pre-university students. These students encounter distinct educational obstacles due to the requirements of pre-university programmes to prepare them for higher education, particularly in chemistry. Since chemistry is a fundamental subject for many science students, understanding the impact of self-efficacy and metacognition and gender and locality on their motivation is essential for developing more efficient instructional strategies.

In Malaysia, several studies have examined self-efficacy, metacognition, and academic motivation, particularly within the broader field of science education (Leong et al., 2018; Choy et al., 2020; Wong et al., 2021). However, despite these efforts, similar to the global scenario, the integration of these variables, specifically within chemistry education, remains limited. The existing literature has predominantly focused on general science disciplines, while the unique cognitive demands of chemistry, a core subject in pre-university programmes, are often overlooked. Moreover, while studies have investigated the role of location (rural vs. urban) on academic outcomes for science students, few have explored the specific impacts on chemistry teaching and learning. The same can be said for gender, where most research has centred on general academic performance or participation in science subjects, leaving a gap in understanding how gender moderates explicitly the relationship between self-efficacy, metacognition, and academic motivation in chemistry education. The six hypotheses presented in this study aim to address the gap in the existing literature regarding these factors. This is crucial for designing chemistry-specific content and targeted teaching strategies that cater to the diverse needs of students. Examining these under-researched influences contributes to a deeper understanding of the cognitive and demographic factors affecting academic motivation in chemistry education, ultimately supporting the development of more effective pedagogical approaches (Graham et al., 2019).

Fig. 1 illustrates this study's conceptual framework. It depicts the hypothesised relationships among self-efficacy, metacognition, and academic motivation in chemistry, with gender and locality as moderating variables.

Fig. 1 Conceptual framework of the influence of self-efficacy and metacognition on academic motivation with gender and locality as moderators.

3. Purpose of the study

Current research on the relationship between chemistry self-efficacy and metacognition concerning students' academic motivation in chemistry lacks sufficient exploration of how locality and gender moderate these factors. This study seeks to address this gap by investigating how students' chemistry self-efficacy and metacognition interact with their academic motivation in chemistry, specifically focusing on the moderating influences of locality and gender within the Malaysian pre-university setting. Thus, this leads to the formulation of six hypotheses:

(a) H1: self-efficacy significantly positively affects academic motivation in chemistry among pre-university students.

(b) H2: metacognition significantly positively affects academic motivation in chemistry among pre-university students.

(c) H3: gender significantly moderates the effect between self-efficacy and chemistry academic motivation among pre-university students.

(d) H4: gender significantly moderates the effect between metacognition and chemistry academic motivation among pre-university students.

(e) H5: locality significantly moderates the effect between self-efficacy and chemistry academic motivation among pre-university students.

(f) H6: locality significantly moderates the effect between metacognition and chemistry academic motivation among pre-university students.

4. Methodology

4.1. Research design and participants

The study employs a cross-sectional research design to capture the relationships between self-efficacy, metacognition, and academic motivation among Malaysian pre-university students and simultaneously measures the moderating roles of gender and locality among 556 pre-university students aged approximately 18–19 from a pre-university institution in Malaysia.

The study was conducted at a government-run pre-university institution in Malaysia that delivers science courses comparable to the Cambridge Advanced Level (A-Level). Typically, students admitted to this institution are high achievers in the Sijil Pelajaran Malaysia (SPM) examination, equivalent to the British General Certificate of Education Ordinary Level (GCE O-Level). After completing their pre-university studies, these students commonly pursue further undergraduate education locally and at international universities, focusing on diverse scientific disciplines. The pre-university institution offers a variety of science-focused courses, including chemistry, biology, physics, computer science, and mathematics. The chemistry course has three main components: physical, inorganic, and organic. Students study fundamental concepts such as thermodynamics, kinetics, bonding, and reaction mechanisms, which provide a solid foundation for future studies in engineering and medicine. The course structure combines theoretical lectures with practical laboratory sessions, allowing students to apply scientific concepts through hands-on experiments. Assessments are conducted through continuous coursework, practical reports, and final examinations, ensuring a balanced evaluation of theoretical understanding and practical skills. Furthermore, the course promotes active learning through group discussions and peer-led activities. The chemistry curriculum is delivered over two semesters and organised under a modular system, which ensures a balanced workload and comprehensive coverage of key topics.

A minimum sample size of 178 was suggested based on the results of the G*Power 3.1.9.7 calculation: F tests family; linear multiple regression, fixed model, R² deviation from zero; A priori, 0.95 level of power, 0.15 effect size (f²) at significance level p < 0.05, and eleven (11) predictors (3 CCSS factors, 6 CMI factors, gender, and locality). Therefore, the sample size acquired for this study (n = 556) was deemed sufficient. The pre-university institution was located using a convenience sampling strategy based on accessibility and willingness to participate. The institution represents typical Malaysian pre-university institutions offering comparable science-focused curricula. This is due to its alignment with national academic standards, the high academic achievement of its students, and the variety of science courses it offers, including chemistry, which mirrors the offerings in other pre-university programmes nationwide. Therefore, performing the research in an institution chosen based on convenience does not limit the generalisation of the findings to other pre-university institutions.

Malaysia has 15 government-run pre-university colleges, each enrolling approximately 1500 to 3000 students per intake. The pre-university institution where this study was conducted typically enrols between 1900 and 2400 science students per cohort. Among the 556 students who participated, the gender distribution included 150 males (27%) and 406 females (73%). Of these students, 301 (54%) came from rural areas, while 255 (46%) were from urban backgrounds. For comparison, the science student population at this pre-university institution during the study comprised 2285 students, with a gender distribution of approximately 30% males and 70% females, closely mirroring our sample's gender ratio. Similarly, the science student population's rural-to-urban ratio was about 60 : 40, comparable to this study's sample's rural and urban composition. This study defines urban students as those whose hometowns and schooling before the pre-university programme were in urban areas, with access to more comprehensive educational resources and facilities to support learning in science subjects. In contrast, rural students are those from rural hometowns, attending primary and secondary schools in areas with more limited educational resources and infrastructure. This classification is crucial in examining how locality influences students' academic experiences and outcomes within the pre-university college setting.

4.2. Instrumentation

In the current study, three (3) psychometric instruments were administered to measure the key variables: chemistry academic motivation, self-efficacy, and metacognition. This study employed the Academic Motivation Scale (AMS), developed initially by Vallerand et al. (1992) and later adapted for chemistry by Liu et al. (2017). The AMS, grounded in Ryan and Deci's (2020) self-determination theory, categorises motivation into three (3) constructs: intrinsic, extrinsic, and amotivation. Liu et al.'s (2017) adaptation maintains these categories but applies them specifically to chemistry, enabling a more focused examination of motivation in this context. Intrinsic motivation in chemistry includes subtypes such as motivation to know (desire to understand chemical phenomena), motivation to accomplish (satisfaction from mastering tasks like solving chemical problems), and motivation to experience stimulation (excitement from hands-on activities, such as experiments). Extrinsic motivation involves external regulation (actions driven by rewards like grades), introjected regulation (behaviour motivated by internal pressures like guilt avoidance), and identified regulation (recognising chemistry's value for personal goals). Amotivation reflects a lack of intention or purpose in learning chemistry, offering insights into disengagement. Liu et al.'s (2017) adaptation allows for a nuanced exploration of how different motivational factors influence students’ engagement and performance in chemistry, making it a valuable tool for studying academic motivation in STEM. The Academic Motivation Scale-Chemistry (AMS-Chemistry), developed by Liu et al. (2017), consists of 28 items rated on a five-point Likert scale. All AMS-Chemistry items are rated on a scale ranging from 1 (not at all) to 5 (exactly), except for four (4) items related to the amotivation factor, which are rated from 1 (exactly) to 5 (not at all). The items were used in their original wording for the present study.

The adaptation of Bandura's (1986) concept of self-efficacy into instruments developed by Dalgety et al. (2003) and Uzuntiryaki and Yeşim Çapa Aydın (2009) demonstrates the evolving application of self-efficacy theory within chemistry education. Dalgety et al. (2003) designed the Chemistry Attitudes and Experiences Questionnaire (CAEQ) to assess chemistry self-efficacy, focusing on students' confidence in performing task-specific academic skills, such as learning and applying chemistry theory. This aligns with Bandura's (1986) definition of self-efficacy as the judgment of one's ability to organise and execute tasks. However, their scale primarily emphasised cognitive aspects and did not encompass practical laboratory skills like handling equipment, materials, and chemicals in a laboratory setting. Uzuntiryaki and Çapa Aydın (2009) expanded on this by developing the College Chemistry Self-Efficacy Scale (CCSS), which broadened self-efficacy measurement to include three dimensions: cognitive skills, psychomotor skills (e.g., laboratory tasks), and the everyday application of chemistry. This broader approach built upon Dalgety et al.'s (2003) work by integrating both theoretical and practical components of self-efficacy, creating a more comprehensive tool while maintaining Bandura's (1986) focus on task-specificity across various contexts. The second instrument of the current study, the CCSS (Uzuntiryaki and Çapa Aydın, 2009), is used to determine participants’ chemistry self-efficacy. This scale consists of 21 items rated on a five-point scale, from 1 (very poorly) to 5 (very well).

