Adaptation Mechanisms of Hong Kong Secondary School Students to the Sudden Shift to Online Learning During the COVID-19 Pandemic: Path Analysis of PISA 2022 Data

doi:10.21203/rs.3.rs-8527651/v1

Download PDF

Research Article

Adaptation Mechanisms of Hong Kong Secondary School Students to the Sudden Shift to Online Learning During the COVID-19 Pandemic: Path Analysis of PISA 2022 Data

https://doi.org/10.21203/rs.3.rs-8527651/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Based on the Programme for International Student Assessment (PISA), this research investigated the adaptation mechanisms of 2366 Hong Kong secondary school students in the sudden shift in learning mode during the COVID-19 pandemic. This study conducted path analysis based on structural equation modeling to explore how feedback via information and communication technology (ICT) affects the home learning experience through students’ self-directed learning and information literacy, ultimately leading to improved psychosomatic health. The findings indicated that feedback via ICT significantly enhances students' self-directed learning self-efficacy, information literacy, and home learning experience, further exerting positive effects on their well-being. Both self-directed learning self-efficacy and information literacy partially mediated the relationship between feedback and home learning experience. Additionally, the research highlighted the significant moderating roles of growth mindset and collaborative tendency within the related pathways. In summary, this study showed the crucial role of feedback, self-directed learning self-efficacy, and information literacy in the online learning mode. In future online or blended learning environments, educators should pay more attention to providing feedback to students' learning to promote their self-directed learning self-efficacy and information literacy. This will help them maintain learning efficiency and well-being in the absence of traditional learning environment supervision.

Educational Psychology

Psychology

Adolescent

Self-directed Learning Self-efficacy

Information Literacy

Feedback

Well-being

Online Learning

A sudden outbreak of COVID-19 disrupted the pace of teaching and learning in education systems around the world since 2020. According to the World Health Organization (2020), COVID-19 was declared as a Public Health Emergency of International Concern (PHEIC) under the International Health Regulations (IHR). Meanwhile, Hong Kong announced the complete cessation of face-to-face classes soon after the announcement of the first highly suspected case of COVID-19 on 22 January 2020 (Caixin Global, 2020), and put forward the option of “Suspending Classes Without Suspending Learning” (Education Bureau, HKSAR, 2020), which means switching from face-to-face classes to e-learning or combining face-to-face classes with online schooling (blended) until the end of 2022 when it was again announced that face-to-face classes would be resumed gradually on a full-time basis from February 2023 onwards (Government Information Services Department, HKSAR, 2022). In the years between 2020 and early 2023, secondary school students in Hong Kong experienced a prolonged period of fully online or blended learning modes.

The Hong Kong Education Bureau (EDB) has made a lot of efforts and work to help students adapt to the online mode of learning, such as providing equal e-learning opportunities for students with less well-off background through the Quality Education Fund (QEF) (Education Bureau, HKSAR, 2021), including Internet devices such as computers and routers. This sudden change in learning mode, although technically possible, may not have prepared many students for the change. The shift in mode of teaching and the invisible stress caused by the pandemic may have had a significant impact on students' learning styles and mental health. Many studies reported an increase in adolescents’ anxiety symptoms during the pandemic, and many of the symptoms were caused by the lockdown in the pandemic (Magson et al., 2021; Yue et al., 2022; Francisco et al., 2020). Some studies reported the significance of learning motivation (Bryce & Fraser, 2022; Hadiapurwa et al., 2021) and digital skills (Kumpikaitė-Valinienė et al., 2021; Limniou et al., 2021) during the pandemic.

In the online mode of learning, information and communication technology (ICT) is both a cornerstone of teaching and learning, as well as a key pathway to ensure that students receive feedback and support from their teachers and classmates. In ICT-based online learning model, the mechanisms of how feedback and support from teachers and classmates through ICT can positively affect students' psychosomatic symptoms by contributing to the home learning experience through students' personal attributes and information literacy need to be systematically investigated. So far, there is no study that has systematically explored the relationship between teachers’ feedback and students’ self-efficacy and information literacy, and how these factors affect students’ learning experience during online learning. Previous studies using the Programme for International Student Assessment (PISA) dataset and modelling methods have only explored the relationship between some of these factors, such as the effects of teacher feedback on students' resilience (Zhan et al., 2024), students’ self-efficacy (Chang & Bangsri, 2020), and students’ well-being (Yang & Zheng, 2024). This study integrates these factors and aims to contribute to current literature by exploring the underlying mechanisms of these factors and offering a more comprehensive understanding of the dynamics involved in online learning environments.

This study aims to systematically understand the adaptation mechanisms of Hong Kong secondary school students and the challenges they faced during the sudden shift in learning mode during the COVID-19 pandemic. Starting from the feedback via ICT during online learning, we explore how such input affects the home learning experience through students’ self-directed learning and information literacy, ultimately leading to improved psychosomatic health. The model could be found in Fig. 1.

2.1. Feedback from Teachers and Classmates through ICT is a Social Resource

Maintaining positive interactions with teachers and classmates is a key factor in ensuring students’ home learning experience. According to the Social Support Theory, the relationships and social networks that individuals have are important social resources in the face of difficulties and stress (Leahy-Warren, 2014). In the traditional teaching mode, students were able to receive feedback and support from their teachers and classmates during their face-to-face interactions with them. These interactions become important social resources when they face learning difficulties and stress. For example, when facing learning challenges, teachers' learning guidance and encouragement can not only help students academically and enhance the overall experience of learning, but also emotionally alleviate uneasiness and reduce both mental and physical symptoms. In contrast, in the online learning mode, due to the reduced opportunities for face-to-face communication and interaction between students and teachers, the social resources available to students will be reduced accordingly, which will make it more difficult for students to face learning problems and pressures, leading to a decline in students' motivation and subjective initiative to learn, negatively affecting the learning experience at home, and may even be reacted to the physiological level of the students, resulting in headaches and other symptoms.

The online mode of learning has been integrated into students' learning routines during the pandemic, but this also means that many interactions with fellow teachers are missing compared to the traditional face-to-face teaching format before the pandemic (Alam et al., 2022; Adnan & Anwar, 2020). The use of ICT to interact with teachers and classmates (access to social resources) compensates the negative impact of the lack of face-to-face interaction to a certain extent and is an important part of students’ home learning experience. Therefore, in the online learning mode, feedback provided by teachers and classmates through ICT could be identified as an important social resource for students in facing challenges in learning, helping them to alleviate negative feelings and solve learning problems. According to Rasi and Vuojärvi (2017), timely, detailed, specific and positive feedback can empower students through emotions in the learning process. A similar argument is that feedback from teachers can give a feeling of motivation and care to online learners (Cheng, 2017). And from a performance perspective, feedback has also been confirmed to be able to have a facilitating, guiding and monitoring effect. For example, Truskowski and VanderMolen (2017) agree that feedback can be used to improve performance during independent practice, while Nadolski and Hummel (2017) see feedback as a tool that allows learners to monitor their progress. Therefore, we suggest that feedback not only could provide emotional support to students but also help them to achieve higher performance. The following hypothesis is proposed:

H1: In the online learning mode, feedback from teachers and classmates through ICT contribute to students’ positive feelings of learning at home.

H2: In the online learning mode, students’ positive home-learning feelings are inversely related to their psychosomatic symptoms.

2.2. Students’ Self-Efficacy in Self-Directed Learning

The online learning mode also challenged students' independence in their learning. In the traditional learning mode, regulated classroom discipline and structured course formats provide a stable learning environment for students. These tangible and intangible supervision allow students to maintain a stable pace in their learning. Even if they fall behind, they can still have direct supervision from the teacher and mutual help from classmates in the classroom to improve their learning progress. In the online learning mode, these tangible and intangible supervision are greatly limited by ICT, and students need to rely more on their self-efficacy of self-directed learning in order to maintain a certain pace of learning (Chang & Siew, 2024). Studies showed that individuals’ self-efficacy has a direct impact on their motivation and initiative (Schunk & DiBenedetto, 2021), especially when facing challenges in learning. Besides, according to the Self-Directed Learning Theory (Garrison, 1997), when individuals are faced with a lack of supervision and guidance, their own self-directed learning capabilities, which include the ability to independently set learning goals, set learning strategies, and evaluate the results of their learning, will play an important role. Some studies have pointed out that when students have low self-directed learning abilities, they may develop avoidance and boredom behaviors (Toraman et al., 2021; Clausen, 2021), which may negatively affect the home learning experience and may eventually lead to psychosomatic symptoms.

Effective learning requires learners to take the initiative in setting learning goals, choosing learning strategies, self-monitoring the learning process, and evaluating the final results (Garrison, 1997). In the traditional classroom, the teacher can be directly involved in most of the steps in the process, such as helping students to set appropriate learning goals and guiding them to choose appropriate learning methods, so as to cultivate students' self-directed learning ability. In the online learning mode, feedback can play a similar role. As Chen et al. (2018) argue, simple, accurate, and complete feedback can help learners identify and correct errors, motivate them to acquire knowledge, and increase their confidence and motivation in learning. At the same time, in this process of feedback helping learners to identify and correct errors, their input in learning can be felt by students to have a real effect on learning (Filius et al., 2018), which in turn motivates them to invest more time and effort (Truskowski & VanderMolen, 2017). By increasing motivation and confidence in learning, feedback fosters and promotes individual self-directedness in learning (Cheng, 2017), thus self-efficacy in self-directed learning improves. Based on these findings from the literature, we propose the following hypothesis:

H3: In the online learning mode, feedback from teachers and classmates through ICT can promote students' self-efficacy in self-directed learning.

