Differentiated instruction in L2 teaching: two extensive reading programmes conducted during COVID-19 pandemic

ABSTRACT

Purpose

This study investigates two extensive reading (ER) programmes carried out in a secondary school in Beijing during the COVID-19 pandemic, with the aim of analysing how differentiated instruction (DI) is reflected in the ER implementation and how this forced online learning inform foreign/second language (L2) pedagogy in relation to ER and DI.

Design/methodology/approach

Adopting an exploratory case study approach, the study collected data from multiple sources: interview with six students and two teachers; questionnaire survey (N = 74 + 27); teachers’ reflective journal; documents related to the two reading programmes. Data collection lasted for a semester (six months). Thematic analysis was used to code and interpret data.

Findings

Following the five-category framework for differentiated instruction (Tomlinson, C. A., and T. Moon. 2013. Assessment and Student Success in a Differentiated Classroom. Alexandria: Association for Supervision & Curriculum Development), this study collects evidence of DI implementation in terms of content (three-tiered system of providing reading materials), process (personalised reading frequency and quantity), and product (diversified forms/platforms of presentation of reading results) of the reading programmes. However, due to the online teaching mode, DI was not fully reflected in aspects of affect (inadequate collaborative activities and interaction between students and teacher) and learning environment (lack of studying-together atmosphere). Results also indicate a widening gap between higher and lower achieving learners regarding engagement with ER activities due to differences in student autonomy and parental involvement.

Originality/value

This study provides some insights into differentiated instruction in ER implementation in the lockdown period, and proposes implications for L2 pedagogical design and innovation with a DI approach to ER.

KEYWORDS:

• Differentiated instruction
• extensive reading
• online learning
• l2 teaching
• secondary EFL teaching
• high achieving students

To accommodate student difference and optimise individual growth and development, differentiated instruction (DI) as a pedagogical philosophy and approach has been gaining recognition and changing classroom practice in the past two decades (Suprayogi, Valcke, and Godwin 2017). DI could be embodied in learning content, the process of learning, the assessment of learning outcomes, interaction and communication between learners and teacher, and the social and physical learning environment (Tomlinson and Moon 2013). In the actual implementation of DI, three central principles are learners’ readiness, interest, and learning profile which serve as the basis for teachers’ design and use of DI in classrooms (Tomlinson 2001). In recent years, an exponential number of studies have been carried out concerning DI, which could be broadly classified into two groups: the benefits of DI and difficulty of its implementation (in most cases due to inadequate teacher training/education).

As a type of individualised reading, extensive reading (ER) resonates with the essence of DI, and has been gaining popularity as an L2 pedagogical approach in the past two decades. Due to the evolving and expanding contexts of ER practice, the implementation frameworks are concurrently expanding and accommodating diversified teaching environments and learner needs (Day 2015; Waring and McLean 2015). Aligned with the expanded interpretations of ER, a DI approach to ER implementation may add more nuances and layers to this pedagogical practice, and therefore merits exploration and discussion. This study examines two ER programmes from the perspective of DI, followed by discussions of L2 pedagogical implications.

Literature review

Differentiated instruction

Students in contemporary classrooms are getting increasingly diverse and heterogeneous. Correspondingly, student diversity and inclusive education have been gaining attention of teachers and educators (Nusser and Gehrer 2020). Differentiated instruction (DI) is an ‘educational practice that provides learners with multiple ways to master the objectives’ (Fox and Hoffman 2011, 21). Different from whole-class instruction or one-size-fits-all instruction, DI respects and prizes student difference to achieve optimal learning outcomes and individual development by adopting varied teaching strategies and activities in accordance with the individual needs of students (Suprayogi, Valcke, and Godwin 2017).