The third psychometric tool employed in this study was the Chemistry Metacognition Inventory (CMI). A modified version of Taasoobshirazi et al.'s (2015) Physics Metacognition Inventory (PMI). Building on Flavell's framework, Taasoobshirazi et al. (2015) extended this conceptualisation in their studies by distinguishing between knowledge of cognition and regulation of cognition in problem-solving. They identified declarative, procedural, and conditional knowledge as critical for solving complex scientific problems, such as those in physics and chemistry, while highlighting the role of planning, monitoring, and evaluation in regulating learning during problem-solving. This integration of Flavell's (1979) holistic view of metacognition into educational contexts underscores the importance of understanding and managing cognition for successful scientific problem-solving. Taasoobshirazi and Farley (2013) recommended two (2) adjustments for researchers in chemistry education who wished to use the PMI. Firstly, they suggested substituting the word “physics” with “chemistry.” Secondly, they proposed revising the information management items to specify “molecular diagrams” instead of “free-body diagrams.” Following Taasoobshirazi and Farley's recommendation, the term “molecular diagram” was used. In chemistry education, molecular diagrams serve as visualisation tools to support understanding of molecular structures and the behaviour of atoms and molecules at the sub-microscopic level (Chittleborough and Treagust, 2008). The CMI psychometric tool includes 26 items rated on a five-point scale, extending from 1 (never true of myself) to 5 (always true of myself).

Science courses (physics, chemistry, and biology) at the pre-university level in Malaysia are taught in English, requiring students to possess considerable English language proficiency. The participating students demonstrated sufficient English proficiency to comprehend the items in the instruments, as reported in the pilot study.

4.2.1. Pilot study. The data from each instrument were evaluated for evidence to support content validity, construct validity, and criterion-related validity, as well as internal consistency reliability, with the study population. A pilot study was conducted to address four validation elements – test content, response process, internal structure, and reliability (Lewis, 2022). The sample for the pilot study consisted of 96 pre-university students who were not participants in the real study and three chemistry education experts.

4.2.2. Test content validity. The current study's content validity was ensured by consulting three subject matter experts in chemistry education, who reviewed the items to confirm that the AMS-Chemistry adequately represented the spectrum of academic motivation, encompassing intrinsic, extrinsic, and amotivation dimensions. Experts also reviewed the CCSS and CMI to ascertain that the items were appropriate for measuring chemistry self-efficacy and metacognition. The subject matter experts indicated that the instruments were suitable for pre-university students and comprehensible for this group. However, they recommended clarifying the instructions, items, and sections in the instruments for the study participants. Consequently, participants were given a briefing before completing the instruments. Additionally, the experts suggested revising the term “periodically”, which appeared in several items in the CMI, to “from time to time” for improved clarity.

4.2.3. Response process. Interviews were conducted with five pre-university students recruited randomly from the pilot study group to validate the response process. They were instructed to respond to a subset of items from the AMS-Chemistry, CCSS, and CMI. For the AMS-Chemistry, student responses revealed uncertainty regarding the terms “enter the job market” and “quest for knowledge”. To resolve this, “enter the job market” was revised to “finding a job”, and “quest for knowledge” was simplified to “wanting to know more”. For the CCSS, students indicated a lack of understanding of the term “particulate nature of matter”, so a descriptive explanation was included to clarify the item. In the CMI, the term “periodically” appeared in several items, which students found confusing; therefore, it was revised to “from time to time” as the subject matter experts who reviewed the items recommended.

4.2.4. Internal structure. The instruments were administered to 96 pre-university students from the pilot study for internal structure validation. Confirmatory factor analysis was conducted for this purpose. The analysis confirmed the factor structures of the CCSS and CMI as documented in their original versions. However, for the AMS-Chemistry, the findings differed from previous research, revealing a five-factor structure instead of the seven-factor structure reported by Liu et al. (2017). Notably, items related to the factors “to know,” “to accomplish,” and “to experience” from Liu et al.'s study (2017) consolidated into a single factor in the current study's pilot data. Consequently, this study recognised AMS-Chemistry as having a revised five-factor structure. The model fit for AMS-Chemistry was evaluated using several indices. The chi-square test was significant, χ²(340) = 606.95, p < 0.001, suggesting a poor fit; however, this may be due to the sensitivity of the chi-square statistic to smaller sample sizes (Kline, 2015). Other fit indices provided a more nuanced assessment. The comparative fit index (CFI = 0.98) and Tucker-Lewis index (TLI = 0.97) exceeded the recommended threshold of 0.90 (Hu and Bentler, 1999), indicating excellent incremental fit. The root mean square error of approximation (RMSEA = 0.06, 90% CI [0.08, 0.10]) suggested a moderate but acceptable fit, while the standardised root mean square residual (SRMR = 0.08) indicated a good fit (Hu and Bentler, 1999). Overall, the combination of these indices supports the conclusion that the proposed five-factor model adequately fits the data. Similar to AMS-Chemistry, the model fit for CCSS was evaluated using multiple indices. The chi-square test was significant, χ²(186) = 367.26, p < 0.001, suggesting a poor fit; however, this is likely due to the sensitivity of the chi-square statistic to smaller sample sizes. Other fit indices provided a more comprehensive assessment. The CFI = 0.97 and TLI = 0.97 exceeded the recommended threshold of 0.95, indicating excellent incremental fit. The RMSEA = 0.09, (90% CI [0.08, 0.11]) suggested a mediocre fit, which is considered acceptable in the context of smaller sample sizes (MacCallum et al., 1996; Chen et al., 2018). Additionally, the SRMR = 0.08 indicated a good fit (Hu and Bentler, 1999). Overall, these indices suggest that the proposed CCSS model adequately fits the data. Similarly, the model fit for CMI was evaluated using multiple indices. The chi-square test was significant, χ²(284) = 486.74, p < 0.001, suggesting a poor fit, likely due to the sensitivity of the chi-square statistic to smaller sample sizes. Other fit indices provided a more comprehensive assessment. The CFI = 0.96 and TLI = 0.97 exceeded the recommended threshold of 0.95, indicating excellent incremental fit. The RMSEA = 0.09, (90% CI [0.07, 0.10]) suggested a mediocre fit, which is considered acceptable in the context of smaller sample sizes (MacCallum et al., 1996; Chen et al., 2018). Additionally, the SRMR = 0.08 indicated a good fit. Overall, these indices suggest that the proposed CMI model adequately fits the data. Fig. 2 shows the confirmed factor structures mapped on the current study's conceptual framework.

Fig. 2 Mapping of factor structures of the study's conceptual framework.

4.2.5. Reliability. Reliability was assessed through internal consistency measures for all three (3) instruments. The data from 96 pre-university students from the pilot study were used for this assessment. The Cronbach's alpha of 0.94 for the AMS-Chemistry, closely aligning with previous research, which reported a reliability estimate of approximately 0.80 (Pratt et al., 2023). The factors of AMS-Chemistry showed Cronbach's alpha values ranging from 0.74 to 0.95. The CCSS demonstrated excellent internal consistency, with a Cronbach's alpha of 0.96. The alpha values for the CCSS factors ranged from 0.86 to 0.93, surpassing those reported by Ferrell and Barbera (2015), who found values ranging from 0.82 to 0.88. The CMI also displayed excellent internal consistency, with a Cronbach's alpha of 0.94, in line with Taasoobshirazi et al. (2015), who observed a reliability of around 0.87. The CMI factors had Cronbach's alpha values from 0.77 to 0.95. These reliability metrics confirm that the instruments used in this study consistently measure the intended constructs, minimising random error. Table 1 summarises the Cronbach's alpha values for the psychometric instruments obtained from the pilot study.

