Effectiveness of Implementing Differentiated Instruction in the English for Specific Purposes Classroom in Hong Kong

The academic benefits of differentiated instruction are supported in western culture, although to date the majority of research conducted has concentrated in the Middle East and in Greece, and in primary and secondary instruction. Despite its relative novelty in Hong Kong, and the fact that it is under-researched regarding its effectiveness, a pilot study has been conducted at a tertiary institution in Hong Kong to explore implementing differentiated instruction in the ESL/EFL/ESP Classroom for learners from different educational backgrounds. To examine and determine its effectiveness, all Year 2 Nursing and Social Work students (N = 319) were included in the research study, and the students’ English skills were assessed by classifying them by a screening test into 3 ability groups, for each of whom three tiered tasks were assigned for completion at three different points of the semester. At the end of the semester, the findings were analysed with t-tests and a post-course student survey to determine the change in comprehension skills, learning motivation, classwork preferences, and study approaches. The relationship between each factor and English skill was assessed using correlation coefficients. The study has demonstrated that adopting the differentiated instruction approach in English and communication courses enhances students’ English ability in tertiary education in Hong Kong. In particular, substantial improvement is observed in both “lower ability” students and “mid-level ability” students, notably in grammar, syntax and inference-making skills, while the student survey responses reveal students’ learning preferences, as well as their cognitive and psychological elements to warrant further research.


Background
In recent years to address this issue of ever-growing classroom diversity in English language teaching, both teachers and researchers in the field worldwide have been calling for a change of approach and strategy (Tomlinson & Imbeau, 2010). Today, students are diverse and come in all categories in terms of background, socio-economic status, readiness for instruction, culture, learning profile, interest, familiarity with technology that aids and supports teaching and learning, and so on (Lawrence-Brown, 2004;Sharma, 2006;Tomlinson, 1999). Various researchers and practitioners in the field have been calling explicitly for the introduction of differentiated instruction as an effective approach to attend to the learning needs of different students in the classroom with high heterogeneity, so that students with diverse needs will be able to attain a good proficiency level (Tomlinson, 1999;McQuarrie & McRae, 2008).

Rationale for our research study
As a result of inclusive education in Hong Kong, as well as worldwide, today's English language classroom everywhere reveals an ever-increasing student diversity, so much so that the traditional 'one-size-fits-all' approach in teaching and learning is no longer tenable (Tomlinson, 2017;Bondie et al., 2019). Students come to a class bringing with them their diverse background, learning styles, motivation, attitudes, interest, abilities and family support. The diversity of students in a mixed-ability classroom can result in a significant challenge for the teacher in meeting the learning needs of all students. To address this problem, differentiated instruction (DI) has been considered as an approach as well as a philosophy to cater for learner diversity (Tomlinson, 2002(Tomlinson, , 2015(Tomlinson, and 2017Tomlinson & Imbeau, 2010;Tomlinson & Allan, 2000) and also a possible way to enhance teaching and learning (Watts-Taffe et al., 2012). It is the goal of differentiated instruction (DI) to reach out to each student and deliver the lesson in a way that fits their readiness, interest and learning profile.
Differentiated instruction is based on a strong theoretical foundation which includes constructivist theories, brain-based research and multiple intelligences (Felder & Soloman, 2004;Gardner, 1983;Vygotsky, 1978). To date, many of the studies are qualitative in nature, indicating positive affective outcomes in terms of motivation, task commitment, and attitudes about learning (Burkett, 2013;Maeng, 2011), and past studies undertaken to investigate the effectiveness of differentiated instruction focused on such subject areas as Mathematics (Tieso, 2005), Reading (Marulanda et al., 2006) and Science (Dooley, 2009), principally in L1 (first/native language) context and seldom applied to second language acquisition (SLA).
However, extending the research focus to college students is important, because the growing diversity within today's classroom underscores the necessity for a "pedagogical shift" (Swaffield & Guske, 2011) from a teacher-centred, 'one-size-fits-all' method to a learner-centred differentiation of instruction, at all levels in the entire education system. Such a pedagogical shift has motivated this research team to conduct the present study.

