Introduction

Classical Chinese was an official written form of the Chinese language used in ancient China. In Hong Kong, learning to read classical Chinese was an integral part of the Chinese language curriculum. According to the curriculum and assessment guidelines, students at the secondary school level are expected to develop a solid understanding of the characteristics of classical Chinese vocabulary and sentence structures and have the ability to grasp the literal and profound meanings of classical Chinese texts (HKCDC-HKEAA, 2021). However, owing to linguistic differences between modern and classical Chinese, students experience great difficulties at the lexical level when reading classical Chinese texts (Lau, 2018; Chen & Chen, 2020; Dong et al., 2021). A lack of effective reading strategies and adequate content knowledge further affects students’ higher-order text comprehension (Sun, 2009; Zhao, 2004). Students’ low confidence and reluctance to read classical Chinese due to its difficulty result in poor motivation and hinder progress in developing proficiency (Lau, 2019a; Chi & Chiou, 2015; Zhang, 2021). Influenced by the traditional Confucian culture, the current approach to classical Chinese reading instruction is mainly teacher-dominated, focusing on knowledge transmission and text translation. This approach has been questioned by many Chinese scholars for its ineffectiveness in facilitating students’ development in classical Chinese reading and nurturing their motivation. (Lau, 2019b; Liu, 2020; Zhang, 2021).

Self-regulated learning (SRL), involving metacognition, motivation, and strategic actions, has been widely advocated as an effective form of learning (Schunk & Greene, 2018; Zimmerman, 2000). With technological advancements, the application of eLearning, which refers to learning through diverse forms of technology (Bates, 2005), has emerged as a crucial focus in recent research on promoting SRL (Lai & Hwang, 2021). While the effectiveness of both SRL instruction and eLearning on facilitating student learning have received support across many subject areas, they differ substantially from traditional approaches to teaching classical Chinese reading in Hong Kong. By adopting a flipped classroom (FC) model, this study made the first attempt to integrate instructional principles of SRL and incorporate out-of-class online learning activities into a two-year intervention program in the specific context of classical Chinese reading. Compared with most SRL intervention studies which generally last a short duration, this study adopted a two-year longitudinal design to guide students’ development of SRL through various stages. It aimed to provide valuable insights into the applicability and effectiveness of SRL and FC instruction in a school subject long dominated by teacher-centered instruction.

Development of SRL and SRL-based instruction

SRL encompasses three major components: metacognition, which involves awareness and competence in goal setting, monitoring, and self-evaluation; effective strategy use, which entails employing strategic behaviors to enhance learning processes and outcomes; and motivation, which encompasses high levels of self-efficacy and intrinsic motivation in learning (Usher & Schunk, 2018; Winne, 2018; Zimmerman, 2000). According to Schunk and Zimmerman (1997), learners must go through four stages to develop into self-regulated learners. In the observational stage, learners induce the major strategic features from watching a proficient model. They move to the emulation stage when they attempt to apply the learning strategies by emulating the patterns observed under the model’s supervision. Attainment of the self-controlled stage occurs when learners can independently use the strategies in transfer tasks. The final self-regulation stage is achieved when learners can systematically adapt their performance to changing personal and contextual conditions.

Instructional factors play an important role in supporting students’ development into mature self-regulated learners. Based on previous studies (Cousins et al., 2022; Dignath & Veenman, 2021; Dignath et al., 2008; Housand & Reis, 2008; Jayawardena et al., 2019; Lombaerts et al., 2009; Mohammadi et al., 2020; Neitzel & Connor, 2017; Perry & Rahim, 2011; Stoeger et al., 2015; van Grinsven & Tillema, 2006), four major principles are identified to group various effective instructional practices promoting student SRL: (1) Task nature (T): direct instruction on subject knowledge and learning strategies helps students establish a good foundation of learning and open tasks enhance students’ motivation and facilitate knowledge transfer. (2) Teacher support (S): teachers use various types of scaffolding techniques to support students’ development through various stages of self-regulatory skills. (3) Student autonomy (A): student-directed activities enable students to practice their SRL skills by gaining control over the learning process. (4) Evaluation practices (E): mastery-oriented and student-involved assessments are adopted to enhance students’ metacognitive knowledge and skills in goal setting, monitoring, and self-evaluation.