In the implementation of DI, five components are worth teachers’ attention and thinking: content, process, product, affect, and learning environment (Tomlinson and Moon 2013). These five categories are underpinned by three principles–readiness, interest, and learning profile (Tomlinson 2001). First, the teaching content should be compelling and cater to different needs of learners (Tomlinson 2014). To be specific, the suitability of learning materials is dependent on students’ varied competence levels, personal interests, and learning styles and strategies. Second, the process of learning, including the enhancement of knowledge and skills, is based on students’ initial learning profiles, allowing for multiple learning strategies and studying paces. Third, the product of learning, closely related to summative assessment, also legitimises variety and alternatives regarding presentation modes of learning outcomes. Fourth, to meet the varied affective needs of students, communication, collaboration, and interaction between students and teachers are important in that these interpersonal processes contribute to empathetic and dynamic relationships and a supportive learning community. Fifth, the physical and social learning environment should be inclusive and learner-friendly to members of different preferences and behavioural patterns. To effectively implement DI, these five categories underpinned by the three principles require an integrated approach rather than being considered and implemented separately (Tomlinson 2014).

Set against the backcloth of increasing recognition and attention to DI, an exponential number of studies relating to DI implementation and its benefits have been conducted in various educational contexts in recent years. With regard to the efficacy of DI on different groups of learners, results diverge: In Dutch primary schools, evidence shows that high-ability and average groups of students benefit more than low-ability students from DI in terms of mathematical achievement (Faber, Glas, and Visscher 2018). In German elementary schools, a quantitative longitudinal study demonstrates that differentiated reading instruction (with treatment group) led to higher growth in reading fluency, especially for students with lower reading skills (Förster, Kawohl, and Souvignier 2018). In slight contrast with this finding, a study carried out in German secondary schools did not find evident effect of DI on student reading competence partly due to the difficulty teachers encountered in applying tiered assignments to various competence levels within class (Nusser and Gehrer 2020). The benefit of DI is also reflected in the moderating role it plays in the improvement of student autonomous motivation as a result of teacher structure mediated by student perceived competence, according to the research carried out in Canadian elementary schools (Guay, Roy, and Valois 2017). Similarly, a study conducted in Indonesian public schools evidences that DI promotes gifted students’ motivation and achievement in English learning (Kamarulzaman, Azman, and Zahidi 2017). Regarding factors that facilitate DI implementation, data collected from 200 English language teachers in the United Arab Emirates indicate that cooperative learning is perceived by teachers as conducive to magnifying the effect of DI (Ismail and Al Allaq 2019).

In parallel with the growing popularity of research into the benefits of DI, an array of studies have dug into the difficulty and challenges of DI implementation. A large proportion of research in this direction reports the mismatch between teachers’ awareness of DI and their actual teaching practice. A systematic review of 12 empirical studies carried out in secondary schools in different geographical contexts show that although DI demonstrated small or moderate effect on student achievement, knowledge gap remained for teachers to competently implement DI (Smale-Jacobse et al. 2019). Reasons for the mismatch include lack of related teacher training (e.g. for Indonesian primary school teachers) (Suprayogi and Valcke 2016), inadequate pre-teacher education (e.g. teacher education targeted at DI implementation among high-achieving secondary school students in Norway) (Brevik, Gunnulfsen, and Renzulli 2018), unsupportive institutional environment including the limitations of course content and lack of guidance on DI strategies (e.g. in primary and secondary schools of Saudi Arabia) (Aldossari 2018), and the discrepancy between contextual factors (or ‘institution decisions’) and personal beliefs (or ‘teacher decisions’) as reflected in 28 US-based studies (Bondie, Dahnke, and Zusho 2019). A specific example is that data collected from 492 primary school teachers in Ethiopia shows that DI is generally regarded as time-consuming and challenging (Merawi 2018). Similarly, teachers of multigrade classes of Maldives revealed problems such as lack of competence and time for DI implementation (Shareefa 2020). Promisingly, some teacher development programmes have collected evidence showing that DI-oriented training could not only positively influence teachers’ attitudes and practices, but also student achievement. For example, one programme was carried out in Cyprus which prepared teachers with knowledge and practice concerning DI in the form of seminars (Valiandes and Neophytou 2018); another example was a school project in Iran which provided teachers with several DI-related scaffolding sessions (Karimi and Nazari 2021).