Differentiating the instruction or the curriculum
Taking reference of the contention that the 'one-size-fits-all' approach is no longer effective in today's ESL Classroom (Bondie et al., 2019;Magableh & Abdullah, 2020), the point of departure for the present study is therefore to examine, in a Hong Kong post-secondary institution how English teachers attempt to innovate in their pedagogical practice by implementing differentiated instruction in the local ESL context. To date our study has shed some light on how differentiated instruction may be integrated into the Hong Kong ESL Classroom to improve student learning in similar L2/ESL contexts, contributing to the existing research base for differentiated instruction for ESL in particular. Indeed, since merely differentiating the instruction on any single English language course, albeit as a starting point, may not be too impactful, and necessarily limited in both scope and dimension, therefore, this study proving effective, the next step forward should be and would be differentiating the curriculum for ESL teaching and learning at tertiary level in Hong Kong.

The teacher's perspective
Extant research has also shown that differentiation of instruction is beneficial not only to students but also to teachers. Affholder (2003) reported that teachers in the study employing higher levels of differentiated instruction strategies showed elevated self-efficacy and readiness to assume greater responsibility for student growth and learning. Another study by McAdamis (2001) reported that teachers in the study initially resisted changing their teaching practice; however, the adoption of strategies like peer coaching, action research and staff development workshops offered them on-going support and feedback. This resulted in the teachers becoming eventually convinced of the benefits of differentiation, as they witnessed improvements in their students' learning and motivation. They were then eager to continue adopting differentiated instruction on a regular basis.
Nevertheless, despite the positive impacts of differentiated instruction on teachers' development of teaching practice as shown in the above studies, a number of challenges are faced by college instructors in implementing differentiated instruction, such as large class size, time constraints and limited contact hours with students (Ernst & Ernst, 2005). In Hong Kong, pre-service and in-service teachers perceived some potential in implementing DI in primary and secondary school teaching, however, time management, class management in large classes, and limited training and support are also potential drawbacks for the implementation (Wan, 2015;Wan, 2017). Furthermore, as shown in related literature, while English teachers overseas have long been enacting some of the strategies of differentiated instruction such as collaborative teaching, grouping, tiered activities, tiered assignments in their teaching practice, there is as yet no evidence that such strategies have been widely adopted in Hong Kong schools or colleges.

The research question
In view of the above considerations, one of the objectives of the present study is therefore to explore how local English college teachers go about implementing differentiated instruction as a pedagogical approach to cater for learner diversity through collaborating with the researchers. This study, an action research in its nature and indeed research by teachers for teachers, will thus pedagogically contribute specifically to the professional development of English teachers both as teachers and as researchers, as well as to the future development in implementing differentiation instruction in the Hong Kong ESL setting at college level. The findings generated from this study will provide preliminary data for translating research-based differentiated instruction into workable, school-based interventions and, eventually, into ESL classroom practice (Paris & Paris, 2001;Pressely & El-Dinary, 1997).
This study has investigated the development and implementation of differentiated instruction for ESL teaching and learning in a self-financing post-secondary institution in Hong Kong, where the conventional 'one-size-fits-all' pedagogical approach has been adopted in English teaching and learning for many years. It has also explored the effect of differentiated instruction on college students' English proficiency and motivation. Mixed methods have been used in this case study, involving multiple sources of data, including questionnaires, teacher and student interviews, pre-and post-tests, and lesson observations. The present study therefore aims to fill the research gaps identified earlier and answer the following research questions:

How do college teachers develop and implement differentiated instruction in English
teaching and learning? 2. What is the effect of differentiated instruction on college students' English proficiency and motivation?

Participating teacher and students
All Nursing and Social Work students enrolled in Year 2 ESP courses at a higher education in Semester 2 of 2021-2022 were included in this study. Three English teachers who had attended a series of training sessions in differentiated instruction participated in the research study.

The procedure
The following presents the procedures and activities designed to implement the Project Programme in the participating Bachelor of Nursing (BN) and Bachelor of Social Sciences in Social Work (SW) classes in Semester 2 of the 2021-2022 academic year, in chronological order: 4. Weeks 3 and 4: Conducting activities relating to the first set of tiered tasks 2 as well as the first class observation 5. Weeks 7 and 8: Conducting activities relating to the second set of tiered tasks as well as the second class observation 6. Weeks 11 and 12: Conducting activities relating to the third and last set of tiered tasks as well as the third and last class observation 7.
Week 13: Administering the post-study proficiency test 8.
Week 13: Administering the feedback questionnaire 9. Interviewing participating teachers 10. Interviewing a select sample of participating students 11. Administering the feedback questionnaire to the participating teachers