The above instructional principles served as a “TSAE” framework in the authors’ previous studies to support teachers’ implementation of SRL instruction in Chinese language classes (Lau, 2011, 2013). While the positive impact of SRL instruction on enhancing students’ strategy use, motivation, and reading comprehension was supported in these studies, Chinese language teachers tended to adopt the principles with more emphasis on the supportive role of the teacher (T and S) than those offering students control (A and E). Recently, the effectiveness of SRL instruction in enhancing Chinese students’ classical Chinese reading was examined in a one-year intervention study (Lau, 2020). Although the findings showed that SRL instruction was more effective than traditional teacher-centered instruction in enhancing students’ prior knowledge and comprehension performance, no significant changes in students’ strategy use and motivation were found in the quantitative measures. One possible reason is that Chinese students who are used to learning under teacher-centered instruction need more time and support to adapt to the new learning approach of SRL (So et al., 2019). Moreover, the effects of SRL instruction may be limited by students’ weak foundation of classical Chinese reading (Otto & Kistner, 2017).

Promoting SRL through FC approach

Researchers propose that SRL and eLearning are mutually reinforcing (Blau & Shamir-Inbal, 2017; Palalas & Wark, 2020). On one hand, learning online provides students with a high degree of autonomy by allowing them to learn at their own pace per their individual needs (Akçayır & Akçayır, 2018; Bond, 2020). Different eLearning tools can serve as scaffolds for students during the cyclical phases of SRL (Bai et al., 2022; Perez-Avalverdelvarez et al., 2022). Students’ motivation, especially self-efficacy, can be enhanced in eLearning when they execute control over their learning and witness their progress toward self-set goals (Bai et al., 2022; So et al., 2019). On the other hand, since eLearning is more demanding than the traditional learning environment, students must possess adequate self-regulatory skills (Lai & Hwang, 2021; Sletten, 2017) and receive proper teacher and instructional design support (Akçayır & Akçayır, 2018; Dianati et al., 2022) to learn effectively.

FC—a widely adopted blended pedagogical approach comprising in-class instruction and out-of-class eLearning—is one way to maximize the benefits and compensate for the limitations of eLearning. FC reverses the traditional teacher-centered classroom into student-centered learning by having students study content before class through online learning to free in-class time for more interactive and higher-level learning activities (Berrett, 2012; Fulton, 2012; Strelan et al., 2020). Different from the traditional drilling approach of classical Chinese learning, the out-of-class eLearning component of FC provides students more chances to practice various SRL skills through diversified pre-class preparation and post-class application activities, which are more effective in enhancing their strategic reading skills and intrinsic motivation (Al Mulhim, 2021; Rasheed et al., 2020). Compared with solely web-based learning, the FC approach is more suitable for students with weak classical Chinese reading ability because it places more emphasis on the teacher’s role as a facilitator in face-to-face teaching to enable students to perform SRL (Dianati et al., 2022; Öztürk & Çakıroğlu, 2021; Sletten, 2017). The effectiveness of integrating SRL intervention into FC courses to enhance student learning is substantially supported across studies on different school subjects (e.g., Lo et al., 2021; Sletten, 2017; Sun et al., 2018), particularly on language learning (e.g., Öztürk & Çakıroğlu, 2021).

Certain challenges, such as students’ inadequate preparation, poor quality of eLearning materials, and lack of teacher support, have also been reported in studies using FC (Akçayır & Akçayır, 2018; Rasheed et al., 2020). To tackle these problems, Blau and Shamir-Inbal (2017) proposed a “Re-designed model of flipped learning (RDFC model)” that offers a holistic pedagogical design for FC. Unlike the traditional approach where the acquisition of new content primarily occurs through pre-class video watching, the RDFC model emphasizes extensive independent learning with teacher scaffolding and peer collaborations to foster active learning in both in-class and out-of-class settings, facilitating knowledge construction before, during, and after lessons (see Fig. 1).

Fig. 1
figure 1

Re-designed model of flipped learning (extracted from Blau and Shamir-Inbal (2017), pp.77)

Full size image

Theoretical framework and purpose of the study

Given the potential benefits of SRL instruction and technology on enhancing students’ learning of classical Chinese reading discussed above, this study aimed to design and evaluate the effectiveness of a two-year flipped SRL intervention program in enhancing Chinese secondary school students’ classical Chinese reading ability and motivation. The literature on the four developmental stages of SRL, the TSAE framework, and the RDFC model was synthesized into a cohesive theoretical framework for this study (see Fig. 2). When integrating the TSAE framework and the four developmental stages of SRL, the importance of different instructional principles was adjusted according to students’ developmental stages of SRL (Schunk & Zimmerman, 1997). Specifically, more emphasis was placed on the T and S principles in the first two stages of SRL to help students establish a solid foundation for learning through guidance and support (Neitzel & Connor, 2017). As students progressed to later stages, greater focus was given to the A and E principles to promote independent learning through student-directed tasks and self-assessment (van Grinsven & Tillema, 2006). In reference to the RDFC model, various out-of-class eLearning and in-class activities, designed based on the TSAE principles and students’ SRL developmental stages, were organized in a coherent manner to provide students with extensive independent learning supported by teacher scaffolding and peer collaboration to facilitate knowledge construction, application, and further development before, during, and after lessons (Blau & Shamir-Inbal, 2017).