2.3. Information Literacy of Students

The online learning mode also placed higher demands on students' information literacy. Information literacy is a set of abilities requiring individuals to recognize when information is needed and to have the ability to locate, evaluate, and use the needed information effectively (Association of College & Research Libraries, 2000). When it is impossible to interact with teachers and classmates face-to-face to get timely guidance and feedback, searching for answers on the Internet may become the choice of most students. The highly fragmented and potentially misleading information on the Internet requires students' information literacy skills to filter, integrate and understand. Students with higher information literacy skills will be able to get ahead of other students in the process, access more learning resources more efficiently, achieve better academic results (Yang & Kim, 2014), and in turn develop a more positive home learning experience, which ultimately promotes mental and physical health.

Learning is a process that requires individuals to select, process and integrate new information based on their existing experience and knowledge, and finally construct a new knowledge system. When students are limited by the online learning mode and need to search for learning resources online, their information literacy plays an essential role in this learning process. Feedback may contribute to or improve individual information literacy, and theoretically, teachers' feedback can guide students to improve their resource searching methods and help them to identify reliable sources. As in the study of Peter et al. (2015), they examined whether students' competence in information literacy improved significantly by providing them with individual feedback, and the results showed that students who received and used individual feedback had a significant improvement in their information literacy competence. However, in another study (Avcı & Ergün, 2019), the researchers compared the information literacy of two student groups that mainly differed in the level of activity in communicating feedback, and the results did not find a significant difference. In other words, it seems that whether or not students interact more with others (including teachers and classmates) does not necessarily promote information literacy skills. However, considering that the differences between these two groups of students in this study were not limited to the activeness of communicating feedback, and that there may be other remaining factors interfering with the results, we propose the following hypothesis:

H4: In the online learning mode, the feedback and support from teachers and classmates through ICT can promote students' information literacy.

2.4. The Role of Self-Directed Learning Self-Efficacy and Information Literacy

Students' self-efficacy in self-directed learning ability could positively influence the feelings of learning. According to the research of Son, Kim, and Cho (2018), there is a significant correlation between students' satisfaction with their subjects and their self-directed learning ability. This is further proved in Chang and Siew's (2024) study where students' self-directed learning competencies had a significant positive impact on their e-learning course assessment. What’s more, some research indicates that self-efficacy has a significantly negative relationship with academic burnout (Brouwers & Tomic, 2000; Charkhabi et al., 2013), meaning that individuals with higher self-efficacy are less likely to experience emotional exhaustion during learning and have a more positive learning experience. Some researchers also suggest that students with high academic self-efficacy have higher intrinsic rewards and concentration during learning, enabling them to have more positive learning experiences (Bassi et al., 2007). Furthermore, academic self-efficacy has been confirmed to significantly predict students' satisfaction with online learning (Jan, 2015). Also, the self-directed learning theory states that individuals with higher self-directed learning ability can maintain higher learning efficiency under the lack of external supervision. Therefore, combining this with the above findings, higher self-directed learning self-efficacy may contribute to a more positive learning experience.

Higher information literacy may also have a positive effect on their learning in the online learning mode. Information literacy is also known by other names such as digital literacy, ICT literacy, and e-literacy (Martínez-Bravo et al., 2020). Studies on the relationship between information literacy and student achievement have shown different or even opposite results. For example, Yang and Kim's (2014) study supports that information literacy positively affects learning outcomes in online learning. Akhyar et al. (2021) hold a similar perspective, suggesting that information literacy in online learning has a significant positive impact on students' science learning outcomes. However, there are some studies that hold the opposite view, suggesting that information literacy does not have a significant impact on students' cumulative GPA (Abbas et al., 2019; Darlis & Sari, 2021). A meta-analysis addressed the causal relationship between information literacy and academic achievement, based on analyzing multiple litterateurs, and suggested that information literacy promotes academic performance (Li et al., 2025). Better academic performance, in turn, may promote students’ better home-learning feelings by increasing their confidence and satisfaction with the learning process. Therefore, we propose the following hypotheses:

H5: Self-directed learning self-efficacy and information literacy can positively influence students' home-learning feelings.

Combining this with previous literature about the relationship of feedback with self-directed learning self-efficacy and with information literacy, we propose the following hypotheses:

H6: Self-directed learning self-efficacy and information literacy both play a mediating role in the pathways of feedback from teachers and classmates through ICT to promote students' positive home-learning feelings.

2.5. Growth Mindset and Cooperative Tendencies

Students' own growth mindset and propensity for co-operation may affect the pace of learning and the actual home learning experience. From the perspective of learning itself, Social Constructivist holds that knowledge is co-created by a group, and that learning is not a passive acceptance of knowledge, but a process that requires individuals to actively select, process, and integrate new information based on their own existing experiences and foundations in order to construct a new knowledge system (Vygotsky & Cole, 2018). During this growth process, the individual's growth mindset or fixed mindset bias may have a significant impact. When challenges are encountered during this process, individuals with a growth mindset may take more proactive actions to overcome them, whereas individuals with a fixed mindset are more likely to be less proactive and take passive-avoidant actions (Blackwell et al., 2007). Also, Social Constructivist emphasizes the fact that the process of selecting, processing and integrating new information needs to take place in a certain social context in interaction with others, and therefore students' perceptions of cooperation may also influence this process (Vygotsky & Cole, 2018). For example, when students with a high propensity for cooperation encounter problems in finding learning resources online, they are more likely to discuss them with their teachers or classmates in order to obtain their support in terms of online information.

Studies supported the moderating role of growth mindset in the process of self-directed learning influencing home learning feelings. According to Dweck (2006), growth mindset has a strong correlation with students' learning behaviors and suggests that learners with a growth mindset show more curiosity, openness and persistence in their learning process (Duckworth, 2016). Consequently, growth mindset is recognized as an approach that can be effective in promoting self-directed learning self-efficacy and deep learning, especially when students are challenged in their learning (Brandt, 2020). In a study of how growth mindset affects academic achievement in high school students across different socioeconomic contexts, it was found that after controlling for all common variables that could predict academics, students' growth mindset still explained 11.8% of the variance in academic achievement and showed the positive relationship between growth mindset and academic achievement to a certain extent (r = 0.343; Claro et al., 2016). And their results also found that students with a growth mindset were able to show higher adaptability in facing uncertainty and academic stress, reinforcing the learning experience. These results showed that growth mindset has a positive relationship with home learning feelings (Brandt, 2020), and individuals can promote self-directed learning self-efficacy through growth mindset. It can be hypothesized that growth mindset may moderate the effect of self-directed learning self-efficacy on home learning feelings, whereby the higher an individual's growth mindset, the more significant the effect of self-directed learning self-efficacy on home learning feelings. Growth mindset may moderate the effect of self-directed learning self-efficacy on home learning feelings, which means that the higher the individual's growth mindset, the stronger the effect of self-directed learning self-efficacy on home learning feelings. Based on the above arguments, we propose the following hypothesis:

H7: Students' growth mindset can moderate the effect of self-efficacy in self-directed learning on the positive feelings of learning at home.

Furthermore, studies supported the moderating role of individuals' attitudes towards cooperative behaviors in the influence of information literacy on feelings about learning at home. Students with a higher tendency for cooperative behaviors tend to interact more with their teachers and classmates to obtain guidance and encouragement (feedback) when they encounter problems in learning. Some studies have confirmed that when classmates work together to overcome learning problems, they form a team learning environment which is constructive and supportive and can improve learning performance and intrinsic motivation (Nielsen et al., 2018), allowing individuals to be more satisfied and motivated in the learning processes. At the same time, this teamwork improves students' information literacy skills, with students doubling the likelihood of finding the correct item knowledge question after working together (Schroeter et al., 2020). Schroeter et al.’s (2020) study also suggests that teamwork may enhance individuals’ information literacy skills by improving their critical thinking. These results show that team learning environment created by the positive interaction and feedback between students and their teachers and classmates can improve academic performance, and this teamwork can promote information literacy skills by improving students’ critical thinking skills. It can be hypothesized that individual cooperation may moderate the effect of information literacy on the feeling of learning at home, which means the more cooperative an individual is, the more significant the effect of information literacy on the feelings of learning at home is. Based on the above arguments, we propose the following hypothesis:

H8: Students' cooperation can moderate the role of information literacy on positive home-learning feelings.

2.6. Impact of Gender, Parental Involvement and Home Resources

Studies indicated that females are more affected by the COVID-19 pandemic in terms of homework, social isolation and mental health compared to males (Prowse et al., 2021). At the same time, the level of parental involvement in their children's academic matters, such as the frequency of communication with teachers, may be related to students' negative psychological conditions. For example, in the study by Liu et al. (2021), students whose parents maintained frequent communication with teachers reported higher levels of depressive symptoms compared to students whose parents did not maintain frequent communication with teachers. In addition, the economic situation represented behind home resources may have a potential impact on students' learning. For students with lower socioeconomic status, it may be difficult to complete homework and online learning in the online learning mode (Van Lancker & Parolin, 2020), even if one could consider applying for education funding. Considering these variables may have a potential impact on students' psychological status and learning, they were included in the analyses as control variables.