Measurement and Tiered Tasks
It is to be noted that the pre-test and the post-test were modelled on the EU B2 Track Test, with well-defined parameters for scores, indicating specifically the areas for remedial work, if any; that students' learning approaches were assessed with the Approaches and Study Skills Inventory for Students (ASSIST).The effectiveness of the differentiated instruction activities has been critically measured. Furthermore, all the teachers on all the sections of the LENG 203 and LENG 204 courses participated through the entire process, from the preparation stage in Semester 1 throughout Semester 2 and in the weeks following the end of Semester 2. This ensured that the Project Objectives have been reliably examined for achievement.
The short version of "Approaches and Study Skills Inventory for Students" (ASSIST-S; Entwistle et al., 2013) was employed to measure the depth of students' learning approach, also on account of its high reliability (Cronbach's Alpha = range between .75 and.83) and high validity (KMO = .769). The inventory contains 18 items for assessing students' preference for deep, strategic or surface approach, on a 5-point Likert scale (5 = Agree, 1 = Disagree).
Participants were asked to rate on a 5-point Likert scale (1 = Disagree; 5 = Strongly Agree) for all the 33 items. The cut-off scores of test anxiety, communication comprehension and fear of evaluation were 45, 33 and 21 respectively. Test takers with a total score of 98 or lower were considered "low"; 99 was for indicating "moderate" and 100 or above was an indication of "high" in anxiety.
The survey posed ten questions to learn about students' learning motivation; self-perceived English ability; perception on educational and affective factors; classwork preferences (e.g., individual work or group work); among others. Responses were collected using 4-point or 5-point Likert scales and checkboxes. (NB: The screening survey and post-study survey were identical. Survey completion was voluntary, and self-identification optional.)

Brief description of the Tiered Tasks
Tiered Task 1 aims to strengthen students' skills in vocabulary building. Medical terms with Greek word roots, prefixes, and suffixes are introduced to Nursing students; Latin word roots, prefixes and suffixes are introduced to Social work students. Group activities are changed to individual activities for Nursing students. Nursing students are asked to complete one reading task and three in-class activities while Social work students are split into two groups to complete one reading task and one group activity in class. Out of the three in-class activities, Activity 3 is challenging and students with high ability are expected to score high while the students with low and averaged ability are expected to score lower. Class observation is made for Tiered Task 1.
Tiered Task 2 aims to facilitate peer learning in inference-making. Students are encouraged to discuss with peers to complete a comprehension exercise. Group activities are changed to individual activities for Nursing students. Social Work students are split into four groups for an assigned reading task. Each group discusses, presents, and comments on one other group's presentation. Class observation is made for Tiered Task 2 for Social Work and students' involvement in discussion is observed and recorded for analysis.
Tiered Task 3 aims to facilitate peer learning in reading strategies. Students are encouraged to discuss with peers to complete a comprehension exercise. Group activities are changed to individual activities for Nursing students. Social Work students are put into four groups; free riders are put into two groups to encourage active involvement. Class observation is made for Tiered Task 2 for Social Work and students' involvement in discussion is observed and recorded for analysis.

Results
This research study has revealed not only the overall effectiveness of differentiating the instruction on the English for Specific Purposes (ESP) Classroom in our context, but fortuitously also both the students' affective and other educational dimensions and the participating teachers' positive feedback as well as the pressure that they felt contributing to the process of differentiation. The following presents a detailed report on all these aspects for discussion:

Submission rate
A total of 310 responses were collected on the screening assessment, with a submission rate of 99% from Nursing students and 81.40% from Social Work students, while a total of 319 responses were collected on cognitive assessment, with a submission rate of 99.33% from Nursing students and 75.86% from Social Work students. Baseline measures (Screening assessment & survey) were analysed and compared between the experimental and control groups. (Data from Social Work tutorial session A was excluded from the analysis because no post-assessment was submitted.) The submission rates of Tiered Task  A total of 282 responses were collected on Post-study assessment, with a submission rate of 94.31% from Nursing students and a submission rate of 65.52% from Social Work students. A total of 155 responses were collected on cognitive assessment, with a submission rate of 45.15% from Nursing students and 68.97% from Social Work students. All these data were analysed and a comparison was attempted between the experimental and control groups.