Fig. 2
figure 2

The theoretical framework of the study. The abbreviations used in the figure represent the four major instructional principles of the TSAE framework. Specifically, T refers to the nature of Task; S refers to teacher Support; A refers to student Autonomy; E refers to the Evaluation practices

Full size image

Specifically, this study addressed three research questions:

RQ1

Can all SRL instructional principles be feasibly implemented in experimental schools’ regular Chinese language lessons?

RQ2

Does SRL instruction have more positive effects on enhancing students’ strategy use, motivation, and performance in classical Chinese reading than traditional teacher-centered instruction?

RQ3

Does the inclusion of out-of-class eLearning activities produce more positive effects on enhancing students’ learning of classical Chinese reading than when only SRL instruction is adopted?

Methodology

Experimental design

This study adopted a quasi-experimental design using a pre-test-post-test treatment–control group. All participants were assigned to one of the following conditions: SRL instruction plus out-of-class eLearning (EG-I+E), SRL instruction only (EG-I), and control (CG). Students in EG-I+E and EG-I had the same number of Chinese language lessons, reading materials, and in-class SRL instruction, but only those in EG-I+E were assigned out-of-class eLearning activities. In the CG, teachers employed a traditional teacher-centered approach that focused on explaining the vocabulary, sentences, content, and writing techniques of each classical Chinese text in the students’ textbooks. To ensure that each group spent a similar time studying classical Chinese in the out-of-class context, EG-I students received post-class paper-based SRL tasks and CG students received all classical Chinese texts in the intervention package in the form of traditional comprehension exercises (see Table 1). The study adopted a concurrent mixed methods design to collect both quantitative and qualitative data to evaluate the implementation and effectiveness of the intervention. This design is useful in combining the advantages of quantitative and qualitative methods and utilizing qualitative data to triangulate and supplement the quantitative results (Creswell et al., 2003). The reading comprehension test and questionnaire pre- and post-tests were administered to the entire student sample to objectively assess students’ reading ability and motivation. After each phase, semi-structured interviews were conducted with teacher and student representatives from both EGs to delve into their specific comments regarding the implementation of the intervention program and what impacts it had on students’ reading ability and motivation throughout the program.

Table 1 Different treatments of the three groups
Full size table

Participants

Three secondary schools in Hong Kong agreed to participate in the study voluntarily. The sample comprised 352 secondary students (172 boys; 179 girls; 1 unreported). In the first year of the study, the students were in the 8th grade with ages ranging from 12–14 years (mean = 12.94 years, SD = 0.63). In the second year of the study, the students were in the 9th grade with ages ranging from 13–15 years (mean = 13.72 years, SD = 0.66). Students from each school were assigned to one of three conditions: EG-I+E (N = 115), EG-I (N = 117), and CG (N = 120). As the intervention was integrated into the schools’ regular curriculum, the assignment of conditions was determined based on the readiness of each school to implement SRL instruction and the FC approach. To ensure comparability among student groups, the three schools were selected based on similar student achievement levels and socioeconomic status. All three schools were government subsidized schools locating in urban areas of Hong Kong and mainly enrolled students with a moderate level of achievement from nearby public estates. They used the same Chinese language textbook and had a comparable number of Chinese language lessons per week. The study design was approved by the ethics review board of the authors’ university. Informed consent was obtained from the schools and participants. Quantitative data was collected from the entire student sample. A purposive sampling strategy was employed to select participants for the interviews. One teacher from each of the experimental schools, who was responsible for coordinating the teaching of Chinese language lessons, was invited to participate in teacher interviews. Sixteen student representatives (two students, one with high classical Chinese reading ability and one with low classical Chinese reading ability from each class of each experimental school) were nominated by their class teachers to represent their classmates in the student interviews.