Table 1 Pilot study: Cronbach's alpha values of AMS-chemistry, CCSS, and CMI

Psychometric instrument	Cronbach's alpha
1. Academic motivation scale-chemistry (AMS-chemistry)
a. Factor A (AM-FA): amotivation	0.74
b. Factor B (AM-FB): external regulation	0.95
c. Factor C (AM-FC): introjected regulation	0.89
d. Factor D (AM-FD): identified regulation	0.95
e. Factor E (AM-FE): to experience, to accomplish, and to know	0.94
2. College chemistry self-efficacy scale (CCSS)
a. Factor A (SE-FA): self-efficacy for cognitive skills	0.93
b. Factor B (SE-FB): self-efficacy for psychomotor skills	0.93
c. Factor C (SE-FC): self-efficacy for everyday applications	0.86
3. Chemistry metacognition inventory (CMI)
a. Factor A (MC-FA): knowledge of cognition (declarative, procedural, and conditional)	0.92
b. Factor B (MC-FB): regulation of cognition (information management – chemistry diagrams)	0.94
c. Factor C (MC-FC): regulation of cognition (monitoring)	0.90
d. Factor D (MC-FD): regulation of cognition (evaluation)	0.95
e. Factor E (MC-FE): regulation of cognition (debugging)	0.81
f. Factor F (MC-FF): regulation of cognition (planning)	0.77

---

4.3. Procedure for data analysis

Partial least squares structural equation modeling (PLS-SEM) was employed using Smart-PLS 4.0 to address the research questions. PLS-SEM is a variance-based alternative to the commonly used covariance-based SEM (CB-SEM) for estimating relationships within a structural equation model (Hair et al., 2017; Cheah et al., 2024). Hair et al. (2017) further suggested that PLS-SEM should be used rather than CB-SEM when the research aims to predict key target constructs or extend an existing structural theory. Since this study focuses on extending structural theories related to self-efficacy, metacognition, and academic motivation in chemistry within the Malaysian context, PLS-SEM was deemed the appropriate analytical approach. Additionally, the greater statistical power of PLS-SEM compared to CB-SEM (Hair et al., 2021) further justified its selection. Hair et al. (2021) noted that higher statistical power increases the likelihood of detecting significant relationships when they exist in the population. Although PLS-SEM is less commonly applied in Chemistry Education Research Practice—only two studies have utilized it (Ross et al., 2018; 2020)—it is widely adopted in educational research (Hair and Alamer, 2022). Its use in STEM education is also growing (Karpudewan et al., 2023; Subramaniam et al., 2023).

PLS analysis involved two models: the measurement and structural models (Hair et al., 2017). The measurement model, which reflects the relationship between latent and observed variables, was tested through reliability and validity analyses. Latent variables are theoretical constructs that cannot be directly measured but are inferred from relationships among multiple observed variables, representing underlying traits, attitudes, or characteristics in a model (Hair et al., 2019). These latent variables are indirectly measured through observed variables, which are directly measurable data points reflecting the presence or extent of these constructs (Hair et al., 2019). In this study, the three (3) chemistry-related latent variables (i.e., constructs) are academic motivation (AM), self-efficacy (SE), and metacognition (MC). The observed chemistry-related variables (i.e., items) include academic motivation (measured as AM1 to AM28), self-efficacy (measured as SE1 to SE21), and metacognition (measured as MC1 to MC26).

On the other hand, the structural model tests the hypothesised relationships between the latent variables and incorporates the role of moderators, such as gender and locality, in these relationships. The structural model evaluates how the constructs influence one another, quantified by path coefficients, effect sizes, R², and Q² values (Hair et al., 2017). Specifically, the structural model in this study assesses whether self-efficacy (SE) and metacognition (MC) influence academic motivation (AM). Additionally, it explores whether gender and locality moderate the strength and direction of these relationships. This analysis helps to determine if the impact of self-efficacy (SE) or metacognition (MC) on academic motivation (AM) differs between male and female students or between students from rural and urban areas.

5. Research results

5.1. Reflective model measurements

The model, encompassing all items, was evaluated with the understanding that all indicators were reflective, as Hair et al. (2019) outlined. Analyses were conducted to examine individual item factor loading (λ), Dijkstra-Henseler's rho_A (rho_A), composite reliability (CR), and average variance extracted (AVE). As presented in Table 2, the variation indicates that not all items were assumed to contribute equally to their respective latent factors. Consequently, a uniform weighting approach for item aggregation was deemed inappropriate for this analysis. To ensure the accuracy and relevance of the calculated scores, a weighted composite method was employed, where each item's contribution to the overall score of a construct is weighted according to its factor loading. This approach acknowledges the differential impact of each item on the latent construct, aligning with the psychometric recommendations discussed by McNeish and Wolf (2020). Table 2 summarises the analysis of factor loadings and convergent validity and reliability results.

Table 2 Measurement items

Constructs	Items	λ	Convergent validity and reliability
			rho_A	CR	AVE
Academic motivation Factor A (AM-FA): amotivation	AM1	0.886	0.883	0.907	0.712
	AM2	0.709
	AM3	0.869
	AM4	0.897
Academic motivation Factor B (AM-FB): external regulation	AM5	0.858	0.960	0.959	0.855
	AM6	0.941
	AM7	0.954
	AM8	0.943
Academic motivation Factor C (AM-FC): introjected regulation	AM9	0.856	0.886	0.903	0.699
	AM10	0.841
	AM11	0.765
	AM12	0.878
Academic motivation Factor D (AM-FD): identified regulation	AM13	0.907	0.942	0.959	0.852
	AM14	0.927
	AM15	0.921
	AM16	0.938
Academic motivation Factor E (AM-FE): to experience, to accomplish, and to know	AM17	0.757	0.959	0.963	0.687
	AM18	0.765
	AM19	0.765
	AM20	0.825
	AM21	0.844
	AM22	0.852
	AM23	0.840
	AM24	0.844
	AM25	0.851
	AM26	0.874
	AM27	0.875
	AM28	0.842
Self-efficacy Factor A (SE-FA): self-efficacy for cognitive skills	SE1	0.773	0.930	0.940	0.612
	SE2	0.794
	SE3	0.758
	SE4	0.759
	SE5	0.789
	SE6	0.796
	SE7	0.799
	SE8	0.801
	SE9	0.759
	SE12	0.791
Self-efficacy Factor B (SE-FB): self-efficacy for psychomotor skills	SE10	0.807	0.927	0.941	0.696
	SE11	0.742
	SE13	0.821
	SE14	0.875
	SE15	0.856
	SE16	0.862
	SE17	0.870
Self-efficacy Factor C (SE-FC): self-efficacy for everyday applications	SE18	0.815	0.882	0.919	0.738
	SE19	0.839
	SE20	0.899
	SE21	0.882
Metacognition Factor A (MC-FA): knowledge of cognition (declarative, procedural, and conditional)	MC1	0.840	0.908	0.927	0.679
	MC2	0.837
	MC3	0.827
	MC4	0.818
	MC5	0.838
	MC6	0.783
Metacognition Factor B (MC-FB): regulation of cognition (information management – chemistry diagrams)	MC7	0.883	0.913	0.939	0.795
	MC8	0.922
	MC9	0.925
	MC10	0.833
Metacognition Factor C (MC-FC): regulation of cognition (monitoring)	MC11	0.854	0.907	0.934	0.779
	MC12	0.868
	MC13	0.892
	MC14	0.915
Metacognition Factor D (MC-FD): regulation of cognition (evaluation)	MC15	0.863	0.909	0.933	0.777
	MC16	0.898
	MC17	0.902
	MC18	0.860
Metacognition Factor E (MC-FE): regulation of cognition (debugging)	MC19	0.900	0.827	0.897	0.744
	MC20	0.901
	MC21	0.782
Metacognition Factor F (MC-FF): regulation of cognition (planning)	MC22	0.804	0.846	0.882	0.600
	MC23	0.814
	MC24	0.703
	MC25	0.728
	MC26	0.817

---

The analysis of convergent validity and reliability for the constructs of academic motivation (AM), self-efficacy (SE), and metacognition (MC) demonstrates strong support for the measurement model. Hair et al. (2019) suggest that factor loadings (λ) exceeding 0.7 are good, implying that the items effectively capture the construct they are designed to measure. Based on the data in Table 2, all items for the current study's three (3) latent variables (i.e., constructs) fulfilled this requirement. Moreover, in Table 2, the composite reliability (CR) and Dijkstra-Henseler's rho_A values are similarly high and exceed the recommended threshold of 0.7 (Hair et al., 2020), further supporting the constructs' reliability. The average variance extracted (AVE) values for the factors range from 0.600 to 0.855, suggesting that their respective constructs explain a substantial proportion of variance in the items (Hair et al., 2020). These results affirm that the measurement scales effectively capture the intended constructs, providing a solid foundation for further analysis in this study.

The discriminant validity of the constructs in this study was assessed using the Heterotrait–Monotrait (HTMT) ratio of correlations, as shown in Table 3. HTMT values below 0.85 indicate adequate discriminant validity, signifying that the constructs are distinct (Henseler et al., 2015). The constructs in this analysis demonstrated acceptable discriminant validity, as all HTMT values fell well below the 0.85 threshold. For instance, the relationship between AM-FA and MC-FA yielded a value of 0.234, and between MC-FB and AM-FB, a value of 0.082. i.e., both well within acceptable limits. Thus, the constructs' discriminant validity has been confirmed for this study.