Groupings and performance of each ability group
Students were organized into three ability groups to determine if the tiered tasks would make different impacts on students with different English abilities.
The three ability groupings were carried out by the mean score rather than by the standard EU-B2Track Test passing score, because the mean score is 16.44, which is 1.56 below the passing score of EU-B2Track Test and the standard deviation is 2.36. As such, students with a score below 14 were grouped as low ability, students with a score above 20 were grouped as high ability, and students who scored between 14 and 20 were indicated as the average ability group.
ANOVA: Single Factor suggests that students' performance changed between the screening and post-study assessments and between each of the tiered tasks. The result suggested that the performance of both low ability [(F (2, 18) = 6.06, ρ<0.01)] and Averaged group [(F (2, 18) = 3.88, ρ<0.05)] differed significantly between the tiered tasks.
A strong positive correlation was found in both low and average ability group between the screening score and that for Activity 3 of Tiered Task 1 (r = .76 & r = .78). Moderately negative correlations were found in the high ability group between the screening score and that for Activity 3 of Tiered Task 1 (r = -.65). Likewise, the same phenomenon transpired the low ability group, indicating a negative correlation between the screening score and the total of Activity 1 and Activity 2 of Tiered Task 1 (r = -.79). A moderate negative correlation was revealed in the high ability group between the screening score and the total for Activity 1 and Activity 2 of Tiered Task 1 (r = -.58), and a strong negative correlation is observed in the low ability group between the screening score and that for Tiered Task 3 (r = -.74), Tiered Task 2 (r = .71).

English proficiency
The Nursing students performed better than the Social Work students in the post-study assessment. The class average score of all 10 Nursing tutorial classes was 17.16 and the class average score of the Social Work class was 15.10 in the screening assessment, while the Nursing students scored on average 19.99 in post-study assessment and the Social Work students scored an average of 16.42.
The class averages between experimental and control group were significant in the screening but not in the post-study assessment. Improvements were observed in both the Nursing and Social Work students in English ability, with the significant differences revealed in the pre-course screening (M = 16.97, SD = 3.47) and the post-study test (M = 19.49, SD = 2.43), t(10) = 9.60, ρ <0.001. In the low ability groups of students, significant improvements were observed in overall score [t(10) = 4.67, ρ <0.001], Grammar [t(11) = 2.58, ρ <0.05] and Inference-making [t(10) = 5.47, ρ<0.001] (as shown in Figure 2). In the averaged ability groups, significant improvements were observed in overall score [t(10) = 11.20, ρ <0.001], Grammar [t(10) = 3.06, ρ =0.01] (as shown in Figure 2), Inference-making [t(10) = 5.69, ρ <0.001] (as shown in Figure 3). No significant change was observed in the high ability groups, which may warrant some pondering and should provide "food for the next research study".

Figure 2: Comparison of Pre-and Post English Assessment on Grammar Skills
Comparison is made between the three ability groups. Significant improvements is observed in low and average ability groups.

Figure 3: Comparison of Pre-and Post English Assessment on Inference-making Skills
Comparison is made between the three ability groups. Significant improvements is observed in low and average ability group.

Study approach
A weak correlation was found between students' deep learning approach and English ability (r = .18) in the pre-study screening assessment scores. An increase of 20.18% and 3.95% were observed in the average ability group and the high ability group for deep learning approach adherence. By contrast, a decrease of 14.58% was observed in the low ability group for deep learning approach adherence.

Survey questions
This survey did not reveal any significant change in self-rated learning motivation, the influence of education factors, English ability, comprehension ability, and the English ability gap between peers.
Of the four affective factors, self-esteem accounted for 33.06% for selection and became the most impactful affective factor in English learning in the screening survey (as shown in Figure  4), while language learning motivation accounted for 31.22% for selection as the most impactful affective factor in English learning in the post-study survey. These two affective factors were frequently selected by participating students in both screening and post-study surveys.

Figure 4: Influential Affective Factors
The four opinons of affective factors -sense of belonging, self-esteem, language learning anxiety and lanuage learning motivation are made available for students to select in the screening survey.
Of the four areas of comprehension abilities -Interpreting the author's attitude or preference, Vocabulary, Understanding the main idea of the passage, and Grammar -understanding the meaning of the passage accounted for 47.02% and 43.06% in the pre-study screening (as shown in Figure 5) and the post-study surveys respectively as the most selected self-perceived strength in reading. Interpretation of the author's attitude or preference ranked as the second strongest strength in reading, accounting for 24.11% and 22.01% for selection in the surveys respectively. Vocabulary accounted for 36.85% and 31.58% for selection as the ability students intended to strengthen (as shown in Figure 6). Vocabulary was also the most selected weakness in both screening (39.53%; as shown in Figure 7) and post-study surveys (38.43%).