DI was originated and developed in western countries, and relatively a less explored domain in Asian countries (Kamarulzaman, Azman, and Zahidi 2017). In Hongkong, DI has been highlighted in the curriculum reform, but factors such as large class size and the one-size-fits-all curriculum obstruct DI implementation in teaching practices (Wan 2017). In mainland China, rather than propagating DI, stratified teaching is widely applied to different levels of education as the main pedagogical approach to address student difference (Shi, Cheung, and Cheung 2020). Consequently, limited research concerning DI has been conducted in this context. This article aims to fill this gap by proving some insight into DI in English as a Foreign Language (EFL) teaching by investigating two extensive reading programmes carried out in a secondary school in Beijing during the COVID-19 pandemic.

Extensive reading and its variations

Aligned with the principles of student-centredness and differentiation of instruction, extensive reading (ER) is an individualised form of reading (Eisenmann 2013). Specifically, ER is ‘real-world reading but for a pedagogical purpose’ (Day and Bamford 1998, 7), with a relatively fast speed and in large quantity (Nation and Waring 2019). The ten principles of ER created by the two seminal authors in this field (Day and Bamford 1998) have been regarded as the basic tenets and benchmark for ER implementation in the past two decades:

1. The reading material is easy.

2. A variety of reading material on a wide range of topics is available.

3. Learners choose what they want to read.

4. Learners read as much as possible.

5. The purpose of reading is usually related to pleasure, information, and general understanding.

6. Reading is its own reward.

7. Reading speed is usually faster rather than slower.

8. Reading is individual and silent.

9. Teachers orient and guide their students.

10. The teacher is a role model of a reader. (Day 2002, 137–141)

These ten principles could be summarised as follows: easy materials of various topics and genres are available; learners select their own reading; the purpose of reading is for pleasure or general understanding; teachers orient and guide student reading, and act as a role model of a reader. Whilst the ten principles establish the basis for ER implementation, teachers’ practices vary widely in response to contextual features. Based on a comprehensive review of 44 ER-related empirical articles, Day (2015) envisaged and extended ER implementation into three directions: supervised ER, independent ER, and blended extensive and intensive reading. Meanwhile, Waring and McLean (2015) proposed several formats for ER implementation: classical ER; class reading; integrative ER; ER as literature; ‘easy ER’; ‘ER as i+1’ (161). It is evident that the expanding frameworks for ER implementation manifest the increased emphasis on student difference and diversity of teaching contexts. In this sense, ER and DI share some principles: diversified reading materials (content); different ways of engaging with materials (process); varied manners of assessing reading (product); individual reading facilitated by collective activities and teacher scaffolding (affect); class/school library and other resources (learning environment) (Tomlinson and Moon 2013). This said, it would be interesting and enlightening to examine ER implementation from the perspective of DI, which is one objective of this study.

As a variation of ER, ER as literature has drawn considerable interest and attention from the L2 teaching community, although it still remains a contentious topic and a challenging task. One side of the debate is that literature is an inexhaustible reservoir of lexical, syntactic, and semantical features, presented in rich and nuanced textual contexts, and therefore could be beneficially used in L2 classrooms (Nation 2017). The other side of the debate focuses more on the difficulty of original literary works. For example, Macalister and Webb (2019) hold that children’s literature is too difficult for intermediate L2 readers because approximately 8000 word families are needed to achieve the aim of 98% vocabulary coverage. Notwithstanding the ongoing debate, there has been a phenomenal growth in research and practice of using literature in L2 classrooms (e.g. Bloemert et al. 2019; Saka 2018; Sun 2020; Viana and Zyngier 2020). Taking into account that literature teaching often involves collective reading and unified materials, therefore posing challenges for DI, it is worth exploring the feasibility of implementing DI in L2 literature teaching. The present study explores two ER programmes which used literature as the main ER materials, thus may provide some insight into the question raised above.