Table 3 HTMT ratio of correlations

Factors	AM-FA	MC-FA	SE-FA	AM-FB	MC-FB	SE-FB	AM-FC	MC-FC	SE-FC	AM-FD	MC-FD	AM-FE	MC-FE	MC-FF
AM-FA = amotivation; AM-FB = external regulation; AM-FC = introjected regulation; AM-FD = identified regulation; AM-FE = to experience, to accomplish, and to know; MC-FA = knowledge of cognition (declarative, procedural, and conditional); MC-FB = regulation of cognition (information management – chemistry diagrams); MC-FC = regulation of cognition (monitoring); MC-FD = regulation of cognition (evaluation); MC-FE = regulation of cognition (debugging); MC-FF = regulation of cognition (planning); SE-FA = self-efficacy for cognitive skills; SE-FB = self-efficacy for psychomotor skills; SE-FC = self-efficacy for everyday applications
AM-FA
MC-FA	0.234
SE-FA	0.122	0.779
AM-FB	0.115	0.116	0.190
MC-FB	0.074	0.575	0.501	0.082
SE-FB	0.132	0.562	0.736	0.257	0.372
AM-FC	0.151	0.369	0.384	0.414	0.239	0.352
MC-FC	0.136	0.616	0.595	0.171	0.395	0.520	0.437
SE-FC	0.040	0.663	0.807	0.253	0.479	0.666	0.337	0.522
AM-FD	0.262	0.374	0.427	0.535	0.276	0.450	0.520	0.359	0.441
MC-FD	0.223	0.477	0.479	0.178	0.381	0.456	0.373	0.520	0.410	0.370
AM-FE	0.316	0.517	0.563	0.260	0.426	0.512	0.657	0.538	0.519	0.643	0.472
MC-FE	0.215	0.419	0.404	0.202	0.382	0.454	0.358	0.517	0.386	0.421	0.602	0.452
MC-FF	0.177	0.675	0.664	0.199	0.516	0.558	0.438	0.693	0.583	0.427	0.658	0.540	0.692

---

5.2. Structural model measurements

It is essential to confirm that there are no multicollinearity issues (collinearity between predictor and criterion variables) before assessing the structural model, as these can distort the model's causal effects. To do so, the inner model's variance inflation factor (i.e., inner VIF) was assessed. The literature indicates that inner VIF values below 5.0 for the constructs suggest that lateral multicollinearity is not an issue (Hair et al., 2017). Table 4 displays the results of the lateral collinearity test. The inner VIF values for all constructs listed in Table 4 are below 5.0, indicating that multiple collinearities do not pose an issue for this study (Hair et al., 2017).

Table 4 The variance inflation factor (VIF) values for the inner model

Factors	AM-FA	MC-FA	SE-FA	AM-FB	MC-FB	SE-FB	AM-FC	MC-FC	SE-FC	AM-FD	MC-FD	AM-FE	MC-FE	MC-FF
AM-FA = amotivation; AM-FB = external regulation; AM-FC = introjected regulation; AM-FD = identified regulation; AM-FE = to experience, to accomplish, and to know; MC-FA = knowledge of cognition (declarative, procedural, and conditional); MC-FB = regulation of cognition (information management – chemistry diagrams); MC-FC = regulation of cognition (monitoring); MC-FD = regulation of cognition (evaluation); MC-FE = regulation of cognition (debugging); MC-FF = regulation of cognition (planning); SE-FA = self-efficacy for cognitive skills; SE-FB = self-efficacy for psychomotor skills; SE-FC = self-efficacy for everyday applications.
AM-FA
MC-FA	2.512			2.512			2.512			2.512		2.512
SE-FA	3.534			3.534			3.534			3.534		3.534
AM-FB
MC-FB	1.482			1.482			1.482			1.482		1.482
SE-FB	2.083			2.083			2.083			2.083		2.083
AM-FC
MC-FC	1.869			1.869			1.869			1.869		1.869
SE-FC	2.333			2.333			2.333			2.333		2.333
AM-FD
MC-FD	1.710			1.710			1.710			1.710		1.710
AM-FE
MC-FE	1.686			1.686			1.686			1.686		1.686
MC-FF	2.471			2.471			2.471			2.471		2.471

---

The results of the structural model assessment are summarised in Tables 5–7, highlighting the significant relationships between the predictors (i.e., self-efficacy and metacognition) and academic motivation. Table 5 displays the path coefficients (i.e., direct hypothesis testing), where self-efficacy (β = 0.288) and metacognition (β = 0.412) strongly influence academic motivation in chemistry, with T statistics exceeding the threshold for significance and p-values of p < 0.001. Table 6 provides the metrics, revealing that the model accounts for approximately 42.3% of the variance in academic motivation, as evidenced by the R² value of 0.423. The R² value of above 0.260 indicates the model exhibits substantial level of predictive accuracy (Cohen, 1988). Effect sizes of f² = 0.070 for self-efficacy and f² = 0.142 metacognition demonstrates small and medium effect sizes of the predictive constructs on AM (Cohen, 1988). Finally, Table 7 presents the cross-validated redundancy analysis, which indicates a Q² value of 0.179 for academic motivation, signifying the predictive relevance of the model (Hair et al., 2019). These results underscore the importance of self-efficacy and metacognition in enhancing academic motivation among Malaysian pre-university students, specifically in chemistry.

Table 5 Direct hypothesis testing

Hypotheses	β	M	SD	t	p	Decision
β = path coefficient; M = sample mean; SD = standard deviation; t = T statistics; p = P Values.
H1: self-efficacy → academic motivation	0.288	0.285	0.063	4.564	<0.001	Supported
H2: metacognition → academic motivation	0.412	0.425	0.058	7.136	<0.001	Supported

---

Table 6 Model fit and effect sizes

Metric	Value
R^2 for academic motivation	0.423
Adjusted R^2 for academic motivation	0.421
f^2 (effect sizes):
H1: self-efficacy → academic motivation	0.070
H2: metacognition → academic motivation	0.142

---

Table 7 Cross-validated redundancy (Q²)

Constructs	SSO	SSE	Q^2 (=1-SSE/SSO)
SSO = sum of squared observations; SSE = sum of squared errors.
Academic motivation	2780	2282.816	0.179
Self-efficacy	1668	1668
Metacognition	3336	3336

---

5.3. Moderating effects

The moderating effects of gender and locality on the relationships between self-efficacy, metacognition, and academic motivation in chemistry were examined using R² values and path coefficients. The inclusion of gender as a moderating factor slightly increased the explained variance in academic motivation for self-efficacy and metacognition models, with R² values increasing from 0.387 to 0.388 and 0.359 to 0.363, respectively. Similarly, including locality resulted in minor increases in R² values for both models. Table 8 presents the R² values analysis. However, path coefficient analysis revealed that only the moderating effect of gender on the relationship between self-efficacy and academic motivation was statistically significant (β = −0.07, p = 0.023), suggesting that gender plays a notable role in shaping the impact of self-efficacy on academic motivation. In contrast, neither gender nor locality had significant moderating effects on the relationship between metacognition and academic motivation, nor did locality moderate the link between self-efficacy and academic motivation. These findings suggest that while gender influences the relationship between self-efficacy and academic motivation in chemistry among Malaysian pre-university students, the moderating roles of gender and locality in other relationships within the model are minimal. Table 9 summarises the findings.

Table 8 R² values analysis

Variable	R^2 Excluded	R^2 Included	Difference
Self-efficacy
Gender as moderator	0.359	0.363	0.004
Locality as moderator	0.352	0.354	0.002
Metacognition
Gender as moderator	0.387	0.388	0.001
Locality as moderator	0.390	0.392	0.002

---

Table 9 Path analysis for coefficient (hypothesis testing – moderated by gender and locality)

Hypotheses	β	M	SD	t	p	Decision
β = path coefficient; M = sample mean; SD = standard deviation; t = T statistics; p = P values.
H3: self-efficacy → academic motivation (moderated by gender)	−0.07	−0.08	0.03	1.99	0.02	Supported
H4: metacognition → academic motivation (moderated by gender)	−0.03	−0.05	0.05	0.51	0.30	Rejected
H5: self-efficacy → academic motivation (moderated by locality)	−0.05	−0.03	0.06	0.78	0.22	Rejected
H6: metacognition → academic motivation (moderated by locality)	−0.05	−0.02	0.08	0.62	0.27	Rejected

---

6. Discussion

The findings support hypothesis H1, indicating a significant positive effect of self-efficacy on academic motivation in chemistry among Malaysian pre-university students (β = 0.288, p < 0.001). This aligns with social cognitive theory, which posits that individuals with higher self-efficacy are more motivated to engage in tasks, persevere through challenges, and set ambitious goals (Bandura, 1986). The current results echo Alci's (2015) findings, which showed a positive correlation between self-efficacy and intrinsic and extrinsic motivation in chemistry undergraduates. In the Malaysian context, this relationship may be attributed to the structured, academically rigorous nature of the pre-university science programme, which requires students to develop a strong belief in their abilities to succeed (Ministry of Education Malaysia, 2024a). Social factors in Malaysia, where academic success is highly valued, may further enhance students' confidence and drive, especially among those who consistently perform well in science subjects (Wong et al., 2021).