Intervention program design

The intervention program comprised four phases, starting from the beginning of the academic year of 2022/2023 and ending at the end of the academic year of 2023/2024. The program design was guided by the theoretical framework outlined in Fig. 2. One reading module was designed for each phase and implemented in one of the semesters of the two academic years. Each module selected classical Chinese texts based on a humanistic theme of ancient Chinese culture. Various types of reading strategies were arranged in each module to develop students’ classical Chinese reading ability. Interesting materials and student-centered activities were used to encourage students’ active application of the learned strategies in solving various reading tasks. Pre- and post-class eLearning activities were designed for students to carry out pre-class preparation and practice the strategies in new tasks.

During Phase 1, students focused on learning and applying word interpretation strategies specifically designed to aid students in interpreting the meaning of classical Chinese words based on linguistic differences between modern and classical Chinese. An online platform provided teaching videos, web-based materials, and online games for students’ out-of-class eLearning activities. In Phase 2, cultural knowledge, cognitive and metacognitive strategies were added to enhance students’ text-level comprehension, such as analyzing the main ideas, implicit meaning, and structure of the texts. In Phase 3, most class time was spent on open and interactive activities to foster in-depth and higher-level text comprehension. The degree of metacognitive control was also increased in this phase. In Phase 4, all learning tasks were student-led. Students were responsible for preparing pre-class teaching videos and leading in-class activities for their self-selected texts. The detailed instructional design of each phase is shown in Appendix 1.

Instruments

Classical Chinese reading comprehension test

Two reading comprehension tests were designed to assess students’ classical Chinese reading ability in each academic year, respectively. Each test comprised one narrative and one argumentative text. Ten open-ended “word interpretation” questions assessed students’ word-level comprehension and eighteen “text comprehension” (half multiple-choice and half open-ended) questions assessed text-level comprehension, including identifying information, integration, inference, and evaluation. All items are scored from 0 to 2 points. The total score of each test was 56. Rubrics with clear assessment criteria were used to guide the marking of open-ended questions. Two trained research assistants were responsible to mark the tests. Initially, they collectively assessed 10% of the open-ended questions. Discrepancies in scoring were discussed and resolved to ensure coding validity. Once the agreement between the two raters exceeded 90%, indicating good interrater reliability (Stemler, 2019), the remaining tests were scored independently.

SRL reading instruction questionnaire

This questionnaire was used and validated in the authors’ previous study (Lau, 2020). It assessed how students perceive the degree of different SRL instructional principles in their reading classes. Five subscales with 28 items rated on a 5-point Likert scale were derived from the four principles of the TSAE framework: task nature-instructional content, task nature-instructional materials and activities, teacher support, student autonomy, and evaluation practices.

Classical Chinese reading strategy questionnaire

This questionnaire, comprising two subscales with 26 items, was developed and validated in the authors’ previous studies (Lau, 2018, 2020). The “word interpretation strategies” subscale was designed to measure students’ use of strategies to interpret the meaning of classical Chinese words based on linguistic differences between modern and classical Chinese (Hu, 2010; Wei, 2009). Another subscale, “text-level strategies” subscale was adapted from the Metacognitive Awareness of Reading Strategy Inventory (Mokhtari & Reichard, 2002) to measure students’ use of different cognitive and metacognitive strategies to facilitate their comprehension of the main ideas, implicit meaning, and structure of a text. Using a five-point Likert scale, students reported how frequently they used each strategy during CC reading.

Classical Chinese reading motivation questionnaire

This questionnaire was adapted from the Chinese Reading Motivation Questionnaire (Lau, 2004) and validated in the authors’ previous studies (Lau, 2018, 2020). Two subscales, each with 6 items rated on a 5-point Likert scale, were chosen from the questionnaire to measure students’ self-efficacy and intrinsic motivation in classical Chinese reading because these two types of motivation were the core components of SRL.

Student and teacher interviews

Semi-structured student and teacher group interviews were conducted among the two EGs in each phase. Each interview consisted of two sets of questions. The first set focused on students’ learning progress during the program, including students’ improvements in their reading ability, strategy use, and motivation in classical Chinese reading. The second set was based on the TSAE framework to solicit teachers’ and students’ comments on different features of the in-class instruction and out-of-class eLearning activities designed in the intervention program. The interview questions were mainly developed based on the study’s theoretical framework, with some adjustments made considering the quantitative findings obtained in the early stage (Table 2).

Table 2 Reliability estimates for the instruments
Full size table

The English translation of all questionnaires, interview questions, sample items, and rubrics of reading tests are provided in Appendices 2, 3, 4 and 5.