Research questions

To gain an insight into how DI is embodied in two EFL extensive reading programmes carried out in a secondary school in Beijing during the COVID-19 pandemic, the research aims to answer the following two questions:

1. How is DI reflected in two EFL extensive reading programmes during the pandemic?

2. How do the two online ER programmes inform L2 pedagogy in relation to DI and ER implementation?

Methods

Research context

The two reading programmes that the current study investigates were conducted in a key middle school in Beijing during the COVID-19 pandemic from January to July 2020. During this period, teachers delivered synchronous lessons online and set daily assignments for students. Some WeChat Apps were used for teachers to check students’ assignments and give feedback. Except for two weeks in June when the lockdown was eased and students returned to school temporarily, for the rest time of the semester, students studied at home. In the participating school, stratified teaching has been implemented and evolving for many years. As a teaching mode which recognises student difference and competence variation, stratified teaching hierarchises teaching objectives, materials, and methods in accordance with student aptitude and learning needs (Shi, Cheung, and Cheung 2020). In different grades of the school under discussion, varying types of stratified teaching were implemented. For example, in Junior One to Junior Three, two main subjects–math and English–involved stratified teaching, while in Senior One to Senior Three, teaching stratification was mainly embodied in experimental classes (comprised of students good at science subjects) and regular classes. In this school, ER had been integrated into EFL teaching for more than ten years, mainly in the forms of ER as literature and magazine reading.

Participants

This study was part of the researcher's PhD thesis project exploring ten extensive reading programmes carried out in Chinese secondary schools. From these ten cases, two were selected for this study for twofold reasons. First, these two programmes were from the same middle school in Beijing (one of the top schools whose students are more competent and competitive than those of other public schools), therefore bearing some commonality such as students’ sociocultural backgrounds and teacher training trajectories. Second, the two programmes lasted for a whole semester, unlike other programmes existing for a (much) shorter time partly due to the pandemic. The two teacher participants, Xia and Yun (pseudonyms) were former colleagues of the researcher, and had 12 and 21 years of EFL teaching experience respectively. During the data collection period, Xia was teaching two Junior One Level A classes (students aged 12–13). In this grade, English classes were divided into Level A (mostly upper-intermediate English learners, equivalent to CEFR B2) and Level R (intermediate and lower-intermediate English learners, appropriately CEFR B1 and A2 levels) based on students’ scores in the English placement exam at the beginning of the semester. Yun was teaching two Senior Two classes (students aged 16–17): one experimental class (students have higher academic competence especially in science subjects; their English is approximately CEFR B2 level) and one regular class (students have average academic competence, whose English is CEFR B1 level on average). Voluntarily, 74 students of Xia and 27 students of Yun completed the questionnaire and three students from each teacher were interviewed at the end of the semester.

Data collection

Data were collected from four sources: questionnaires (students), interviews (teachers and students), reflective journal (teachers), and relevant documents including teaching plans, reading materials, students’ works etc. (teachers and students). During the data collection period, at the end of each month, teacher participants submitted their reflective journal entries to the researcher, together with documents in various forms related to the reading programme. In the reflective journal, teachers could record anything (e.g. teaching activities, difficulties, confusions, reflections etc.) concerning their ER implementation. At the end of the semester, individual interviews were conducted with each teacher participant, and focus group interviews or individual interviews (depending on interviewees’ preferences) were carried out with voluntary student participants (six in total; three from each programme). Interview questions focused on students’ reading experiences (e.g. reading materials, reasons for the reading, reading frequency and amount) in the lockdown period and their perceptions of the difference between normal times and the special circumstances in relation to ER. Meanwhile, an electronic survey was conducted with student participants of the two teachers (see Table 1).

Table 1. Data collection schedule and participant numbers.