Hypothesis H2 is also supported, with metacognition having a significant and stronger positive impact on academic motivation (β = 0.412, p < 0.001) than self-efficacy. This finding aligns with existing research suggesting that metacognitive awareness and regulation promote academic motivation by enabling students to plan, monitor, and evaluate their learning effectively (Landine and Stewart, 1998; Abdelrahman, 2020). In Malaysia, metacognition's substantial impact on motivation may be particularly relevant due to the challenging content of pre-university chemistry (Matriculation Division, 2022a, b), which necessitates careful planning and self-monitoring. Malaysian students are often encouraged to take ownership of their learning, which can foster metacognitive skills (Choy et al., 2020). These skills allow students to navigate complex chemistry topics more confidently, preventing burnout and helping them achieve a sense of progress and engagement.

The results support hypothesis H3, showing that gender statistically significantly moderates the relationship between self-efficacy and academic motivation (β = −0.07, p = 0.023). Female students in this study demonstrated a stronger link between self-efficacy and academic motivation than male students, consistent with findings in other STEM fields, such as those reported by Wang and Yu (2023). Interestingly, while higher self-efficacy among male students was associated with slightly lower academic motivation than females, this trend may reflect sociocultural influences unique to Malaysia. In Malaysia, societal expectations often encourage male students to project confidence and self-assurance (Saadat and Sultana, 2023); however, this outward confidence does not necessarily translate into greater motivation, particularly if they lack genuine engagement with the subject. Conversely, the pre-university educational environment in Malaysia, which includes supportive educators and challenging assignments, appears especially influential for female students. Opportunities for autonomy in learning and exposure to positive role models can further enhance their motivation and academic performance, highlighting the critical role of self-efficacy in fostering their academic engagement (Khalid and Rahman, 2023).

Hypothesis H4 is not supported, as gender did not significantly moderate the relationship between metacognition and academic motivation among Malaysian pre-university students (β = −0.03, p = 0.30) despite the study's expectations based on prior literature. This suggests that gender differences in motivation (Salta and Koulougliotis, 2020; Meydan, 2021; Ardura et al., 2023; Salsabila and Huda, 2023) and metacognition (Jumani et al., 2010; Ahmed et al., 2019; Adiansyah et al., 2021) exist independently. The findings do not support hypothesis H5. In contrast to the study's expectations informed by prior literature, locality did not significantly moderate the relationship between self-efficacy and academic motivation among pre-university chemistry students (β = −0.05, p = 0.22). Similarly, hypothesis H6 is not supported. Contrary to expectations and despite documented locality differences in academic motivation and metacognition, the findings reveal no significant moderation by locality on the relationship between metacognition and academic motivation in chemistry among Malaysian pre-university students (β = −0.05, p = 0.27). Although previous studies suggest that locality can independently influence self-efficacy, metacognition, and academic motivation, with urban students often reporting higher self-efficacy (Ibrahim et al., 2019; Astalini et al., 2020), metacognition (Taghieh et al., 2019; Alam, 2020), and motivation (Saleh, 2014; Zhang and Zhou, 2023) due to better access to resources (Pimvichai et al., 2015; Hill et al., 2018; Hariyati et al., 2021), these effects did not extend to moderating the relationship between self-efficacy and metacognition with academic motivation in this study. The absence of moderating effects of locality on self-efficacy and metacognition's impact on academic motivation can be attributed to the homogenising influence of Malaysia's educational policies. The standardised pre-university chemistry curriculum (Matriculation Division, 2022a, b) and centralised STEM education policies (Ministry of Education Malaysia, 2024b) likely create an equalisation effect, minimising disparities. This environment promotes systematic study habits and strategic problem-solving approaches, reducing notable differences in academic motivation across localities and possibly genders.

In summary, this study confirms the significant roles of self-efficacy and metacognition in predicting academic motivation among Malaysian pre-university students, with metacognition having a more substantial influence. Gender significantly moderates the relationship between self-efficacy and motivation, highlighting the need for educational strategies that support both genders differently in STEM fields, especially chemistry. Although locality does not moderate the relationship between self-efficacy and motivation or between metacognition and motivation, the findings indicate that Malaysian pre-university students benefit from a relatively equalised educational experience. This is guided by the standardised pre-university curriculum and centralised STEM education policies. Educators should consider emphasising metacognitive strategies in the curriculum to enhance academic motivation, especially for challenging subjects like chemistry. Additionally, interventions to enhance self-efficacy may be particularly beneficial for female students, as this could further support their motivation in science-related subjects.

7. Conclusion

This study aimed to investigate the relationships among chemistry self-efficacy, metacognition, and academic motivation, focusing on the moderating roles of gender and locality among Malaysian pre-university students. The findings revealed that self-efficacy significantly influences academic motivation, particularly among female students, while metacognition consistently impacts motivation across genders and localities. Gender significantly moderated the self-efficacy-motivation relationship, whereas locality did not significantly moderate the relationships between self-efficacy, metacognition, and academic motivation.

Theoretically, these results contribute to the ongoing discourse on academic motivation in chemistry education by highlighting the role of self-efficacy and metacognition as crucial predictors. The study reinforces existing theories suggesting that self-efficacy is a critical motivational driver, significantly when moderated by gender, adding nuance to our understanding of how male and female students engage with chemistry. This finding challenges traditional assumptions of gender-neutral self-efficacy models, suggesting that gender-specific interventions may be more effective.

Practically, the findings suggest that chemistry educators should focus on enhancing self-efficacy, particularly for female students, to foster stronger academic motivation. Additionally, integrating metacognitive strategies into the curriculum can benefit all students, providing tools to monitor and regulate their learning effectively. These insights could inform the design of gender-sensitive educational interventions and targeted teaching practices to improve motivation, confidence, and academic success in chemistry.

Despite its contributions, this study has limitations that may affect the generalisability of the findings. Firstly, focusing on a specific group of pre-university students limits applicability across broader educational contexts and diverse demographics. The relatively homogeneous sample and geographical scope may not capture varied educational experiences or motivational factors among different populations. Secondly, this study's cross-sectional design provides a snapshot rather than a longitudinal understanding of how self-efficacy and metacognition interact with academic motivation over time. Several methodological steps were taken to address these limitations and enhance the validity of the findings. The study administered well-validated chemistry psychometric instruments (i.e., AMS-Chemistry, CCSS, and CMI), ensuring reliable and accurate measurements across the target demographic. A pilot study involving a comparable pre-university student group was conducted to refine the instruments through cognitive interviews and expert reviews, confirming their relevance and applicability to Malaysian students. Finally, using PLS-SEM allowed for a detailed examination of the relationships among variables, including moderating effects, to ensure that findings accurately reflect the influences of self-efficacy and metacognition on academic motivation. The third limitation of this study is the relatively small sample size used for the CFA. Despite this, the CFA results indicate adequate model fit indices (MacCallum et al., 1996; Chen et al., 2018). Increasing the sample size would further enhance the robustness of the model fit.

Future research should address these limitations by including more diverse samples across educational levels and contexts to enhance generalisability. Longitudinal studies are also recommended to examine changes in the relationships among self-efficacy, metacognition, and academic motivation over time. Moreover, investigating the effects of educational interventions targeting self-efficacy and metacognition may provide insights into how these factors can be effectively enhanced to improve academic motivation in chemistry and other STEM fields. Such studies could provide valuable data on tailoring interventions to different demographic groups, further informing gender-sensitive and location-responsive educational practices.

In conclusion, this study offers valuable theoretical and practical insights into the key factors influencing academic motivation in chemistry, laying the groundwork for future research and educational strategies to improve student engagement and learning outcomes.

8. Ethics

In conducting this research, we adhered strictly to the ethical standards set forth by the Universiti Sains Malaysia Human Ethics Committee under JEPeM Code: USM/JEPeM/PP/24070603. Participants' confidentiality and anonymity were safeguarded throughout the study, and informed consent was obtained before data collection. The study ensured voluntary participation, allowing individuals to withdraw at any stage without repercussions. All procedures were designed to minimise potential risks and discomfort to participants, ensuring their rights and welfare remained prioritised. The research was conducted following beneficence, respect for persons, and justice.