Procedures and data analysis plan

The appropriateness of the intervention design, questionnaires, and interview questions was reviewed by two scholars, four experienced teachers, and all EG teachers. Since all questionnaires used were validated in previous studies, only the newly developed reading tests were piloted among 319 students in three secondary schools with the same achievement level as the participating schools. Refinements were made based on the reviewers’ feedback and the pilot study results.

Two teacher workshops were held for all EG teachers to enhance their understanding of SRL instruction and introduce the design of the intervention program. Each of the reading modules was arranged in one semester and delivered by EG teachers in their regular lessons. One classroom observation was arranged for each module and researcher-teacher collaborative meetings were held before and after each module to ensure fidelity of implementation and to support teachers’ preparation and evaluation of their implementation. The pre- and post-reading comprehension tests were administered at the beginning and end of each academic year, respectively. The questionnaire pre-test was administered at the beginning of the first academic year and two waves of post-test were administered at the end of each academic year. The reading test took 30 min, and the questionnaire took about 10 min to complete. They were administered to students during regular class periods by their teachers. Standard instruction on the administration procedures was prepared for the teachers. Descriptive statistics and repeated measures ANOVA were performed to examine changes among different treatment groups. After each module, face-to-face semi-structured group interviews were conducted with teachers and students separately to gather their views on the implementation and effectiveness of that module. Each teacher interview lasted around one hour, while each student interview lasted approximately 30 min. All interviews were audiotaped for transcription. The transcripts were analyzed through the process of open coding, developing inductive categories, making comparisons and connections across informants and situations, and finally, generating larger themes to understand how teachers and students in the two EGs perceived the implementation and effectiveness of different types of activities in the intervention program for enhancing students’ learning of classical Chinese reading and their underlying factors.

Results

Intervention program implementation

As indicated by the mean scores of different subscales of students’ perceived reading instruction (Table 3), all groups perceived a higher degree of the principles of T (instructional content) and S than other principles in the pre-test. Repeated measures ANOVA using time as the within-subject factor and student group as the between-group factor were performed to compare changes among different groups. The significant time × group interaction effects revealed in all SRL instructional principles indicated that there were significant differences among the three groups in terms of the changes in their perceived reading instruction measured at different time points, with the largest effect observed for Principle A (see Table 3). Follow-up repeated measures were separately conducted for each school. As shown in Table 4, both EGs perceived significant increases in the degree of most SRL instructional principles between the pre-test and the first post-test and maintained a high degree of all principles in the second post-test. Although CG also perceived significant increases in some of the principles, the magnitude of most changes was much smaller than the two EGs. These findings addressed RQ1, providing support for the feasibility of implementing all SRL instructional principles within the regular Chinese language lessons of the experimental schools.

Table 3 Results of repeated measures ANOVA involving all treatment groups
Full size table
Table 4 Results of repeated measures ANOVA conducted separately for each treatment group
Full size table

In the interviews, EG teachers reported following all the instructional designs of the intervention package to deliver the intervention program. As reported by the teachers, while EG-I+E completed all lessons on time, EG-I sometimes lagged behind the planned schedule. EG-I+E teachers reflected that the out-of-class e-learning component of the program saved a lot of class time and provided a high degree of autonomy for their students. Contrastingly, EG-I teachers took longer to shift gradually from teacher-centered instruction to SRL instruction.

The descriptions of EG students’ learning experiences reflected that their teachers generally adhered to the major principles of SRL instruction when delivering the intervention program. In the interviews, most students were able to give concrete examples of the strategies and cultural knowledge they learned, the interactive classroom activities, the support received from their teachers, the kinds of autonomy they had, and different types of assessment activities. The word interpretation strategies and student-directed learning activities were the most impressive elements of the intervention program for students, highlighting the significant difference between the intervention program and their previous classical Chinese learning experiences. Students found the activities more interesting and that they facilitated more in-depth comprehension than previous teacher-centered instruction. However, some of the students opined that certain self-directed activities were substantially difficult for them. While students generally held positive attitudes toward peer evaluation activities, they did not like setting goals and self-evaluating. Since students were accustomed to following the predetermined learning objectives set by teachers and relying on teacher evaluations, they faced challenges in setting specific goals and lacked the motivation to make self-evaluations. Regarding the eLearning component of the EG-I+E program, most students expressed positive feedback about the various out-of-class eLearning activities that were implemented. However, some students mentioned that they did not complete all the out-of-class activities owing to the heavy workload of homework or their laziness.