3.1. Data Source and Sample Characteristics

This study primarily used the data of the PISA 2022 in Hong Kong, which was collected by Organisation for Economic Co-operation and Development (OECD), and it focuses on the learning, health, and family background information of adolescents born in 2006 or 2007 (Organisation for Economic Co-operation and Development [OECD], 2023). PISA used a two-stage international standard stratified sampling design. The first stage of sampling was conducted at the school level, with the selection probability referencing the percentage of students born in 2006 and 2007 in the school. A total of 163 schools participated in this project. The second stage of sampling was taken at the student level, students born in 2006 and 2007 were randomly selected within the same school. And each school generally selected 42 students (no more than 50 students). So, 5907 students were initially selected. Furthermore, PISA used two versions of questionnaire in Hong Kong based on two languages: English and Chinese. To ensure the accuracy of the results, we chose the data collected through the Chinese version. After removing the data with missing values or collected through the English version questionnaire, a total of 2366 samples were selected in this study.

Among these 2366 students, there were 1167 females (49.3%) and 1199 males (50.7%), 2023 students were born in 2006 and 343 in 2007. In terms of grade, the majority of the sample was in grades 10 (1625, 68.7%) and 9 (648, 31%). Most samples were either native student (1337, 60.4%), defined as those with at least one parent born in the country/economy of assessment, or Second-Generation student (732, 33.1%), defined as those born in the country but whose parent(s) were born abroad. And First-Generation student refers to students who were born outside the country/economy of assessment, with both parents also born in another country/economy. As for parental education, most of the mothers in the sample were identified as International Standard Classification of Education (ISCED) level 3.4 (57.7%) or level 2 (21.3%), and most of the fathers were also identified as level 3.4 (60.7%) or level 2 (20.2%). Please see Table 1 for the sample characteristics.

Table 1

*Sample Characteristics*
Variable	Category	Number (n)	Percentage (%)
Grade	Grade 7	7	0.3
	Grade 8	79	3.3
	Grade 9	648	27.4
	Grade 10	1625	68.7
	Grade 11	7	0.3
Gender	Female	1167	49.3
Gender	Male	1199	50.7
Mother’s Education Level	ISCED Level 3.4	1365	57.7
	ISCED Level 3.3	162	6.8
	ISCED Level 2	504	21.3
	ISCED Level 1	211	8.9
	Did not complete ISCED Level 1	72	3.0
	Missing	52	2.2
Father’s Education Level	ISCED Level 3.4	1437	60.7
	ISCED Level 3.3	162	6.8
	ISCED Level 2	477	20.2
	ISCED Level 1	159	6.7
	Did not complete ISCED Level 1	44	1.9
	Missing	87	3.7
Immigration Background	Native student	1337	56.5
	Second-generation student	732	30.9
	First-generation student	143	6.0
	Missing	154	6.5
Birth - Year	2006	2023	85.5
Birth - Year	2007	343	14.5

3.2. Measurement of Variables

We selected the indicator variables which have been officially calculated and issued by PISA for research, and they are: ICT (ICTFEED), self-directed learning self-efficacy (SDLEFF), students’ practices regarding online information (ICTINFO), feelings about learning at home (FEELLAH), psychosomatic symptoms (PSYCHSYM), cooperation (COOPAGR), growth mindset (GROSAGR), parental involvement (PARINVOL), home possessions (HOMEPOS), and gender (ST004D01T). Missing values were removed based on listwise deletion to maintain a complete sample for analysis according to these ten indicator variables.

These nine variables were all calculated by PISA after a series of complex processing methods with well-accepted confidence. In the PISA, the scale scoring system for students was built using an Item Response Theory (IRT) model that considers the difficulty and discrimination of each item. After the IRT model had been determined, scale scores were computed in the form of weighted likelihood estimates (WLE), which provided a stable and unbiased estimate of a student's underlying traits. To ensure reliability, WLE scores were calculated only when students responded to at least three valid items on the scale. Missing or invalid responses were coded as ‘99’ in the SPSS data file. And the tables about these indicator variables’ internal consistency could be found in the supplementary Excel file provided via StatLink on page 441 441 of the PISA 2022 Technical Report (OECD, 2024).

For the frequency of feedback and support provided by teachers and classmates through ICT, we used ICTFEED index in PISA, which measures the frequency of using digital resources in various activities related to support or feedback from others, including teachers and classmates. This indicator variable contains four questions, such as “read or listen to feedback sent by my teachers regarding my work and academic results”, and there are five answer choices for these questions (“never or almost never”, “about once or twice a year”, “about once or twice a month”, “about once or twice a week”, “every day or almost every day”). Higher scores indicate higher frequency. The internal consistency of this indicator variable was 0.87, which could be found in Table 19.A.82 (titled “Cronbach’s alpha for the IRT scales in the ICT Familiarity Questionnaire”) in the supplementary Excel file provided via StatLink (p. 441) of the PISA 2022 Technical Report (OECD, 2024).

For the self-directed learning self-efficacy, we used SDLEFF index in PISA, which refer to the ratings of how confident student felt about having to do various self-directed learning tasks when their school close again in the future. In other words, SDLEEF indicates the index of self-directed learning self-efficacy. This indicator variable has eight items, such as “finding learning resources online on my own”, and there are four answer choices for these questions (“not at all confident”, “not very confident”, “confident”, “very confident”). Higher scores indicate higher self-efficacy. The internal consistency of this indicator variable was 0.89, which could be found in Table 19.A.11 (titled “Cronbach's alpha for the IRT scales in the Student Questionnaire”) in the supplementary Excel file provided via StatLink (p. 441) of the PISA 2022 Technical Report (OECD, 2024).

For the information literacy, we used ICTINFO index in PISA, which refers to the index of “students’ practices regarding online information”, it contains six items, such as “when searching for information online I compare different sources”, and there are four answer choices for these questions (“strongly disagree”, “disagree”, “agree”, “strongly agree”). Higher scores indicate higher frequency. Higher scores indicate better information literacy. The internal consistency of this indicator variable was 0.84, which could be found in Table 19.A.82 (titled “Cronbach's alpha for the IRT scales in the ICT Familiarity Questionnaire”) in the supplementary Excel file provided via StatLink (p. 441) of the PISA 2022 Technical Report (OECD, 2024).

For the feelings of home-based learning, we used FEELLAH index in PISA, which refers to students’ feelings about learning at home when school buildings were closed during COVID-19. This indicator has six items, such as “I enjoyed learning by myself”, and there are four answer choices for these questions (“strongly disagree”, “disagree”, “agree”, “strongly agree”). Higher scores indicate better feelings. The internal consistency of this indicator variable was 0.71, which could be found in Table 19.A.11 (titled “Cronbach's alpha for the IRT scales in the Student Questionnaire”) in the supplementary Excel file provided via StatLink (p. 441) of the PISA 2022 Technical Report (OECD, 2024).

For the psychosomatic symptoms, we used PSTCHSYM index in PISA, which refers to the frequency of psychosomatic symptoms experienced by students, like headache and stomach pain and so on. This indicator variable has nine items, and each item has five answer choices (“rarely or never”, “about every month”, “about every week”, “more than once a week”, “about every day”). Higher scores indicate higher risk of psychosomatic symptoms. The internal consistency of this indicator variable was 0.91, which could be found in Table 19.A.99 (titled “Cronbach's alpha for the IRT scales in the Well-Being Questionnaire”) in the supplementary Excel file provided via StatLink (p. 441) of the PISA 2022 Technical Report (OECD, 2024).

For the cooperation tendency, we used COOPAGR index in PISA, which refers to the ratings of students’ agreement with statements about a series of behaviors indicative of cooperation, and represents the tendencies in behavior and attitudes. This indicator variable has ten items, such as “I work well with other people”. Before scaling, some items were reverse-coded to maintain consistent direction. Each of the ten items have five answer choices (“strongly disagree”, “disagree”, “neither agree nor disagree”, “agree”, “strongly agree”). Higher scores indicate more positive behaviors and attitudes about cooperation. The internal consistency of this indicator variable was 0.63, which could be found in Table 19.A.11 (titled “Cronbach's alpha for the IRT scales in the Student Questionnaire”) in the supplementary Excel file provided via StatLink (p. 441) of the PISA 2022 Technical Report (OECD, 2024). The internal consistency of this indicator variable is barely acceptable, but it is worth noting that the internal consistency of this indicator in other countries or regions is generally low. Most of them fall on the range of 0.5 to 0.7, and even less than 0.5 in some regions. This may be related to the different interpretations of cooperation in different cultural background, especially in a globalized environment like Hong Kong, where differences in cultural background are highly common.