Data availability

The data that support the findings of this study are not publicly available due to restrictions set by the participating institution.

Conflicts of interest

There are no conflicts to declare.

References

1. Abdelrahman R. M., (2020), Metacognitive awareness and academic motivation and their impact on academic achievement of Ajman University students, Heliyon, 6(9), e04192 DOI:10.1016/j.heliyon.2020.e04192.
2. Abdolrezapour P., Jahanbakhsh Ganjeh S. and Ghanbari, N., (2023), Self-efficacy and resilience as predictors of students’ academic motivation in online educationm, PLoS One, 18(5), e0285984 DOI:10.1371/journal.pone.0285984.
3. Adiansyah R., Corebima A. D., Zubaidah S. and Rohman F., (2021), The correlation between metacognitive skills and scientific attitudes towards the retention of male and female students in South Sulawesi, Indonesia, Int. J. Eval. Res. Educ., 10(4), 1272–1281 DOI:10.11591/ijere.v10i4.21597.
4. Ahmed N., Wiyarsi A. and Senam R. N., (2019), Comparison of students’ metacognitive skills by gender on chemical bonding in chemistry, J. Pendidikan Sains, 7(2), 137–146 DOI:10.26714/JPS.7.2.2019.137-146.
5. Alam M. A., (2020), Metacognitive ability and academic achievement in biology, Bhartiyam Int. J. Educ. Res., 9(2), 1–16.
6. Alci B., (2015), The influence of self-efficacy and motivational factors on academic performance in general chemistry course: a modelling study, Educ. Res. Rev., 10(4), 453–461 DOI:10.5897/ERR2014.2003.
7. Allen A., Conner B., Gantman B., Warner K., Nash R., Janes B., Hellum M. and Singer C., (2022), Developing a well-received pre-matriculation program: the evolution of MedFIT, Discover Educ., 1(1), 12 DOI:10.1007/s44217-022-00012-z.
8. Amaliyah S., Suryaningsih S. and Yunita L., (2021), Gender differences in the relationship between anxiety, self-efficacy and students learning outcomes on chemistry subject, EDUSAINS, 13(1), 8–14 DOI:10.15408/ES.V13I1.12991.
9. Ardura D., Zamora Á. and Pérez-Bitrián A., (2021), The role of motivation on secondary school students’ causal attributions to choose or abandon chemistry, Chem. Educ. Res. Pract., 22(1), 43–61 10.1039/D0RP00168F.
10. Ardura D., Zamora Á. and Pérez-Bitrián A., (2023), On the effect of gender on secondary school students’ causal attributions to choose or abandon physics & chemistry, Chem. Educ. Res. Pract., 24(4), 1174–1189 10.1039/D3RP00070B.
11. Asfaw M. D., (2022), Assessment on students’ self-efficacy, academic achievement, locality and gender in chemistry at Woreillu Secondary School, Ethiopia, J. Addict. Res. Ther., 13(6), 473 DOI:10.4172/2155-6105.100473.
12. Astalini A., Darmaji M. I., Kuswanto R. P., Anggraini L. and Putra I., (2020), Attitude and self-confidence students in learning natural sciences: rural and urban junior high school, Universal J. Educ. Res., 8(6), 2569–2577 DOI:10.13189/ujer.2020.080640.
13. Bakkaloglu S., (2020), Analysis of metacognitive awareness of primary and secondary school students in terms of some variables, J. Educ. Learn., 9(1), 156–163 DOI:10.5539/jel.v9n1p156.
14. Bandura A., (1986), Social foundations of thought and action: A social cognitive theory, Prentice-Hall.
15. Bolshakova V. L., Johnson C. C. and Czerniak C. M., (2011), “It depends on what science teacher you got”: urban science self-efficacy from teacher and student voices, Cultural Stud. Sci. Educ., 6, 961–997 DOI:10.1007/s11422-011-9346-2.
16. Brown A. L., (1987), Metacognition, executive control, self-regulation, and other more mysterious mechanisms, in Weinert F. E. and R. Kluwe (ed.), Metacognition, motivation, and understanding, L. Erlbaum Associates, pp. 65–116.
17. Cayubit R. F. O., (2022), Why learning environment matters? An analysis on how the learning environment influences the academic motivation, learning strategies and engagement of college students, Learn. Environ. Res., 25(2), 581–599 DOI:10.1007/s10984-021-09382-x.
18. Cheah J. H., Magno F. and Cassia F., (2024), Reviewing the SmartPLS 4 software: the latest features and enhancements, J. Mark. Anal., 12, 97–107 DOI:10.1057/s41270-023-00266-y.
19. Chen F., Curran P. J., Bollen K. A., Kirby J. and Paxton P., (2008), An empirical evaluation of the use of fixed cutoff points in RMSEA test statistic in structural equation models, Sociol. Methods Res., 36(4), 462–494 DOI:10.1177/0049124108314720.
20. Chittleborough G. and Treagust D., (2008), Correct interpretation of chemical diagrams requires transforming from one level of representation to another, Res. Sci. Educ., 38, 463–482 DOI:10.1007/s11165-007-9059-4.
21. Choy S. C., Yim J. S. C. and Tan P. L., (2020), A metacognitive knowledge, metacognitive experience, and its effects on learning outcomes for STEM and non-STEM Malaysian students, Int. J. Adv. Res. Educ. Soc., 2(1), 1–14. https://myjms.mohe.gov.my/index.php/ijares/article/view/8302.
22. Cohen J., (1988), Statistical power analysis for the behavioural sciences, 2nd edn, Hillsdale, NJ: Lawrence Erlbaum Associates.
23. Cousins A. and Mills M., (2015), Gender and high school chemistry: student perceptions on achievement in a selective setting, Camb. J. Educ., 45(2), 187–204 DOI:10.1080/0305764X.2014.934202.
24. Dalgety J., Coll R. K. and Jones A., (2003), Development of chemistry attitudes and experiences questionnaire (CAEQ), J. Res. Sci. Teach., 40(7), 649–668 DOI:10.1002/tea.10103.
25. Dierendonck C., Tóth-Király I., Morin A., Kerger S., Milmeister P. and Poncelet D., (2021), Testing associations between global and specific levels of student academic motivation and engagement in the classroom, J. Exp. Educ., 91, 101–124 DOI:10.1080/00220973.2021.1913979.
26. Eccles J. S. and Wigfield A., (2020), From expectancy-value theory to situated expectancy-value theory: a developmental, social cognitive, and sociocultural perspective on motivation, Contemp. Educ. Psychol., 61, 101859.
27. Elford D., Lancaster S. J. and Jones G. A., (2022), Fostering motivation toward chemistry through augmented reality educational escape activities. A self-determination theory approach, J. Chem. Educ., 99(10), 3406–3417 DOI:10.1021/acs.jchemed.2c00428.
28. Ferrell B. and Barbera J., (2015), Analysis of students' self-efficacy, interest, and effort beliefs in general chemistry, Chem. Educ. Res. Pract., 16(2), 318–337 10.1039/C4RP00152D.
29. Ferrell B., Phillips M. M. and Barbera J., (2016), Connecting achievement motivation to performance in general chemistry, Chem. Educ. Res. Pract., 17(4), 1054–1066 10.1039/C6RP00148C.
30. Flavell J. H., (1979), Metacognition and cognitive monitoring: a new area of cognitive-developmental inquiry, Am. Psychol., 34(10), 906.
31. Glynn S. M., Brickman P., Armstrong N. and Taasoobshirazi G., (2011), Science Motivation Questionnaire II: validation with science majors and nonscience majors, J. Res. Sci. Teach., 48(10), 1159–1176 DOI:10.1002/tea.20442.
32. Graham S., (2020), An attributional theory of motivation, Contemp. Educ. Psychol., 61, 101861 DOI:10.1016/j.cedpsych.2020.101861.
33. Graham K. J., Bohn-Gettler C. M. and Raigoza A. F., (2019), Metacognitive training in chemistry tutor sessions increases first year students’ self-efficacy, J. Chem. Educ., 96(8), 1539–1547 DOI:10.1021/acs.jchemed.9b00170.
34. Guo X., Deng W., Hu K., Lei W., Xiang S. and Hu W., (2022), The effect of metacognition on students’ chemistry identity: the chain mediating role of chemistry learning burnout and chemistry learning flow, Chem. Educ. Res. Pract., 23(2), 408–421 10.1039/D1RP00342A.