Effectiveness of the intervention on reading comprehension

Students of all groups performed unsatisfactorily in the reading comprehension pre-tests. The findings of repeated measures ANOVA indicated a significant time effect on reading comprehension test scores in each academic year, but no significant time × group interaction effects were found (see Table 3), indicating that all groups demonstrated similar degrees of improvement in their reading comprehension post-test scores. In response to RQ2 and RQ3, these findings suggest that SRL instruction and the inclusion of out-of-class eLearning activities did not have more positive effects on enhancing students’ classical Chinese reading performance than traditional teacher-centered instruction.

In the interview, all student representatives from the two EGs regarded classical Chinese reading as very difficult before the intervention program. Since most reading difficulties they encountered in classical Chinese reading were at the lexical level, they found that reading classical Chinese texts became less difficult after learning the word interpretation strategies. Additionally, since the texts in each module centered around a humanistic theme of ancient Chinese culture, and interactive and high-level activities were designed to facilitate students’ text-level comprehension by applying their knowledge of the theme, students expressed that their understanding of cultural knowledge and the meanings of texts deepened through active participation in these activities. EG teachers also observed obvious improvements in students’ word translation and comprehension of the in-depth meanings of classical Chinese texts.

Effectiveness of the intervention on strategy use

All students reported a moderate to low frequency of using classical Chinese reading strategies in the pre-test measures. The results of repeated measures ANOVA indicated significant time × group interaction effects in both types of strategies (see Table 3). Findings of follow-up analyses indicated that while students in the two EGs showed similar significant increases of their strategy use between the pre-test and the first post-test and maintained a high frequency of using the strategies in the second post-test, CG did not have any significant change in their strategy use throughout the two academic years (see Table 4). These findings indicated that SRL instruction had more positive effects on enhancing students’ strategy use than the traditional teacher-centered reading instruction (RQ2) but the inclusion of out-of-class eLearning activities did not yield additional positive effects beyond SRL instruction (RQ3).

Both EG teachers and students held positive perceptions of the word interpretation strategies in the program. Teachers observed their students gradually developing habits of using strategies to interpret unfamiliar words when reading a new classical Chinese text. Most EG-I+E students said they had already mastered the strategies in the first semester and agreed that the strategies were useful in understanding classical Chinese. Some EG-I+E students mentioned that they always played the program’s online games to practice the strategies outside of class. EG-I students’ learning of the word interpretation strategies was slower than EG-I+E students but most said they mastered the strategies in the second semester. In the second year of the intervention program, teachers and students reported that most students used the word interpretation strategies more skillfully and automatically and, thus, teachers spent much less time explaining the words in class. In comparison, there was less discussion among teachers and students about the text-level strategies. However, they acknowledged the usefulness of certain text-level strategies such as selection, questioning, prediction, and text structure strategies in enhancing students’ text understanding.

Effectiveness of the intervention on reading motivation

All groups reported a low level of self-efficacy and intrinsic motivation in classical Chinese reading in the pre-test. The results of repeated measures ANOVA indicated a significant time × group interaction effect in self-efficacy (see Table 3). While the two EGs reported similar significant increases in their self-efficacy and intrinsic motivation in the post-test measures, the CG did not have any significant changes in both types of motivation throughout the two academic years (see Table 4). These findings indicated that SRL instruction had more positive effects on enhancing students’ motivation than the traditional teacher-centered reading instruction (RQ2), but the inclusion of out-of-class eLearning activities did not yield additional positive effects beyond SRL instruction (RQ3).

In the interviews, teachers and students commented positively on the effectiveness of the intervention program in enhancing students’ classical Chinese reading motivation. Regarding self-efficacy, students said they after learning the word interpretation strategies, they were not as afraid of classical Chinese reading as before. Teachers also observed that their students became more willing to interpret classical Chinese words by themselves instead of getting model answers from the teachers or the internet. During the second year, most students demonstrated increased confidence in engaging in higher-order activities and were able to discuss the text meaning in-depth. For intrinsic motivation, EG teachers said their students used to be extremely reluctant to read classical Chinese texts. During the intervention program, their students participated more actively in the interactive activities and the learning atmosphere significantly improved. Most EG students said they increased their intrinsic interest in reading classical Chinese because the instructional materials and activities were interesting and relevant to their daily lives. EG-I+E students also found the eLearning activities to be more interesting and attractive than traditional classroom teaching.