For growth mindset, we used GROSAGR index in PISA, which measures the growth mindset. It has four items, such as “your intelligence is something about you that you cannot change very much”. Some items have been reverse-coded before calculating. Each question has four answer choices (“strongly disagree”, “disagree”, “agree”, “strongly agree”). Higher scores indicate a stronger tendency to growth mindset, which means students are more convinced that effort can change ability. The internal consistency of this indicator variable was 0.78, which could be found in Table 19.A.11 (titled “Cronbach's alpha for the IRT scales in the Student Questionnaire”) in the supplementary Excel file provided via StatLink (p. 441) of the PISA 2022 Technical Report (OECD, 2024).

For the degree of parental participant, we used PARINVOL index in PISA, which refers to the level of parents’ involvement in their child’s education in the past years. This indicator variable has ten items, such as “discussed my child’s progress with a teacher on my own initiative”, and each item has three answer choices (“yes”, “no”, “not supported by school”). Some items have been reverse-coded before calculating the scores. High scores indicate that parents participate more in their child’s education. The internal consistency of this indicator variable was 0.79, which could be found in Table 19.A.111 (titled “Cronbach's alpha for the IRT scales in the Parent Questionnaire”) in the supplementary Excel file provided via StatLink (p. 441) of the PISA 2022 Technical Report (OECD, 2024).

For the home resources, we used HOMEPOS index in PISA, which refers to the index of family’s economic conditions, higher scores indicate richer household material resources, including education-related items. This indicator variable has 31 items, such as “a room of your own”, “educational software or apps” and “how many books are there in your home?”. The internal consistency of this indicator variable was 0.81, which could be found in Table 19.A.11 (titled “Cronbach's alpha for the IRT scales in the Student Questionnaire”) in the supplementary Excel file provided via StatLink (p. 441) of the PISA 2022 Technical Report (OECD, 2024).

3.3. Data Analysis

In addition to descriptive statistics, correlation and other analytical methods conducted with SPSS 29.0, this study primarily used R Studio with the R version 4.5.0 for path analysis based on structural equation modeling to explore the mechanism of ICTFEED’s influence on FEELLAH. At the same time, regarding the complex sampling design of PISA data, the weighing variable (W_FSTUWT) and the replication weighting variable were used in the analysis to guarantee the precision and representativeness of the data results. Before testing the model, ICTFEED, SDLEFF, ICTINFO, FEELLAH, PSYCHSYM, COOPAGR, GROSAGR, HOMEPOS, and PARINVOL had been mean-centered, and the gender variable (ST004D01T) was coded as a dummy variable (Female = 0, Male = 1).

In terms of modeling, this study used structural equation modeling for path analysis based on these variables. This model set ICTINFO as the independent variable, SDLEFF (M1) and ICTINFO (M2) as the parallel mediator variables, FEELLAH as the first-level dependent variable, and PSYCHSYM as the second-level dependent variable, while GROSAGR and COOPAGR were imported as moderating variables to regulate the paths of M1 to FEELLAH and M2 to FEELLAH. The model could be found in Fig. 1. We constructed interaction items to explore the moderating effects, and calculated the effects of mediating, and moderated mediating, and so on. When we were analyzing these effects, the Balanced Repeated Replication weights were referenced to provide more accurate standard errors and confidence intervals.

4.1. Descriptive Statistics

The descriptive statistics results showed that the continuous variables all have their means in the range of -0.25 to 0.25, the standard deviation ranged from 0.82 to 1.05. The average level of the global sample on this variable was noted as 0, with a mean lower than 0 indicating a level below the global average. For most of the indicator variables, the absolute values of skewness were less than 1 and the absolute values of kurtosis were less than 2. The absolute value of COOPAGR was significantly higher than 7 in kurtosis, indicating the students’ scores on this variable were more concentrated. However, since this variable is a standardized indicator calculated by PISA through the WLE, this variable was used directly in this study without transformation or adjustment. The descriptive statistics results indicated the data in this study followed an approximately normal distribution. Details could be found in Table 2.

Table 2

*Variables Characteristics*
Variables	Mean	SD	Skew	Kurt
ICTFEED (WLE)	0.172	0.975	0.469	0.836
SDLEFF (WLE)	-0.155	0.911	0.189	0.727
ICTINFO (WLE)	0.251	0.859	0.183	3.454
GROSAGR (WLE)	-0.102	1.048	-1.371	3.640
COOPAGR (WLE)	-0.197	0.837	1.659	8.670
FEELLAH (WLE)	0.199	0.902	0.346	1.828
PSYCHSYM (WLE)	-0.188	1.005	-0.466	0.624
HOMEPOS (WLE)	-0.200	0.821	0.265	0.234
PARINVOL (WLE)	0.233	0.845	0.644	1.231

4.2. Correlation Between Variables

A Pearson correlation analysis was conducted. In the analysis, to make the results more representative, the weighing variable (W_FSTUWT) was used to exclude the negative bias from complex sampling. The results showed that ICTFEED had a significantly and positively correlations with SDLEFF (r = 0.256, p < 0.001) and ICTINFO (r = 0.313, p < 0.001); FEELLAH was significantly associated with SDLEFF (r = 0.493, p < 0.001), ICTINFO (r = 0.274 p < 0.001), and PSYCHSYM (r=-0.049, p < 0.001). Furthermore, GROSAGR (moderator variable) was significantly and positively associated with SDLEFF (r = 0.144, p < 0.001), FEELLAH (r = 0.107, p < 0.001), and PSYCHSYM (r=-0.158, p < 0.001); COOPAGR (moderator variable) was also significantly and positively related to ICTINFO (r = 0.208, p < 0.001), FEELAH (r = 0.186, p < 0.001), and PSYCHSYM (r=-0.147, p < 0.001). Both control variables HOMEPOS and PARINVOL were significantly associated with all other variables except GROSAGR. These correlation results provided primary support for the pathways in the model and confirm the inclusion of these variables in the following analyses. Details could be found in Table 3.

Table 3

*Correlation between Variables*
Variables	COOPAGR	GROSAGR	FEELLAH	SDLEFF	HOMEPOS	ICTFEED	ICTINFO	PSYCHSYM	PARINVOL
COOPAGR	1	0.101*	0.186*	0.210*	0.089*	0.162*	0.208*	-0.147*	0.049*
GROSAGR		1	0.107*	0.144*	-0.011	0.115*	0.063*	-0.158*	-0.014
FEELLAH			1	0.493*	0.097*	0.225*	0.274*	-0.049*	0.090*
SDLEFF				1	0.126*	0.256*	0.247*	-0.131*	0.090*
HOMEPOS					1	0.120*	0.121*	0.087*	0.063*
ICTFEED						1	0.313*	-0.056*	0.077*
ICTINFO							1	-0.038*	0.069*
PSYCHSYM								1	0.037*
PARINVOL									1
Note. “ * ” indicates p < 0.001; all correlation coefficients are Pearson correlations.

4.3. Path Analysis

Before path analysis, we did collinearity diagnostics without applying the main sampling weight (W_FSTUWT). The results showed that there was no severe multicollinearity between these variables. These results confirmed the statistical independence of the explanatory variables to a certain extent, supported the robustness of the following path analysis, and ensured that the path analysis’s estimated path coefficients were not biased by multicollinearity. Details could be found in Table 4.

Table 4

*Collinearity Diagnostics*
Variables	Tolerance	VIF
COOPAGR (WLE)	0.906 > 0.1	1.104 < 5
GROSAGR (WLE)	0.955 > 0.1	1.047 < 5
SDLEFF (WLE)	0.858 > 0.1	1.165 < 5
HOMEPOS (WLE)	0.948 > 0.1	1.055 < 5
ICTFEED (WLE)	0.856 > 0.1	1.168 < 5
ICTINFO (WLE)	0.862 > 0.1	1.160 < 5
PSYCHSYM (WLE)	0.907 > 0.1	1.103 < 5
PARINVOL (WLE)	0.980 > 0.1	1.020 < 5
Gender	0.951 > 0.1	1.052 < 5
Note. Dependent variable: FEELLAH (WLE).

However, the analyses showed that a few variables did not satisfy the requirement of approximate normal distribution in the descriptive statistics results, hence, in path analysis, we used the Maximum Likelihood Mean-adjusted Estimator (MLM) to test the model fit to improve the robustness of parameter estimation. In addition, the weighing variable (W_FSTUWT) was used in testing model fit. The results of the model fit showed that χ²(12) = 69.504, p < 0.001, robust CFI = 0.921, robust TLI = 0.751, robust RMSEA = 0.070 (90% CI: [0.054, 0.086]), and SRMR = 0.032. Among these results, the significant result of chi-square test may be caused by the large sample size, and the low robust TLI may be caused by the absence of observed variables in the model. Thus, the model fit was acceptable, has good explanatory power, and was able to respond well to the structural relationship between these variables.