35. Hair J. and Alamer A., (2022), Partial Least Squares Structural Equation Modeling (PLS-SEM) in second language and education research: guidelines using an applied example, Res. Methods Appl. Linguistics, 3(1), 100027 DOI:10.1016/j.rmal.2022.100027.
36. Hair J. F., Hult G. T. M., Ringle C. M. and Sarstedt M., (2017), A primer on partial least squares structural equation modeling (PLS-SEM), Sage.
37. Hair J. F., Hult G. T. M., Ringle C. M., Sarstedt M., Danks N. P. and Ray S., (2021), Partial least squares structural equation modeling (PLS-SEM) using R: A workbook, Springer Nature DOI:10.1007/978-3-030-80519-7.
38. Hair J. F., Risher J. J., Sarstedt M. and Ringle C. M., (2019), When to use and how to report the results of PLS-SEM, Eur. Bus. Rev., 31(1), 2–24 DOI:10.1108/EBR-11-2018-0203.
39. Hair Jr J. F., Howard M. C. and Nitzl C., (2020), Assessing measurement model quality in PLS-SEM using confirmatory composite analysis, J. Bus. Res., 109, 101–110 DOI:10.1016/j.jbusres.2019.11.069.
40. Hariyati N., Wagino W. and Mudjito M., (2021), Investigating virtual learning on students learning outcomes in urban and rural areas, Dinamika Pendidikan, 16(1), 54–63 DOI:10.15294/dp.v16i1.28661.
41. Henseler J., Ringle C. M. and Sarstedt M., (2015), A new criterion for assessing discriminant validity in variance-based structural equation modelling, J. Acad. Mark. Sci., 43, 115–135 DOI:10.1007/s11747-014-0403-8.
42. Hill P. W., McQuillan J., Hebets E. A., Spiegel A. N. and Diamond J., (2018), Informal science experiences among urban and rural youth: exploring differences at the intersections of socioeconomic status, gender and ethnicity, J. STEM Outreach, 1(1) DOI:10.15695/jstem/v1i1.28.
43. Hong W. H., Vadivelu J., Daniel E. G. S. and Sim J. H., (2015), Thinking about thinking: changes in first-year medical students’ metacognition and its relation to performance, Med. Educ. Online, 20(1), 27561 DOI:10.3402/meo.v20.27561.
44. Hu L. T. and Bentler P. M., (1999), Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives, Struct. Equ. Modeling, 6(1), 1–55 DOI:10.1080/10705519909540118.
45. Ibrahim N., A Zakiang M. A. and Damio S. M., (2019), Attitude in learning physics among form four students, Soc. Manag. Res. J., 16(2), 21–42 DOI:10.24191/smrj.v16i2.7060.
46. Janeck A. S., Calamari J. E., Riemann B. C. and Heffelfinger S. K., (2003), Too much thinking about thinking?: metacognitive differences in obsessive–compulsive disorder, J. Anxiety Disord., 17(2), 181–195 DOI:10.1016/S0887-6185(02)00198-6.
47. Jumani N. B., Chaudry M. A. and Abbasi F., (2010), Impact of metacognitive awareness on performance of students in chemistry, Contemp. Issues Educ. Res., 3(10), 39–44. https://files.eric.ed.gov/fulltext/EJ1072651.pdf.
48. Karpudewan M., Krishnan P. and Roth W. M. et al., (2023), What research says about the relationships between Malaysian teachers’ knowledge, perceived difficulties and self-efficacy, and practicing STEM teaching in schools. Asia-Pacific Educ. Res., 32, 353–365 DOI:10.1007/s40299-022-00658-1.
49. Khalid S. A. and Rahman A., (2023), Academic self-efficacy, intrinsic motivation and academic achievement: moderating effect of gender, Int. J. Acad. Res. Bus. Soc. Sci., 13, 1085–1098 DOI:10.6007/ijarbss/v13-i1/16043.
50. Kline R. B., (2015), Principles and practice of structural equation modeling, (4th edn), Guilford Press.
51. Landine J. and Stewart J., (1998), Relationship between metacognition, motivation, locus of control, self-efficacy, and academic achievement, Can. J. Couns. Psychother., 32(3), 200–212, https://cjc-rcc.ucalgary.ca/article/view/58606.
52. Lavi R., Shwartz G. and Dori Y. J., (2019), Metacognition in chemistry education: a literature review, Isr. J. Chem., 59(6–7), 583–597 DOI:10.1002/ijch.201800087.
53. Leong K. E., Tan P. P., Lau P. L. and Yong S. L., (2018), Exploring the relationship between motivation and science achievement of secondary students, Pertanika J. Soc. Sci. Hum., 26(4), 2243–2258, https://eprints.um.edu.my/id/eprint/21487.
54. Lewis S. E., (2022), Considerations on validity for studies using quantitative data in chemistry education research and practice, Chem. Educ. Res. Pract., 23(4), 764–767 10.1039/D2RP90009B.
55. Liu Y., (2021), The effects of online organic chemistry laboratory videos on students’ perceptions and intrinsic motivation, J. Res. Sci. Math. Technol. Educ., 4(3), 239–255 DOI:10.31756/jrmste.435.
56. Liu Y., Ferrell B., Barbera J. and Lewis J. E., (2017), Development and evaluation of a chemistry-specific version of the academic motivation scale (AMS-Chemistry), Chem. Educ. Res. Pract., 18(1), 191–213 10.1039/C6RP00200E.
57. Luszczynska A., Gutiérrez-Doña B. and Schwarzer R., (2005), General self-efficacy in various domains of human functioning: evidence from five countries, Int. J. Psychol., 40(2), 80–89 DOI:10.1080/00207590444000041.
58. MacCallum R. C., Browne M. W. and Sugawara H. M., (1996), Power analysis and determination of sample size for covariance structure modelling, Psychol. Methods, 1(2), 130–149.
59. Matriculation Division, (2022a), Curriculum specifications chemistry (SK015) [Fact sheet], Malaysia Ministry of Education.
60. Matriculation Division, (2022b), Curriculum specifications chemistry (SK025) [Fact sheet], Malaysia Ministry of Education.
61. McAlpin J. D., Kulatunga U. and Lewis J. E., (2023), Using social influence models to characterize student interest in a general chemistry peer-led team learning setting, Chem. Educ. Res. Pract., 24(3), 1003–1024 10.1039/D2RP00296E.
62. McNeish D. and Wolf M. G., (2020), Thinking twice about sum scores, Behav. Res. Methods, 52, 2287–2305 DOI:10.3758/s13428-020-01398-0.
63. Merchán Garzón D. F., Huertas Bustos A. P. and Ugarte Lizarazo J. O., (2020), Relationship between metacognitive skills, gender, and level of schooling in high school students, Suma Psicológica, 27(1), 9–17 DOI:10.14349/sumapsi.2020.v27.n1.2.
64. Meydan E., (2021), Investigating secondary school students' motivation for chemistry class in terms of various variables, Int. J. Prog. Educ., 17(1), 498–512 DOI:10.29329/ijpe.2020.329.31.
65. Ministry of Education Malaysia, (2024a), (November 10), Matriculation, https://www.moe.gov.my/pendidikan/lepas-menengah/matrikulasi.
66. Ministry of Education Malaysia, (2024b), (November 10), Science, Technology, Engineering, and Mathematics (STEM) Education, https://www.moe.gov.my/pendidikan-sains-teknologi-kejuruteraan-dan-matematik.
67. Moreno C., Pham D. and Ye L., (2021), Chemistry self-efficacy in lower-division chemistry courses: changes after a semester of instruction and gaps still remain between student groups, Chem. Educ. Res. Pract., 22(3), 772–785 10.1039/D0RP00345J.
68. Muteti C. Z., Jacob B. I. and Mutambuki J. M., (2023), Metacognition instruction enhances equity in effective study strategies across demographic groups in the general chemistry I course, Chem. Educ. Res. Pract., 24(4), 1204–1218 10.1039/D3RP00103B.
69. Naim M. M. N. and Karpudewan M., (2024), STEM-PT Traveler, a game-based approach for learning elements of the periodic table: an approach for enhancing secondary school students’ motivation for learning chemistry, Chem. Educ. Res. Pract., 25, 1251–1267 10.1039/D4RP00032C.
70. Pimvichai J., Yuenyong C., Thomas G. P. and Art-In S., (2015), Comparison of metacognitive orientation between Thai urban and rural physics classrooms, Mediterr. J. Soc. Sci., 6(3), 193–205 DOI:10.5539/jel.v9n1p156.
71. Pinkham A. E., (2019), Metacognition in psychosis, J. Exp. Psychopathol., 10(2), 2043808719841146 DOI:10.1177/2043808719841146.
72. Pratt J. M., Stewart J. L., Reisner B. A., Bentley A. K., Lin S., Smith S. R. and Raker J. R., (2023), Measuring student motivation in foundation-level inorganic chemistry courses: a multi-institution study, Chem. Educ. Res. Pract., 24(1), 143–160 10.1039/D2RP00199C.