Instrument	Time	Participant
Reflective journal	The end of each month from Jan to Jul 2020	2 teachers
Documents	The end of each month from Jan to Jul 2020	2 teachers; related students
Interview	Jul 2020	2 teachers; 6 students
Questionnaire	Jul 2020	74 (Xia) + 27 (Yun) = 101 students

Data analyses

Journal entries, interview transcripts, and related documents were analysed with the approach of thematic analysis. Following the recommended six steps (Braun and Clarke 2006), the researcher first familiarised herself with the data by transcribing the interviews and reading the data repeatedly. Second, the researcher conducted preliminary analysis by highlighting the important points which led to the initial codes (including App, online sources, narrow reading, literature reading, difference/variety, negative effect, positive effect, sharing, assessment, supervision etc.). Third, based on the initial codes, the researcher searched for emergent themes (including DI in EFL reading; advantages/disadvantages of online learning; variety of reading materials; collaborative activities; online supervision etc.). Fourth, the researcher reviewed emergent themes with reference to research questions, combining some overlapping themes and identifying inherent logic between themes. Fifth, the final themes were generated in relation to the conceptual frameworks for differentiated instruction (Tomlinson 2001; Tomlinson and Moon 2013). Sixth, themes were presented in the Findings section with in-depth description and full citations extracted from the data.

Quantitative data collected from the questionnaire were analysed with Excel and presented in the form of percentage, while free-text data generated by the open questions were analysed with thematic analysis in the same manner as qualitative data from other sources (Dörnyei and Taguchi 2010).

To enhance the trustworthiness of this case study, strategies such as triangulation (collecting data from multiple sources), member checking (interview transcripts and draft of Findings were sent to teacher participants via email for review), and thick description (by presenting detailed introduction to the research context and participants and sufficient direct quotations from participants) are applied to the current study (Creswell and Poth 2018).

Findings

The presentation of the findings will follow the five-component framework for DI: content, process, product, affect, and learning environment (Tomlinson and Moon 2013), underpinned by the three dimensions: readiness, interest, and learning profile (Tomlinson 2001). The reason for the adoption of this set of frameworks is the holistic and comprehensive perspective it presents about DI. Considering that the two reading programmes provide different aspects concerning DI, a cross case analysis was not feasible. Therefore, the complete data set was treated as an integral case – DI reflected in ER programmes carried out in COVID-19 pandemic.

Content

Data from these two reading programmes indicate that differentiation in terms of readiness, interest, and learning profile was reflected in the selection and provision of reading materials (content). First, to ensure students of different language levels had respectively suitable materials to read, Xia devised a three-tiered system: vodka (the most difficult), wine (averagely difficult), and green tea (the easiest) (Xia, Journal 1). During the weekends, Xia provided students with materials of three levels with indicative labels for students to choose from. Topics of the materials were usually related to contents of the English textbook. For example, when Shakespeare appeared in the textbook, Xia selected some Shakespeare’s sonnets, introduction to his hometown, his literary works, and his famous sayings as reading materials provided to students tagged with vodka, wine, and green tea. Furthermore, to cater for students’ various reading interests, Xia's recommendation covered a large array of literary genres, including novels, poems, speeches, stories, and prose. As she explained, ‘exposing students to a variety of genres is key to maintaining their interest in learning English’ (Xia, Interview).

Questionnaire results present an overall view of the English materials students read during the lockdown period (see Table 2). Interestingly, these two reading programmes bear great similarity in material selection. A vast majority of students read original novels because it was part of their English assignments (in Xia's case) or the teacher strongly recommended it (in Yun's case), followed by English magazines and online articles. Some highly motivated students read far more than what teachers required. One student of Xia in the interview revealed that in addition to novels, she also read non-fiction materials, including encyclopaedias, online newspapers and magazines (e.g. The Guardian, The New Yorker, and Fox News). When asked how she learned about these sources, she answered ‘my parents’ (Xia's S2, Interview). In comparison, without guidance or supervision from parents, another student (Xia's S1) only occasionally read some online articles.

Table 2. Genres of materials students read during the lockdown period.