4.4. Individual Direct Path Estimates Between Variables

In the model, ICTFEED had a significant and positive effect on both SDLEFF (β = 0.226, p < 0.001, 95% CI: [0.185, 0.237]) and ICTINFO (β = 0.280, p < 0.001, 95% CI: [0.222, 0.274]), and a similar effect on FEELLAH, but with a smaller effect size (β = 0.068, p < 0.001, 95% CI: [0.051, 0.083]). In addition, SDLEFF and ICTINFO also had significant positive effects on FEELLAH, with path coefficients of β = 0.444 (p < 0.001, 95% CI: [0.420, 0.461]) and β = 0.121 (p < 0.001, 95% CI: [0.100, 0.157]) respectively, while FEELLAH had a significant negative effect on PSYCHSYM (β = -0.149, p < 0.001, 95% CI: [-0.215, -0.112]). Also, GROSAGR had a significant effect on both SDLEFF (β = 0.117, p < 0.001, 95% CI: [0.077, 0.124]) and PSYCHSYM (β = -0.134, p < 0.001, 95% CI: [0.147, 0.110]), and COOPAGR had a significant effect on both ICTINFO (β = 0.133, p < 0.001, 95% CI: [0.106, 0.164]) and PSYCHSYM (β = -0.127, p < 0.001, 95% CI: [-0.177, -0.125]). In terms of control variables, while gender had a significant effect on SDLEFF (β = 0.058, p < 0.001, 95% CI: [0.072, 0.138]), ICTINFO (β = 0.026, p < 0.01, 95% CI: [0.018, 0.071]), FEELLAH (β = -0.026, p < 0.01, 95% CI: [-0.083, -0.019]), and PSYCHSYM (β = -0.182, p < 0.001. 95% CI: [-0.404, -0.327]), PARINVOL had a significant effect on SDLEFF (β = 0.066, p < 0.001), ICTINFO (β = 0.035, p < 0.01), FEELLAH (β = 0.033, p < 0.001, 95% CI: [0.019, 0.055]) and PSYCHSYM (β = 0.055, p < 0.001, 95% CI: [0.038, 0.090]), HOMEPOS had a significant effect on SDLEFF (β = 0.100, p < 0.001, 95% CI: [0.087, 0.128]), ICTINFO (β = 0.075, p < 0.001, 95% CI: [0.056, 0.100]) and PSYCHSYM (β = 0.096, p < 0.001, 95% CI: [0.096, 0.140]), most of these effect sizes were small but all were significant, while there was no significant effect of HOMEPOS on FEELLAH (β = 0.012, p = 0.217, 95% CI: [-0.009, 0.028]).

Based on above results, we can conclude that Hypothesis 1, Hypothesis 2, Hypothesis 3, Hypothesis 4, and Hypotheses 5 were supported.

4.5. Mediating Pathways and Moderating Effects

We further tested the mediating effects in the model. In the ICTFEED to SDLEFF to FEELLAH pathway, the mediating effect of SDLEFF was significant (β = 0.102, p < 0.001, 95% CI: [0.080, 0.104]); in the SDLEFF to FEELLAH to PSYCHSYM pathway, the mediating effect of FEELLAH was significant (β = -0.066, p < 0.001, 95% CI: [-0.097, -0.049]); in the ICTFEED to ICTINFO to FEELLAH pathway, the mediating effect of ICTINFO was significant (β = 0.037, p < 0.001, 95% CI: [0.027, 0.040]); in the ICTINFO to FEELLAH to PSYCHSYM pathway, the mediating effect of FEELLAH was significant (β = -0.018, p < 0.001, 95% CI: [-0.031, -0.014]); in the pathway ICTFEED to SDLEFF to FEELLAH to PSYCHSYM, the overall mediating effect of SDLEFF and FEELLAH was significant, but the amount of effect was weak (β = -0.015, p < 0.001, 95% CI: [-0.021, -0.010]); in the pathway ICTFEED to ICTINFO to FEELLAH to PSYCHSYM, the overall mediating effects were also significant and the effect sizes were weak (β = -0.006, p < 0.001, 95% CI: [-0.008, -0.004]).

We also tested the moderating effects in the model. The moderating effect of GROSAGR was significant in the SDLEFF to FEELLAH pathway (β = 0.062, p < 0.001, 95% CI: [0.032, 0.065]), as well as the moderating effect of COOPAGR in the ICTINFO to FEELLAH pathway was significant (β = 0.082, p < 0.001, 95% CI: [0.044, 0.085]).

4.6. Moderated Mediation Effects

We also examined the moderated mediation effects. The mediating effects of SDLEFF in the ICTFEED to SDLEFF to FEELLAH pathway were β = 0.086 (SD = 0.006, 95% CI: [0.070, 0.094]), β = 0.100 (SD = 0.007, 95% CI: [0.079, 0.103 ]), β = 0.115 (SD = 0.008, 95% CI: [0.088, 0.116]), and the mediating effect was increased with GROSAGR. The mediating effects of FEELLAH in the SDLEFF to FEELLAH to PSYCHSYM pathway were β = -0.057 (SD = 0.012, 95% CI: [-0.085, -0.044]), β = -0.066 (SD = 0.013, 95% CI: [− 0.097, -0.049]), β = -0.076 (SD = 0.015, 95% CI: [-0.107, -0.054]), and the mediating effect was increased with progressively higher levels of GROSAGR (Fig. 2).

The mediating effects of ICTINFO in the ICTFEED to ICTINFO to FEELLAH pathway were β = 0.014 (SD = 0.004, 95% CI: [0.010, 0.027]), β = 0.034 (SD = 0.004, 95% CI: [0.024, 0.038]), β = 0.053 (SD = 0.004, 95% CI: [0.037, 0.052]), and the mediating effect was increased with COOPAGR. The mediating effects of FEELLAH in the ICTINFO to FEELLAH to PSYCHSYM pathway were β = -0.008 (SD = 0.004, 95% CI: [-0.020, -0.006]), β = -0.018 (SD = 0.004, 95% CI: [− 0.031, -0.014]), β = -0.028 (SD = 0.005, 95% CI: [-0.041, -0.020]), and the mediating effect was increased with the gradual increase of COOPAGR level (Fig. 3).

Based on these results, we can conclude that the mediation effect increases with the level of moderator variables, and Hypotheses 6, 7, and 8 were supported.

4.7. Total Effects

In the pathway where ICTFEED affected FEELLAH through the two parallel mediators SDLEFF and ICTINFO, the total effect was the sum of the mediating effect of SDLEFF and the mediating effect of ICTINFO and the direct effect of ICTFEED to FEELLAH, which was β = 0.208 (CI: [0.175, 0.210]), with the mediating effect accounting for 67.15% of the total.

In the pathway where ICTFEED affected FEELLAH through the two parallel mediators SDLEFF and ICTINFO, and then indirectly affected PSYCHSYM, the total effect was the sum of the overall mediated effect of SDLEFF and FEELLAH, and the overall mediated effect of ICTINFO and FEELLAH, and the direct effect of ICTFEED to PSYCHSYM, which was β = -0.031 (CI: [-0.041, -0.022]), with the mediating effect accounting for 68.75% of the total. Both ratio shares indicate that the major part of the total effect of both pathways was delivered by the indirect pathway.

5.1. Potential Effect of Cultural Differences

First, the secondary school students in Hong Kong were slightly weaker than the global average level (which is set at 0) in terms of self-directed learning self-efficacy, growth mindset, and cooperation, and were better than the global average in terms of feedback and support frequency obtained through ICT, information literacy, feelings about home learning, and parental involvement (as shown in Table 2). This difference might reflect the differences in cultural background.

Chinese traditional culture emphasizes more on collectivism, and it has a higher power distance. Both power distance and the tendencies of collectivism/individualism are closely related to self-efficacy. In an individualistic society, individuals are expected to be responsible for their own lives and studies. As a result, they may have more opportunities to cultivate a sense of self-efficacy (Tan et al., 2021). During the process of independent learning, they can experience the growth of their abilities better, further fostering a growth mindset. In a collectivist society, parents pay more attention to their children's academic performance, and thus the parental involvement is higher than the average level.

Furthermore, in a social context with a higher power distance, students may rely more on teachers' teachings and structured curriculum learning, lacking opportunities to express their viewpoints (Ding & Zhu, 2021). Under this context, teachers may provide more detailed and frequent guidance, thus students in Hong Kong can receive higher frequency in feedback. Based on these arguments, collectivist countries usually have lower self-efficacy than individualist countries, and countries with a higher power distance are also usually associated with lower self-efficacy (Jin et al., 2023). In a comparative study on self-efficacy and self-regulation strategies of English learners in Germany and China by Wang et al. (2013), the average self-efficacy of Chinese students (M = 4.64) was significantly lower than that of German students (M = 5.59). This can explain the difference between students in Hong Kong and the global average in these demographic variables.

5.2. The Influence of ICT Feedback on Students' Self-Directed Learning Self-Efficacy and Information Literacy

Our research results confirm the impact of the feedback and support that students obtain from teachers and classmates through ICT on their self-directed learning self-efficacy and information literacy. Specifically, the more frequently students use ICT to receive feedback and support from their teachers and classmates, the stronger the effect on improving their self-directed learning self-efficacy and information literacy.

The role of self-directed learning in self-efficacy can be explained through the Theory of Self-Efficacy. Teachers and students can provide feedback and support to students through ICT, such as homework guidance to help students gain a successful experience, and verbal encouragement to enhance students' confidence in their own abilities, and so on. These are consistent with the conclusions of the existing literature. In the online learning environment, feedback can improve students' self-efficacy in various forms (Wang & Wu, 2008), for example, to help students perceive the effect of their efforts on learning (Filius et al., 2018), and to enhance students' enthusiasm and confidence in learning (Chen et al., 2018).