73. Rieder A., Eseryel U. Y., Lehrer C. and Jung R., (2021), Why users comply with Wearables: the role of contextual self-efficacy in behavioral change, Int. J. Hum.–Comput. Interact., 37(3), 281–294 DOI:10.1080/10447318.2020.1819669.
74. Ross J., Guerra E. and Gonzalez-Ramos S., (2020), Linking a hierarchy of attitude effect to student engagement and chemistry achievement, Chem. Educ. Res. Pract., 21(1), 357–370 10.1039/C9RP00171A.
75. Ross J., Nuñez L. and Lai C. C., (2018), Partial least squares structural equation modeling of chemistry attitude in introductory college chemistry, Chem. Educ. Res. Pract., 19(4), 1270–1286 10.1039/C7RP00238F.
76. Ryan R. M. and Deci E. L., (2020), Intrinsic and extrinsic motivation from a self-determination theory perspective: definitions, theory, practices, and future directions, Contemp. Educ. Psychol., 61, 101860 DOI:10.1016/j.cedpsych.2020.101860.
77. Saadat Z. and Sultana A. M., (2023), Understanding gender disparity: factors affecting higher education self-efficacy of students in Malaysia, J. Sci. Technol. Policy Manage., 15(6), 1157–1181 DOI:10.1108/jstpm-10-2022-0165.
78. Saleh S., (2014), Malaysian students' motivation towards physics learning, Eur. J. Sci. Math. Educ., 2(4), 223–232. https://eric.ed.gov/?id=EJ1107653.
79. Salsabila T. and Huda N., (2023), Chemistry motivation of high school students in Indonesia: grade level and gender differences, J. Penelitian Pendidikan IPA, 9(12), 10845–10852 DOI:10.29303/jppipa.v9i12.6071.
80. Salta K. and Koulougliotis D., (2015), Assessing motivation to learn chemistry: adaptation and validation of Science Motivation Questionnaire II with Greek secondary school students. Chem. Educ. Res. Pract., 16(2), 237–250 10.1039/C4RP00196F.
81. Salta K. and Koulougliotis D., (2020), Domain specificity of motivation: chemistry and physics learning among undergraduate students of three academic majors, Int. J. Sci. Educ., 42(2), 253–270 DOI:10.1080/09500693.2019.1708511.
82. Salta K. and Koulougliotis D., (2022), Exploring factors that affect undergraduate students’ motivation to learn chemistry and physics, J. Balt. Sci. Educ., 21(6A), 1191–1204, https://oaji.net/articles/2022/987-1672585587.pdf.
83. Saw G. K. and Agger C. A., (2021), STEM pathways of rural and small-town students: opportunities to learn, aspirations, preparation, and college enrolment, Educ. Res., 50(9), 595–606 DOI:10.3102/0013189X211027528.
84. Saxena A. S., (2020), Impact of metacognitive strategies on self-regulated learning and intrinsic motivation, J. Psychosoc. Res., 15(1), 35–46 DOI:10.32381/JPR.2020.15.01.3.
85. Schraw G., (2001), Promoting general metacognitive awareness, Metacognition in Learning and Instruction: Theory, Research and Practice, pp. 3–16 DOI:10.1007/978-94-017-2243-8_1.
86. Schraw G. and Moshman D., (1995), Metacognitive theories, Educ. Psychol. Rev., 7, 351–371 DOI:10.1007/BF02212307.
87. Schunk D. H. and DiBenedetto M. K., (2020), Motivation and social cognitive theory, Contemp. Educ. Psychol., 60, 101832 DOI:10.1016/j.cedpsych.2019.101832.
88. Sen S., (2016), The effect of different metacognitive skill levels on preservice chemistry teachers' motivation, Kıbrıslı Eğitim Bilimleri Dergisi, 11(3), 126–143. https://www.ceeol.com/search/article-detail?id=965317.
89. Seyhan H. G., (2016), Effects of self-regulatory strategy on prospective science teachers' chemistry self-efficacy according to class level and gender, J. Educ. Train. Stud., 4(3), 97–117 DOI:10.11114/jets.v4i3.1213.
90. Shofiah V., Taruna R., Asra Y. K., Rajab K. and Sa’ari C. Z., (2023), Academic self-efficacy as a mediator on the relationship between academic motivation and academic achievement of college students during the online learning period, Int. J. Islamic Educ. Psychol., 4(1), 154–168 DOI:10.18196/ijiep.v4i1.18247.
91. Subramaniam V., Karpudewan M. and Roth W. M., (2023), Unveiling the teachers’ perceived self-efficacy to practice Integrated STrEaM Teaching, Asia-Pacific Educ. Res., 32, 327–337 DOI:10.1007/s40299-022-00655-4.
92. Sunny C. E., Taasoobshirazi G., Clark L. and Marchand G., (2017), Stereotype threat and gender differences in chemistry, Instr. Sci., 45, 157–175 DOI:10.1007/s11251-016-9395-8.
93. Syskowski S. and Kunina-Habenicht O., (2023), Motivational orientations in teaching-learning laboratories in chemistry, Lehr-Lern-Labore Digitalisierung, p. 329 DOI:10.1007/978-3-658-40109-2_30.
94. Taasoobshirazi G., Bailey M. and Farley J., (2015), Physics metacognition inventory part II: confirmatory factor analysis and rasch analysis, Int. J. Sci. Educ., 37(17), 2769–2786 DOI:10.1080/09500693.2015.1104425.
95. Taasoobshirazi G. and Farley J., (2013), Construct validation of the physics metacognition inventory, Int. J. Sci. Educ., 35(3), 447–459 DOI:10.1080/09500693.2012.750433.
96. Taghieh M. R., Tadayon Z. and Taghieh R., (2019), Comparison of cognitive and metacognitive strategies in the academic achievement of urban and rural students of Eghlid, IJERI: Int. J. Educ. Res. Innov., (12), 282–293 DOI:10.46661/ijeri.3709.
97. Urdan T. and Kaplan A., (2020), The origins, evolution, and future directions of achievement goal theory, Contemp. Educ. Psychol., 61, 101862 DOI:10.1016/j.cedpsych.2020.101862.
98. Urhahne D. and Wijnia L., (2023), Theories of motivation in education: An integrative framework, Educ. Psychol. Rev., 35(2), 45 DOI:10.1007/s10648-023-09767-9.
99. Uzuntiryaki E. and Çapa Aydın Y., (2009), Development and validation of chemistry self-efficacy scale for college students, Res. Sci. Educ., 39(4), 539–551 DOI:10.1007/s11165-008-9093-x.
100. Vallerand R. J., Pelletier L. G., Blais M. R., Briere N. M., Senecal C. and Vallieres E. F., (1992), The Academic Motivation Scale: a measure of intrinsic, extrinsic, and amotivation in education, Educ. Psychol. Meas., 52(4), 1003–1017 DOI:10.1177/0013164492052004025.
101. Wahyudiati D., Sutrisno H. and Supiah I., (2019), Self-efficacy and attitudes toward chemistry of pre-service chemistry teachers: gender and grades level perspective, Int. J. Sci. Technol. Res., 8(9), 1041–1044. https://staffnew.uny.ac.id/upload/132011628/penelitian/C8-Self-efficacy-And-Attitudes-Toward.pdf.
102. Wang L. and Yu Z., (2023), Gender-moderated effects of academic self-concept on achievement, motivation, performance, and self-efficacy: a systematic review, Front. Psychol., 14, 1136141 DOI:10.3389/fpsyg.2023.1136141.
103. Whitcomb K. M., Kalender Z. Y., Nokes-Malach T. J., Schunn C. D. and Singh C., (2020), Comparison of self-efficacy and performance of engineering undergraduate women and men, Int. J. Eng. Educ., 36(6), 1996–2014. https://www.ijee.ie/1atestissues/Vol36-6/24_ijee4004.pdf.
104. Wong S. Y., Liang J. C. and Tsai C. C., (2021), Uncovering Malaysian secondary school students’ academic hardiness in science, conceptions of learning science, and science learning self-efficacy: a structural equation modelling analysis, Res. Sci. Educ., 51(Suppl 2), 537–564 DOI:10.1007/s11165-019-09908-7.
105. Zhang J. and Zhou Q., (2023), Chinese chemistry motivation questionnaire II: adaptation and validation of the science motivation questionnaire II in high school students, Chem. Educ. Res. Pract., 24(1), 369–383 10.1039/D2RP00243D.
106. Zhou N., Fischer C., Rodriguez F., Warschauer M. and King S., (2020), Exploring how enrolling in an online organic chemistry preparation course relates to students’ self-efficacy, J. Comput. Higher Educ., 32, 505–528 DOI:10.1007/s12528-019-09244-9.

---