Genre of material	Student percentage (Xia's programme, N = 74)	Student percentage (Yun's programme, N = 27)
Original English novels	99	79
English magazines	41	29
Online English articles	35	24
English newspapers	23	14
Graded readers	19	21
English graphic novels	8	7
English picture books	3	17
Other	3	14

Regarding novel reading, in addition to the novel the whole class read together (Wonder by R. J. Palacio), Xia recommended some novel series to students, including Percy Jackson by Rick Riordan, and The Hunger Games by Suzanne Collins. At the end of the semester, Flora learned that some students had finished reading the Hunger Games trilogy, and some had read many books of Percy Jackson. In the interview, one student expressed her strong liking for Percy Jackson series: ‘I finished the five books very quickly … I started reading this series during the winter holiday and got addicted to it.’ (Xia’ S2, Interview) Another student demonstrated her passion for The Hunger Games series: ‘After I started reading the books, I found it quite interesting, so I kept reading, one after another.’ (Xia's S3, Interview)

Compared with Xia who laid great emphasis on diversified language difficulty and literary genres, Yun showed her consideration for the diversity of reading tools. ‘Students could read paper books, or electronic books on kindle or iPad. They could also use Apps such as “WeChat reading”.’ (Yun, Journal 2) Yun explained that electronic reading had the advantage of convenience in looking up new words. Corroborating Yun's preference for using electronic reading tools, two out of three interviewed students from her class mentioned using Apps to look for English articles (Yun's S3) or recommendations on English novels (Yun's S1).

Process and product

Seeing reading as a process, the two teachers provided students with ample room for individualised decisions on reading frequency and quantity, as the survey results demonstrate (see Table 3). Across the two programmes, more than half of the students read English materials several times a week, and each time the largest groups of students read less than ten pages of a book or an equivalent amount in other forms.

Table 3. Students’ frequency and quantity of EFL reading in the lockdown period.

Reading frequency	Xia's programme (%)	Yun's programme (%)	Reading amount each time	Xia's programme (%)	Yun's programme (%)
Several times a week	72	57	≤10 pages	47	43
Every day	20	21	10–20 pages	31	29
Once a week	5	7	≥20 pages	15	29
Less often than once a week	3	14

In these two programmes, different methods were adopted to present students’ product of reading. In Xia's programme, presentation and sharing were the main forms of product demonstration, including analysis of characters (see Appendix A), poem reading, approach sharing etc. Compared with Xia's diversified forms of presentation (designed for Junior One students), Yun's method (targeted to Senior Two students) was simple: online submission of reading feedback (see Appendix B). Yun provided details in her journal: ‘We use the WeChat App Class Housekeeper: students report what they read each day with the name of the book, number of pages they read, and notes (extracts, summary, or feedback).’ (Yun's Journal 1) Yun added in the interview that writing notes was not compulsory because she did not want to demotivate those students who dislike writing after reading. This Class Housekeeper activity was part of the formative assessment. Students who actively took part in it could get at most 10 credits in the overall assessment of the semester. Students’ response to this method of assessment differed. One student revealed that this activity to some degree pushed him to persist in reading because they could learn about others’ reading progress and get influenced by it (Yun's S2). Another student contended that this online checking system only worked for those who had the intention of reading, while for those with little interest in reading, this made almost no difference (Yun's S1).

Affect and learning environment

As shown in Table 4, the majority of students in both reading programmes perceive that they had more time reading English materials during the lockdown period compared with normal times. In addition, more students regard it easier to do EFL reading in the lockdown period than the normal time, providing the following reasons in the survey: First, it is more convenient to look up new words online. Second, it is easier to search for background information about the reading. Third, more time is available for the reading due to reasons such as fewer lessons, no commuting, and the cancellation of some extracurricular courses. In the interview, Xia referred to student having more time for reading as an advantage of the online learning mode, adding that ‘at normal times, students have little time left for extensive reading after finishing assignments and preparing for various tests’ (Xia, Interview).

Table 4. Comparison between lockdown period and normal times regarding L2 ER.