Discussion

This study investigated the feasibility and effectiveness of using a FC approach to combine SRL instruction and eLearning in a longitudinal classical Chinese reading intervention program. Owing to the influence of traditional Confucian culture and students’ weak classical Chinese reading ability, teacher-centered instruction predominates in classical Chinese reading classes (Lau, 2019b; Liu, 2020; Zhang, 2021). Different from the authors’ previous collaborative studies with front-line teachers, which found that Principles A and E were not easy to implement in Chinese classes (Author, 2011, 2013), this study adopted a rigorous experimental design to ensure teachers strictly follow the intervention package. In the interviews, EG teachers said it was easy to implement the intervention by following this. The quantitative results revealed significant increases in EG students’ perceived degrees of all SRL instructional principles. Notably, the largest change was observed for Principle A, which signifies the major difference between SRL instruction and traditional teacher-centered instruction.

Consistent with many SRL intervention studies (Bai et al., 2022; Cousins et al., 2022; Mohammadi et al., 2020; Otto & Kistner, 2017; Stoeger et al., 2015), the significant improvement of EG students’ reading performance in this study supports the effectiveness of SRL instruction in facilitating student learning. Considering that students’ major difficulties in classical Chinese reading were at the lexical level (Lau, 2019b; Chen & Chen, 2020), the word interpretation strategies designed for the intervention program were specifically adapted based on major linguistic differences between modern and classical Chinese to provide concrete ways for students to infer and translate the meaning of classical Chinese words (Chen, 2020; Hu, 2010). The significant increase in EG students’ strategy use and their positive comments on the usefulness of word interpretation strategies echoed the views of many researchers that training students in domain-specific strategies is especially effective in enhancing students’ SRL in specific subject areas (Dignath et al., 2008; Greene et al., 2015; Schunk & Greene, 2018). Additionally, EG teachers and students commented that students’ text-level comprehension was further enhanced by learning more cultural knowledge of ancient China and participating actively in interactive and higher-order activities (Dignath & Veenman, 2021; Perry & Rahim, 2011). In contrast to the view that teacher-centered instruction is less effective than SRL instruction in facilitating students’ classical Chinese reading ability (Lau, 2019b; Liu, 2020; Zhang, 2021), this study found that CG students also demonstrated significant improvement in the reading post-tests even though they did not increase their use of reading strategies. Since CG students read the same set of reading materials as EG students, these findings suggest that while EG students improved their performance through consciously applying different strategies and cultural knowledge in reading classical Chinese texts, CG students improved their reading comprehension by cumulating more learning experiences of classical Chinese reading.

During the intervention program, EG students showed significant increases in their self-efficacy and intrinsic motivation in classical Chinese reading while CG students’ self-efficacy and intrinsic motivation remained low throughout the two academic years. These findings support that SRL instruction is more effective than traditional teacher-centered instruction in enhancing students’ motivation in classical Chinese reading (Dignath et al., 2008; Perry et al., 2002). Owing to immense difficulties that students encounter during reading classical Chinese texts, they usually have poor self-efficacy in classical Chinese reading (Lau, 2019a; Chi & Chiou, 2015). Consistent with the view that students’ self-efficacy can be enhanced through strategy learning (Li et al., 2022; Schunk & Ertmer, 2000), EG students indicated that after learning word interpretation strategies, they had more confidence in reading classical Chinese texts. Contrastingly, although CG students improved their performance in the reading comprehension post-tests, their self-efficacy was not enhanced because they had no idea how they could perform better. Since many students regard classical Chinese as irrelevant to their daily life, they usually have poor intrinsic motivation and read classical Chinese texts solely to pass examinations (Lau, 2019a; Zhang, 2021). In the interviews, most EG students agreed that their motivation was improved because they could apply the cultural themes of the texts in discussing daily life issues and the interactive and student-directed activities made learning classical Chinese interesting. Contrastingly, the consistently low intrinsic motivation of CG students confirmed that the transmissive approach of traditional reading instruction failed to arouse students’ interest in classical Chinese reading (Lau, 2019b; Liu, 2020).