The role of Information Literacy can be explained by the Seven Faces of Information Literacy proposed by Bruce (1997). From the perspective of subjective experience, Bruce summarized seven ways in which people practice information literacy, namely information technology, information sources, information process, information control, knowledge construction, wisdom, and personal empowerment. In the information process, Bruce emphasized that individuals could obtain opportunities to develop information literacy in the process of handling information, including collection, integration, expression and communication. Among them, expression and communication are inevitable processes in the ICT environment. The results of this study are consistent with the experimental results of Peter et al. (2015), which argued that by providing students with specific and detailed personal feedback, it could help students improve their information literacy.

5.3. The Mediating Role of Self-Directed Learning Self-Efficacy and Information Literacy

The mediating role of self-directed learning self-efficacy and information literacy in the path from the feedback and support students obtain from teachers and classmates through ICT to the experience of home learning has been confirmed in this study. According to the results of data analysis, both self-directed learning self-efficacy and information literacy played partial mediating roles in their path.

The mediating role of self-directed learning in self-efficacy can be explained by Social Cognitive Theory and Self-Directed Learning Theory. According to Social Cognitive Theory, an individual's perception of their ability and belief in completing tasks will directly affect their motivation and performance (Bandura, 1989). In the current situation, an individual's perception of their own abilities directly depends on their self-directed learning ability. According to the Self-Directed Learning Theory, this ability will play a key role in learning performance when an individual lacks external supervision. Combined with the results of this study and some existing literature, the improvement of self-directed learning ability promotes students' learning performance in electronic courses (Chang & Siew, 2024), and self-efficacy can promote individuals to obtain more positive emotions (Schweder et al., 2022; Hayat et al., 2020). Therefore, this study suggests that the mediating role of self-directed learning self-efficacy not only promotes positive learning experiences through beliefs, but also enhances learning performance by improving abilities, and thereby promoting the positive feelings of home learning.

Additionally, this study found that the mediating effect of self-directed learning self-efficacy (β = 0.102) was much higher than the direct effect of ICT feedback on the feelings of home learning (β = 0.068). The underlying mechanism may be caused by the fact that the feedback and support provided by teachers and students through ICT are identified as external resources, and these external resources do not directly affect students' home-based learning experience. Instead, students need to transform these external resources into internal self-directed learning self-efficacy, such as enhancing confidence in independent learning or feeling a sense of control over their own learning progress, in order to more significantly enhance students' experience of learning at home.

In this study, our analysis results supported the mediating role of information literacy. In addition to the feedback and support provided by teachers and students through ICT having a significant promoting effect on information literacy, information literacy has a significant positive effect on students' home-based learning experience. This result is similar to some existing literature, such as the studies of Yang and Kim (2014) and Akhyar et al. (2021) support that the improvement of information literacy can promote individual learning performance, and the research of Flierl et al. (2018) points out the promoting effect of information literacy on learning motivation. According to the Self-Determination Theory, better learning performance will enhance individual’s sense of satisfaction and achievement in the learning process (Deci & Ryan, 1985). It’s in line with the suggestions of Pekrun et al. (2017), they believed there is an interaction between learning performance and emotions, better grades predict positive emotions, and at the same time, positive emotions also predict better grades. Therefore, we think the effect of information literacy on home-based learning feelings may be through a similar pathway to self-directed learning self-efficacy, creating a positive cycle. The positive cycle is promoting academic performance will lead to more positive emotions, and more positive emotions will in turn lead to better academic performance.

5.4. The Moderating Role of Growth Mindset and Cooperation

The analysis results showed that growth mindset and cooperation have significant moderating effects in their paths, and these results are consistent with the existing literature.

Under the regulation of the growth mindset, as the growth mindset improves, the effect of feedback and support received by students using ICT on home learning feelings through self-directed learning self-efficacy increased. According to the data analysis results, in terms of the mediating effect of self-directed learning self-efficacy, compared with individuals with a low growth mindset (mean - SD), the mediating effect of individuals with a high growth mindset (Mean + SD) was enhanced by 33.72% ((0.115/0.086) -1), which had a more positive impact on their feelings about learning at home. This result also indicates that individuals with a high growth mindset will demonstrate stronger adaptability when facing learning difficulties, thereby enhancing their learning experience (Claro et al., 2016).

With the regulation of cooperation, as students' cooperation improves, the feedback and support obtained by students using ICT also enhanced their perception of home learning through information literacy. Based on the data analysis results, in terms of the mediating effect of information literacy, compared with students with lower cooperation (mean - SD), the mediating effect of students with higher cooperation (Mean + SD) increased by 278.57% ((0.053 − 0.014) -1), which had a more positive impact on their feelings about learning at home. This result theoretically proved that the trait of cooperation not only stimulates an individual's intrinsic motivation and learning performance, and build a more constructive and supportive learning environment (Nielsen et al., 2018), which in turn leads to a more satisfying learning experience, but also improve information literacy skills (Schroeter et al., 2020). However, it should be noted that although the mediating effect has improved significantly under the moderating effect, its actual effect size is still small. At the practical level, its impact on students' home-based learning experience is relatively limited. This may suggest that information literacy is probably only one of the smaller factors influencing students' home learning experience in the new learning mode and its effect is dependent on other external environmental variables, such as electronic device situation, and individual variables.

5.5. Total Effect

According to the analysis results, the total impact effect of students obtaining feedback and support through ICT on their feelings of home-based learning was β = 0.208, among which 67.15% of the effect was the mediating role of students' self-directed learning self-efficacy and information literacy. This result more directly reflects the importance of students' inner psychological resources and skills to their home learning experience in the distance learning mode. Further to this, the total effect of β= -0.031 in the pathway where students use ICT to get feedback and support on their home learning feelings through self-directed learning self-efficacy and information literacy, which in turn indirectly affects their psychosomatic symptoms. Among them, 68.75% of the effects were mediating effects. This result indicates that the feedback and support obtained by students using ICT promote their psychosomatic symptoms mainly by enhancing their internal psychological resources and skills, and thereby improving the home-based learning experience.

From this, we can find that when facing the new learning mode such as those transformed due to the COVID-19 pandemic, students' inner psychological resources and skills, as well as more positive learning experiences, may play an important role in protecting students' physical and mental health. And these two ratios indicate that students' self-directed learning self-efficacy and information literacy are the most important factors in distance learning when the purpose is to protect and promote students' learning experience and physical and mental health. This suggests that educational stakeholders need to pay more attention than ever to cultivating these abilities of students and provide corresponding support and resources.

5.6. Implications

From a theoretical perspective, this study, based on multiple theories such as the self-efficacy theory and the information literacy theory, has developed a deeper understanding of the learning adaptation mechanism of middle school students in the new learning mode. According to the research results, students' self-directed learning self-efficacy and information literacy have significantly promoted their home-based learning experience, thereby improving their physical and mental health. This not only provides empirical support for constructing the theory on the relationship between students' learning adaptability and mental health during the COVID-19 pandemic, but also emphasizes the dual effect of students' intrinsic psychological resources and skills in learning adaptability. Furthermore, this research addresses a current gap in research, as there are few existing studies exploring how students' intrinsic psychological resources and skills influence their remote or blended learning experiences and their psychosomatic wellbeing. This study reveals how students' intrinsic resources and skills assist them in adapting to remote or blended learning environments, maintaining effective learning and positive learning experiences, leading to enhanced psychosomatic wellbeing.

From a practical perspective, this study provides relatively clear suggestions for student cultivation for the educational side. First of all, schools and parents should pay more attention to cultivating students' self-directed learning ability. By fostering self-directed learning ability, including the setting of learning goals and the selection of learning strategies, their autonomy and sense of efficacy can be enhanced. Secondly, for the educational side, including schools and parents, improving students' information literacy should be one of their main educational goals. Information literacy, during distance learning, or in today's highly information-intensive age of the Internet, is critical to improving students' learning effectiveness, experience, and mental health. Finally, the school can design and launch an online interactive platform to enhance the interaction frequency among teachers and students, building a good atmosphere for common learning, and thereby improving students' information literacy and learning experience. By building a supportive environment and systematic cultivation, the improvement of students' self-directed learning self-efficacy and information literacy helps them enhance their learning efficiency and adaptability in today's Internet world, thereby promoting the healthy development of their physical and mental health.

5.7. Limitations

First, the current study focused on PISA data for Hong Kong students. Different cultural backgrounds may affect students' performance in variables such as self-directed learning self-efficacy and information literacy. Therefore, the conclusions of this study need to be used with caution in countries and regions with cultural backgrounds different from that of Hong Kong. Second, although all the path effects in this study were significant, some of them had a weak effect size. For example, the mediating path effect sizes of the feedback and support obtained by students through ICT on their psychosomatic symptoms were − 0.018 and − 0.006 respectively, indicating that the impact on students' physical and mental health in practice was limited. This suggests that there may exist other more important factors that have an impact on the path. Finally, although this study used path analysis to explore the causal relationship between variables and obtained significant results, the cross-sectional data gives us certain limitations in determining causal relationships. For example, the influence of time factors on variable changes cannot be caught. Therefore, caution is needed when applying the results of the current study to determining causal relationships among relevant variables.