Learning context	Time for reading		Difficulty of reading
	Xia's programme (N = 74)	Yun's programme (N = 27)	Xia's programme (N = 74)	Yun's programme (N = 27)
Lockdown period	More (63%)	More (57%)	Easier (37%)	Easier (29%)
Normal times	More (5%)	More (29%)	Easier (12%)	Easier (14%)

With regard to the disadvantages of online learning in relation to extensive reading, the biggest issue was lack of a favourable reading environment. According to the survey results, 59% and 36% students from the two programmes hoped teachers could organise some online discussion or other collaborative activities concerning ER. A student gave this response: ‘One negative influence (of online learning) is the lack of studying-together atmosphere’ (an answer from the questionnaire). Similar responses include: ‘If possible, I really expect activities about novel reading’; ‘Teacher could organise an online reading club’ (answers from the questionnaire). In the interview, one student (Xia's S3) expressed the same expectation in detail:

I like the idea of group discussion, but the lockdown made it impossible. I also like teachers’ guidance, because opinions of peers may not be as profound as teachers’. After the group discussion, if the teacher could share her ideas, that would be better. (Xia's S3, Interview)

From teachers’ perspective, Xia pointed out another problem of online learning:

Online learning widened the gap between different levels of students. The most distinctive students, A+ students, kept reading what teachers recommended, while the less competent students could not allocate much time for English study due to the lack of self-regulation and self-learning ability, so they read much less than the top learners. (Xia, Interview)

Another reason for the enlarged gap between different levels of students was the disparity in parents’ supervision. One student (Xia's S2) in the interview revealed that she read English materials for 30–40 minutes before going to bed every day, but sometimes needed parents’ reminding. Aligned with this opinion, Yun emphasised the importance of teachers’ supervision and linked it to a downside of online learning: ‘Face-to-face instruction enables in-class tests and other forms of supervision, which is essential for students’ sustained reading.’ (Yun, Interview) Slightly different from Yun's viewpoint, one of her students (Yun's S3) revealed in the interview that the online checking system worked for her, without which she could not persist in the reading.

Discussion and conclusion

DI & ER

The present study provides evidence showing the reciprocal relationship between DI and ER, which could be elaborated as follows. First, recommending or providing students with a variety of reading materials taking into account their varied language competence and interests contributes to students’ reading motivation (Day and Bamford 1998; Nation and Waring 2019). In the current study, one teacher devised the three-tiered system (vodka, wine, and green tea) to classify reading materials with indication of difficulty levels, and students could freely choose one according to their language proficiency. Another demonstration of teachers’ awareness of DI was the flexibility of reading frequency and quantity without giving restricted requirements. Second, introducing students to multimodal texts and electronic readers accommodates learners’ differing learning profiles and therefore makes reading accessible and compelling to students of different reading preferences. In this study, teachers recommended various types of reading materials (including novels, graded readers, picture books, graphic novels etc.), literary genres (e.g. short story, poem, prose etc.), and reading tools (e.g. kindle, iPad, WeChat Apps etc.) to allow for different learning styles and varied learning strategies. Correspondingly, data from questionnaires and interviews display a rich variety of reading choices that students made in terms of reading tools and genres of discourse. Third, multiple forms of assessment feed forward and motivate students to persist in extensive reading. Two teachers in the study adopted different assessment measures: one involved presentation of reading product; the other used a mobile App–Class Management. The common feature of these two methods is the sharing element, which led to students’ enhanced motivation for reading as revealed by some student participants in the interview.

Another important factor for implementing DI is the affect elements which are channelled and displayed through interaction and collaboration between peers and teacher. In relation to reading, as Waxler and Hall (2011) said, ‘the conversation about the narrative is at least as important, if not more so, than the reading of the text itself’ (52). The values of communication and collaboration about reading are manifested in the following aspects. First, group discussion could enhance L2 comprehension, especially when the discussion is held in L1 (Turnbull and Evans 2017). Second, confirming with peers about the understanding of the text improves learners’ confidence and motivation for L2 reading (Ramonda 2020). In the questionnaire and interviews of the present study, a large percentage of students expressed their desire for collaborative activities (online or face to face) with peers surrounding the reading. However, teachers seemed not to be fully aware of students’ such expectations and provided limited chances online for students to exchange ideas about their reading.