Both EGs demonstrated significant increases in their perceived degree of SRL instruction and all outcome measures, suggesting that the integration of out-of-class eLearning into SRL instruction did not yield additional positive effects on student learning. The major advantage of eLearning in promoting SRL is providing a flexible and autonomous environment for students to practice their SRL skills based on their own needs and pace (Akçayır & Akçayır, 2018; Blau & Shamir-Inbal, 2017; So et al., 2019). Since student autonomy is emphasized in SRL instruction (Perry & Rahim, 2011; van Grinsven & Tillema, 2006), even in the absence of out-of-class e-learning activities, EG-I students also experienced a significant increase in the degree of Principle A in their reading lessons. The similar nature of the in-class instruction among the two EGs might produce similar positive impacts on their learning. Nevertheless, some benefits of using FC to facilitate the implementation and effectiveness of SRL instruction were revealed in the qualitative results. Both teachers and students of the EG-I+E commented positively on the flexible and interesting nature of the intervention program’s eLearning activities. Teachers affirmed that the out-of-class eLearning component of FC facilitated a smoother implementation of the intervention program by freeing class time to implement higher-order and interactive in-class activities (Berrett, 2012; Strelan et al., 2020). Students also had deeper impressions of the reading strategies they learned from the pre-class teaching videos because they could review the videos many times (Bond, 2020) and practice the strategies through online games (Lin et al., 2020; Wang, 2021).

Despite EG students’ obvious improvements in their reading comprehension and motivation, notably, students were less receptive to metacognitive control activities, and some of the students opined that some higher-order activities were substantially difficult for them. Since SRL requires students to take responsibility for their own learning, students accustomed to teacher-centered instruction tend to be more hesitant when asked to self-direct their learning (Dignath & Buettner, 2008; Zimmerman, 2008). In addition, consistent with previous studies which found that students’ inadequate preparation is the most common problem observed in FC (Akçayır & Akçayır, 2018; Blau & Shamir-Inbal, 2017; Rasheed et al., 2020), this study also found that some students did not complete eLearning assignments before class. Given that Chinese students are used to playing a passive role and have low motivation in learning classical Chinese reading, they may need more time to develop their metacognitive skills and take full responsibility for their learning (Otto & Kistner, 2017; So et al., 2019). Providing clear guidelines and incentives for out-of-class learning (Akçayır & Akçayır, 2018) and demonstrating concrete benefits of metacognitive control activities (Winne, 2018) can also motivate students to participate in these self-directed learning activities.

Given the practical constraints of implementing the intervention program in the participants’ regular classes, certain limitations of the study’s design and suggestions for future studies should be noted. First, since students in the three conditions came from three different schools, controlling for all confounding factors proved challenging, despite efforts to match the backgrounds of the different groups by accounting for various student- and school-level factors. Future studies should consider assigning the same teachers within the same school to teach all student groups to minimize potential confounding effects resulting from teacher- or school-specific attributes that could impact the treatment effects. Second, although all participating schools were ranked at the same achievement level, it should be noted that the three student groups were not entirely identical in all pre-test measures. While group differences were tackled by the statistical control, students’ initial conditions may influence intervention outcomes. Therefore, pre-test measures should be used to screen participants for different groups to ensure all groups are similar in important variables before the interventions. Third, the design of this study was mainly quantitative, with qualitative data used to complement the main analysis, and interview results were analyzed in a straightforward manner. Therefore, future studies should consider employing more sophisticated qualitative analysis techniques to gain deeper insights into the reasons and mechanisms underlying the effectiveness of the intervention.

Conclusion

The study’s findings provide solid empirical support for the applicability and effectiveness of using FC to integrate SRL instruction and eLearning to facilitate student learning in the context of classical Chinese reading, a traditional school subject long dominated by teacher-centered instruction. Although students of all conditions improved their reading performance at the end of the academic year, only students who received SRL instruction significantly improved their strategy use and motivation, suggesting that SRL instruction was more effective than traditional teacher-centered instruction in developing students into strategic, self-efficacious and intrinsic motivated learners. These successful outcomes are attributed to several important features of the intervention design. First, the rigorous experimental design, incorporating a full intervention package based on the comprehensive TSAE framework, ensured teachers’ adherence to all major instructional principles of SRL during the intervention. Second, the intervention program, designed in a module-based format and aligned with the current Chinese language curriculum in Hong Kong secondary schools, facilitated its integration into the regular school curriculum through a longitudinal design. It allowed students to establish a solid foundation of classical Chinese reading in the first stages and gradually develop their higher-level comprehension in the higher stages of SRL development. Third, the combination of out-of-class eLearning with SRL instruction in FC facilitated a smoother implementation of the intervention program and enhanced students’ strategy learning and motivation. Last, EG teachers’ and students’ positive comments on the cultural themes of the modules and word interpretation strategies—designed based on the specific nature of classical Chinese learning—suggest that embedding SRL instruction in the design of subject-specific interventions should achieve more robust effects in a particular subject area. This is especially important when implementing Western-oriented instruction in subjects rooted in non-Western culture.