5.8. Future Research

We suggest that future research can be conducted from three aspects. First, cultural background factors can be included in the analysis. By comparing the performance differences of students with different cultural backgrounds in this research, the influence of cultural differences on the influence mechanism could be examined. Second, future research could consider introducing the influence of time and using longitudinal experimental designs to compare the level of changes in variables at different times. This could further explore and confirm causal relationships and potential long-term impacts. Third, future research can further discuss whether there are differences in the impact of different types of feedback on students, such as immediate feedback and delayed feedback, and explore the differences in the effects on students' learning motivation, self-efficacy, and learning experience.

Abbas Q, Hussain S, Rasool S (2019) Digital Literacy Effect on the Academic Performance of Students at Higher Education Level in Pakistan. Global Social Sci Rev 4(1):108–116. https://doi.org/10.31703/gssr.2019(iv-i).14
Adnan M, Anwar K (2020) Online learning amid the COVID-19 pandemic: Students' perspectives. J Pedagogical Sociol Psychol 1(2):45–51. https://doi.org/10.33902/jpsp.2020261309
Akhyar Y, Fitri A, Zalisman Z, Syarif MI, Niswah N, Simbolon P, Purnamasari S, Tryana A, Abidin N, Z., Abidin Z (2021) Contribution of Digital Literacy to Students’ Science Learning Outcomes in Online Learning. Int J Elementary Educ 5(2):284. https://doi.org/10.23887/ijee.v5i2.34423
Alam BF, Bashir R, Hussain T, Abbas T, Malik SA, Jan SH, Khurshid M (2022) Online vs. traditional learning: A comparative analysis of student’s responses during COVID-19. Work 74(1):21–29. https://doi.org/10.3233/wor-220082
Association of College & Research Libraries (2000) Information literacy competency standards for higher education. Retrieved from https://alair.ala.org/handle/11213/7668
Avcı Ü, Ergün E (2019) Online students’ LMS activities and their effect on engagement, information literacy and academic performance. Interact Learn Environ 30(1):71–84. https://doi.org/10.1080/10494820.2019.1636088
Bandura A (1989) Regulation of cognitive processes through perceived self-efficacy. Dev Psychol 25(5):729–735. https://doi.org/10.1037//0012-1649.25.5.729
Bassi M, Steca P, Fave AD, Caprara GV (2006) Academic Self-Efficacy Beliefs and Quality of Experience in Learning. J Youth Adolesc 36(3):301–312. https://doi.org/10.1007/s10964-006-9069-y
Blackwell LS, Trzesniewski KH, Dweck CS (2007) Implicit Theories of Intelligence Predict Achievement Across an Adolescent Transition: A Longitudinal Study and an Intervention. Child Dev 78(1):246–263. https://doi.org/10.1111/j.1467-8624.2007.00995.x. Portico
Brandt WC (2020) Measuring Student Success Skills: A Review of the Literature on Self-Directed Learning. 21st Century Success Skills. National Center for the Improvement of Educational Assessment
Brouwers A, Tomic W (2000) A longitudinal study of teacher burnout and perceived self-efficacy in classroom management. Teach Teacher Educ 16(2):239–253. https://doi.org/10.1016/s0742-051x(99)00057-8
Bruce C (1997) Seven Faces of Information. Auslib
Bryce CI, Fraser AM (2022) Students' perceptions, educational challenges and hope during the COVID-19 pandemic. Child: Care. Health Dev 48(6):1081–1093. https://doi.org/10.1111/cch.13036
Caixin Global (2020), January 25 Hong Kong reports five confirmed cases of novel coronavirus. https://china.caixin.com/2020-01-25/101508335.html
Chang WC, Siew WF (2024) Exploring the Mediating Role of Student Satisfaction for the Relationship between Self-directed Learning Abilities and Elearning Outcomes. J Educators Online 21(4). https://doi.org/10.9743/jeo.2024.21.4.16
Chang Y-C, Bangsri A (2020) Thai Students' Perceived Teacher Support on Their Reading Ability: Mediating Effects of Self-Efficacy and Sense of School Belonging. Int J Educational Methodol 6(2):435–446. https://doi.org/10.12973/ijem.6.2.435
Charkhabi M, Azizi Abarghuei M, Hayati D (2013) The association of academic burnout with self-efficacy and quality of learning experience among Iranian students. SpringerPlus 2(1):677. https://doi.org/10.1186/2193-1801-2-677
Chen X, Breslow L, DeBoer J (2018) Analyzing productive learning behaviors for students using immediate corrective feedback in a blended learning environment. Comput Educ 117:59–74. https://doi.org/10.1016/j.compedu.2017.09.013
Cheng G (2017) The impact of online automated feedback on students’ reflective journal writing in an EFL course. Internet High Educ 34:18–27. https://doi.org/10.1016/j.iheduc.2017.04.002
Claro S, Paunesku D, Dweck CS (2016) Growth mindset tempers the effects of poverty on academic achievement. Proceedings of the National Academy of Sciences, 113(31), 8664–8668. https://doi.org/10.1073/pnas.1608207113
Clausen NR (2021) Progression of self-directed learning in PBL: Comparing consecutive semesters at AAU. J problem based Learn High Educ 9(1). https://doi.org/10.5278/ojs.jpblhe.v9i1.6373
Darlis V, Sari DK (2021) The Effectiveness of Blended Learning: The Impact of Student’s Characteristics and Digital Literacy on Student Performance. Proceedings of the 3rd International Conference on Educational Development and Quality Assurance (ICED-QA 2020) (pp. 561–566). https://doi.org/10.2991/assehr.k.210202.097
Deci EL, Ryan RM (1985) Conceptualizations of Intrinsic Motivation and Self-Determination. Intrinsic Motivation and Self-Determination in Human Behavior, 11–40. https://doi.org/10.1007/978-1-4899-2271-7_2
Ding L, Zhu Y (2021) The Effect of Self-Efficacy on Online Learning Efficiency in Cross-Cultural Perspective. Proceedings of the 2021 4th International Conference on Humanities Education and Social Sciences (ICHESS 2021). https://doi.org/10.2991/assehr.k.211220.287
Duckworth A (2016) Grit: The power of passion and perseverance, vol 234. Scribner, New York, NY
Dweck CS (2006) The truth about ability and accomplishment. Mindset: The new psychology of success. Ballantine Books
Education Bureau HKSAR (2020), February 12 Insider Perspective: Learning without boundaries—Continuing education during class suspension. https://www.edb.gov.hk/sc/about-edb/press/insiderperspective/insiderperspective2 0200212.html
Education Bureau HKSAR (2021), May 17 Quality Education Fund launches three-year programme to support schools’ e-learning [Press release]. https://www.info.gov.hk/gia/general/202105/17/P2021051700547.htm
Francisco R, Pedro M, Delvecchio E, Espada JP, Morales A, Mazzeschi C, Orgilés M (2020) Psychological symptoms and behavioral changes in children and adolescents during the early phase of COVID-19 quarantine in three European countries. Front Psychiatry 11:570164. https://doi.org/10.3389/fpsyt.2020.570164
Filius RM, de Kleijn RAM, Uijl SG, Prins FJ, van Rijen HVM, Grobbee DE (2018) Strengthening dialogic peer feedback aiming for deep learning in SPOCs. Comput Educ 125:86–100. https://doi.org/10.1016/j.compedu.2018.06.004
Flierl M, Bonem E, Maybee C, Fundator R (2018) Information literacy supporting student motivation and performance: Course-level analyses. Libr Inform Sci Res 40(1):30–37. https://doi.org/10.1016/j.lisr.2018.03.001
Garrison DR (1997) Self-Directed Learning: Toward a Comprehensive Model. Adult Educ Q 48(1):18–33. https://doi.org/10.1177/074171369704800103
Government Information Services Department, HKSAR (2022), December 30 Updated class arrangements set. https://www.news.gov.hk/eng/2022/12/20221230/20221230_190837_701.html
Hadiapurwa A, Jaenudin AS, Saputra DR, Setiawan B, Nugraha H (2021) The importance of learning motivation of high school students during the COVID-19 pandemic. Proceedings of the International Joint Conference on Arts and Humanities 2021 (IJCAH 2021). https://doi.org/10.2991/assehr.k.211223.217
Hayat AA, Shateri K, Amini M, Shokrpour N (2020) Relationships between academic self-efficacy, learning-related emotions, and metacognitive learning strategies with academic performance in medical students: a structural equation model. BMC Med Educ 20(1). https://doi.org/10.1186/s12909-020-01995-9
Jan SK (2015) The Relationships Between Academic Self-Efficacy, Computer Self-Efficacy, Prior Experience, and Satisfaction With Online Learning. Am J Distance Educ 29(1):30–40. https://doi.org/10.1080/08923647.2015.994366
Jin R, Wu R, Xia Y, Zhao M (2023) What cultural values determine student self-efficacy? An empirical study for 42 countries and economies. Front Psychol 14. https://doi.org/10.3389/fpsyg.2023.1177415
Kumpikaitė-Valiūnienė V, Aslan I, Duobienė J, Glińska E, Anandkumar V (2021) Influence of digital competence on perceived stress, burnout and well-being among students studying online during the COVID-19 lockdown: A 4-country perspective. Psychol Res Behav Manage 1483–1498. https://doi.org/10.2147/prbm.s325092
Leahy-Warren P (2014) Social support theory. Theories guiding nursing research and practice: Making nursing knowledge development explicit, 85–101
Li F, Cheng L, Wang X, Shen L, Ma Y, Islam AYMA (2025) The causal relationship between digital literacy and students’ academic achievement: a meta-analysis. Humanit Social Sci Commun 12(1). https://doi.org/10.1057/s41599-025-04399-6
Limniou M, Varga-Atkins T, Hands C, Elshamaa M (2021) Learning, student digital capabilities and academic performance over the COVID-19 pandemic. Educ Sci 11(7):361. https://doi.org/10.3390/educsci11070361
Liu K, Yang Y, Li M, Li S, Sun K, Zhao Y (2021) Parents’ and adolescents’ perceptions of parental involvement and their relationships with depression among Chinese middle school students during the COVID-19 pandemic. Child Youth Serv Rev 129:106190. https://doi.org/10.1016/j.childyouth.2021.106190
Magson NR, Freeman JY, Rapee RM, Richardson CE, Oar EL, Fardouly J (2021) Risk and protective factors for prospective changes in adolescent mental health during the COVID-19 pandemic. J Youth Adolesc 50(1):44–57. https://doi.org/10.1007/s10964-020-01332-9
Martínez-Bravo M-C, Sádaba-Chalezquer C, Serrano-Puche J (2020) Fifty years of digital literacy studies: A meta-research for interdisciplinary and conceptual convergence. El Profesional de La Información 29(4):1–15. https://doi.org/10.3145/epi.2020.jul.28
Nadolski RJ, Hummel HGK (2017) Retrospective cognitive feedback for progress monitoring in serious games. Br J Edu Technol 48(6):1368–1379. https://doi.org/10.1111/bjet.12503. Portico
Nielsen SG, Johansen GG, Jørgensen H (2018) Peer Learning in Instrumental Practicing. Front Psychol 9:339. https://doi.org/10.3389/fpsyg.2018.00339
Organisation for Economic Co-operation and Development (2023) PISA 2022 Database. https://www.oecd.org/en/data/datasets/pisa-2022-database.html
Organisation for Economic Co-operation and Development (2024) PISA 2022 Technical Report, PISA, OECD Publishing, Paris. https://doi.org/10.1787/01820d6d-en
Pekrun R, Lichtenfeld S, Marsh HW, Murayama K, Goetz T (2017) Achievement Emotions and Academic Performance: Longitudinal Models of Reciprocal Effects. Child Dev 88(5):1653–1670. https://doi.org/10.1111/cdev.12704. Portico
Peter J, Leichner N, Mayer AK, Krampen G (2015) Making information literacy instruction more efficient by providing individual feedback. Stud High Educ 42(6):1110–1125. https://doi.org/10.1080/03075079.2015.1079607
Prowse R, Sherratt F, Abizaid A, Gabrys RL, Hellemans KGC, Patterson ZR, McQuaid RJ (2021) Coping With the COVID-19 Pandemic: Examining Gender Differences in Stress and Mental Health Among University Students. Frontiers in Psychiatry, 12. https://doi.org/10.3389/fpsyt.2021.650759
Rasi P, Vuojärvi H (2017) Toward personal and emotional connectivity in mobile higher education through asynchronous formative audio feedback. Br J Edu Technol 49(2):292–304. https://doi.org/10.1111/bjet.12587. Portico
Schroeter C, Higgins L, Hank H, Stevenson C (2020) Teaching information literacy: A case study of the ripple effect in teamwork. Appl Econ Teach Resour (AETR) 2(2):1–13. https://doi.org/10.22004/ag.econ.302620
Schunk DH, DiBenedetto MK (2021) Self-efficacy and human motivation. Adv Motivation Sci 153–179. https://doi.org/10.1016/bs.adms.2020.10.001
Schweder S, Raufelder D, Wulff T (2020) Adolescents’ goals, self-efficacy, and positive emotions – how important is the learning context? Int J School Educational Psychol 10(1):1–17. https://doi.org/10.1080/21683603.2020.1791771
Son Y-L, Kim G-S, Cho E-H (2018) A Study on Effects of Self-Directed Learning Ability and Self- efficacy on Learner Satisfaction in Nursing Students. J Korean Clin Health Sci 6(2):1136–1146. https://doi.org/10.15205/KSCHS.2018.12.31.1136
Tan TX, Yi Z, Camras LA, Cheng K, Li Z, Sun Y, Chen N (2021) The effect of academic performance, individualistic and collectivistic orientation on Chinese youth’s adjustment. Soc Psychol Educ 24(5):1209–1229. https://doi.org/10.1007/s11218-021-09650-x
Toraman ÇETİN, Yurdal M, Aytuğ Koşan AY Ş. E. N., Baş B (2021) Tıp Öğrencilerinin Motivasyonel Kararlılık, Özyönelimli Öğrenme Hazırbulunuşluğu ve Başarı Yönelimleri Arasındaki İlişki. Tıp Eğitimi Dünyası, 20(62), 123–138. https://doi.org/10.25282/ted.957588
Truskowski S, VanderMolen J (2017) Outcomes and perceptions of annotated video feedback following psychomotor skill laboratories. J Comput Assist Learn 33(2):97–105. https://doi.org/10.1111/jcal.12167. Portico
Van Lancker W, Parolin Z (2020) COVID-19, school closures, and child poverty: a social crisis in the making. Lancet Public Health 5(5):e243–e244. https://doi.org/10.1016/s2468-2667(20)30084-0
Vygotsky L, Cole M (2018) Lev Vygotsky: Learning and social constructivism. Learning Theories for Early Years Practice. UK. SAGE Publications Inc, pp 68–73
Wang C, Schwab G, Fenn P, Chang M (2013) Self-Efficacy and Self-Regulated Learning Strategies for English Language Learners: Comparison between Chinese and German College Students. J Educational Dev Psychol 3(1). https://doi.org/10.5539/jedp.v3n1p173
Wang S-L, Wu P-Y (2008) The role of feedback and self-efficacy on web-based learning: The social cognitive perspective. Comput Educ 51(4):1589–1598. https://doi.org/10.1016/j.compedu.2008.03.004
World Health Organization (2020) Addendum to Fact Sheet 15 on National Implementation Measures for the International Health Regulations (2005) and COVID-19 [PDF]. https://extranet.who.int/sph/sites/default/files/document-library/document/FS15A_IHR_COVID19_EN_MAY_2020.pdf
Yang M, Kim J (2014) Correlation between Digital Literacy and Self-Regulated Learning Skills of Learners in University E-Learning Environment. Adv Sci Technol Lett 71:80–83. https://doi.org/10.14257/astl.2014.71.19
Yang Y, Zheng J (2024) Unfolding the Moderating Role of Gender in the Relationship Between Teacher Support and Students’ Well-being: Evidence from PISA 2022. Child Indic Res 17(6):2503–2526. https://doi.org/10.1007/s12187-024-10172-z
Yue J, Zang X, Le Y, An Y (2022) Anxiety, depression and PTSD among children and their parent during 2019 novel coronavirus disease (COVID-19) outbreak in China. Curr Psychol 41(8):5723–5730. https://doi.org/10.1007/s12144-020-01191-4
Zhan Y, Wan ZH, Chen J, Wang M (2023) How is Student Resilience Affected by Teacher Feedback, Teacher Support, and Achievement Goals? A Mediation Model Based on PISA 2018 Survey Data. Asia-Pacific Educ Researcher 33(3):671–682. https://doi.org/10.1007/s40299-023-00764-8