Regarding the topic of L2 literature teaching, this study adds to the existing literature with evidence that narrow reading, derived from individual interests, is an effective approach to enhancing students’ reading motivation. Narrow reading refers to reading books written by a single author or concerning a single topic that the reader shows interest in (Krashen 2004). In addition to the benefits of improving reading speed and comprehension (Chang and Millett 2017), the present study evidences that narrow reading could also generate considerable enthusiasm among upper-intermediate L2 learners for reading series of novels written by the same author.

Pedagogical implications and challenges

This study contributes to the ongoing discussion about DI in the following aspects. Corroborating the finding that DI contributes to students’ autonomous motivation (Guay, Roy, and Valois 2017), the current study collects evidence that differentiating reading tasks with different levels of materials could enhance student autonomy (especially those with higher language competence) with evidence that high-performing students read much more than other groups of students due to their overall learning abilities and (in some cases) parents’ supervision and guidance. The difference between higher and lower achieving learners is also reflected in the text types students were engaged with. To be specific, Xia's students (aged 12–13) were four years younger than Yun's students (aged 16–17), but more proficient in English compared with same-age peers. Results show that more percentage of Xia's students (99%) than Yun's students (79%) read original English novels, while quite interestingly, a higher percentage of Yun's students (17%) than Xia's students (3%) read English picture books. This finding coupled with teachers’ perceptions of the widening gap between higher and lower achieving students complements earlier results that high-ability or gifted students ‘profited’ more than lower-ability students from DI (Faber, Glas, and Visscher 2018; Kamarulzaman, Azman, and Zahidi 2017). Thus, it is worth further exploration and research how to provide supervision and support to learners of lower proficiency levels, especially those from lower socioeconomical backgrounds with less parental care and involvement.

This study also provides some implication for teacher training on DI implementation, specifically relating to pedagogical knowledge targeted at high-achieving secondary students (Brevik, Gunnulfsen, and Renzulli 2018). Responding to the study carried out in German secondary schools showing that students’ reading competence was not significantly improved due to the difficulty of setting tiered assignments (Nusser and Gehrer 2020), this study provides some feasible solutions as presented above (e.g. the three-tiered reading material system).

Due to the fact that the current research was carried out in the pandemic lockdown period (including two weeks of face-to-face instruction) and teaching mostly took place online, results of the study also contribute to literature concerning online learning. First, more students consider reading in lockdown period was easier than normal times due to the convenience of looking up new words and background information about the reading. Meanwhile, the current study reveals some problems with online learning. Corroborating the notion that online learning is heavily dependent on students’ self-management (Rasheed, Kamsin, and Abdullah 2020), this study finds that some secondary students fell behind in reading without adequate self-regulation or parents’ supervision in the online learning mode. Another disadvantage of online leaning reported by student participants of this study is the lack of atmosphere of ‘learning together’. This is a problem worth considering because the ‘togetherness’ or collaborative learning could facilitate DI as previous research shows (Ismail and Al Allaq 2019) and improve students’ reading motivation as the current study suggests. Due to the nature of the participating school, students are mostly high achieving students from relatively well-off family backgrounds. Therefore, limited data was collected concerning lower achieving students in relation to the discussed topics. Further research may delve into this aspect.

Disclosure statement

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

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.

Notes on contributors

Xiaomei Sun

Xiaomei Sun is a PhD candidate at the University of Edinburgh, where she also teaches research methods and education policy courses. Her research interests include extensive reading in L2, children's literature, teacher training and development, and L2 pedagogy. Her recent research has appeared in TESOL Journal, The Language Learning Journal, The Reading Matrix, and ELTED Journal.

Appendices

Appendix A. Two students’ online sharing of character analysis of the novel Wonder

Student A:

Student B:

Appendix B. Students’ reading feedback on App Class Management