The authors declare no competing interests.

v6uq1n1.xlsx
01820d6den.pdf
PISA 2022 Technical Report

Download PDF

Version 1

posted

You are reading this latest preprint version

Adaptation Mechanisms of Hong Kong Secondary School Students to the Sudden Shift to Online Learning During the COVID-19 Pandemic: Path Analysis of PISA 2022 Data

Status:

Version 1

Abstract

Figures

1. Introduction

2. Literature Review

2.1. Feedback from Teachers and Classmates through ICT is a Social Resource

2.2. Students’ Self-Efficacy in Self-Directed Learning

2.3. Information Literacy of Students

2.4. The Role of Self-Directed Learning Self-Efficacy and Information Literacy

2.5. Growth Mindset and Cooperative Tendencies

2.6. Impact of Gender, Parental Involvement and Home Resources

3. Methodology

3.1. Data Source and Sample Characteristics

3.2. Measurement of Variables

3.3. Data Analysis

4. Results

4.1. Descriptive Statistics

4.2. Correlation Between Variables

4.3. Path Analysis

4.4. Individual Direct Path Estimates Between Variables

4.5. Mediating Pathways and Moderating Effects

4.6. Moderated Mediation Effects

4.7. Total Effects

5. Discussion

5.1. Potential Effect of Cultural Differences

5.2. The Influence of ICT Feedback on Students' Self-Directed Learning Self-Efficacy and Information Literacy

5.3. The Mediating Role of Self-Directed Learning Self-Efficacy and Information Literacy

5.4. The Moderating Role of Growth Mindset and Cooperation

5.5. Total Effect

5.6. Implications

5.7. Limitations

5.8. Future Research

References

Additional Declarations

Supplementary Files

Status:

Version 1