Self-percieved weakness in reading
Interpreting the author's attitude or preference Vocabulary Understanding the main idea of the passage Grammar The three opinons of self-percieved weakness in reading -interpreting the author's attitutde or preference, vocabulary, understanding the main idea of the passage, grammar are made available for students to select in the screening survey.
Among the survey responders, 53.29% perceived educational factors influential in English learning, 57.68% are somewhat motivated to learn English, 65.20% perceived their English skills fair, and 59.56% perceived their comprehension skills fair. More than half of participating students (50.16%) perceive some English ability differences between peers and 45.45% of students perceived their English skills are at the same level. Among three preferences in classwork (i.e., small group for gameplay, small group for presentation, discussion and role-play, and individual work), small groups for gameplay accounts for 39.13% and become the most selected classwork preference in the screening survey. However, individual work accounts for 36.78% of the post-study survey and becomes the most selected classroom preference.

Feedback from students and teachers
Twenty-seven students participated in the feedback interviews, two from the Social Work Stream and 25 from Nursing.
It has transpired that in free conversation students provided us with feedback on much more than what took place for them in working out the Tiered Tasks. This is most rewarding and points to the need to seriously take into account students' learning styles, anxiety factors, preferences and interests and the overall learning atmosphere and environment, and other factors, to meaningfully differentiate instruction for optimal effectiveness.
In summary, most students (55.55%) commented on their lecturers' teaching style and 48.15% commented on the applicability of course material. Educational factors seemed to be influential for many students in English learning. Twenty students (74.07%) indicated an end goal of English learning is to be able to communicate in English. Only 33.33% of the students interviewed regarded communication skills are for work, suggesting that most students regarded enhancing their overall language competency over professional competence in English. Two-thirds of students interviewed (66.66%) favor small group learning, showing that small group in-class activities are conducive to increasing classroom engagement in English learning.
The three opinons of slef-precieve strengths in reading -interpreting the author's attitutde or preference, vocabulary, understanding the main idea of the passage, grammar are made available for students to select in the screening survey.
Generally, the teachers focused rather heavily on students' academic performance and on their grades. They aimed to prepare students to meet the existing language standards or the career-related standards. Therefore, the practical application was the teaching goal. The theoretical benefits of differentiated instruction were not in doubt, however, Nursing teachers hesitated to differentiate students in the classroom because of the extra work to cater to students with different abilities. Furthermore, extra manpower required for addressing different ability groups' needs in a differentiated classroom and the follow-up work, if any, to ensure effective feedback was provided to each ability group -this was not really available for the study at the right time.

Effectiveness of the differentiated instruction
The results of the research study indicate that students' inference-making skills improved markedly. The improvement is believed to be brought by the tiered tasks. The class focuses on profession-related verbal communication skills, through which students are trained to understand and respond to verbal communication rather than comprehension. Although the course materials may have some impact on students' ability in attending to relevant information in conversation, the tiered activities are believed to enhance students' ability to understand the main idea of a paragraph and to look for consistent or inconsistent information in a paragraph.
According to Tomlinson (2002), challenging, analytical questions or activities are expected to engage students with high ability. However, our results do not seem to support this view. Students with high ability scored low on the challenging question of Tiered Task 1. On the other hand, such activity scores may not represent the intelligence or learning ability of the students, but rather their classroom engagement. Hence, we may conclude that the tiered tasks may not help to engage students with high ability in English learning. Nevertheless, our literature review seems to suggest that students with lower abilities would be engaged with repetition and this supposition is in line with our results. The negative correlations between screening scores and all three tiered tasks suggest that their classroom engagement is high. Along with the improvement observed in Grammar and Inference-making, we can conclude that the tiered tasks and differentiated instruction developed are likely to benefit students with low and average abilities.

The participating teachers' overall feedback
The participating teachers' views regarding the effectiveness of differentiating part of the instruction, notably during the three intervention points respectively at Week 3/4; Week 7/8; and Week 11/12, varied, particularly regarding the impact of affective factors and educational factors on learning motivation. Since this research study focused largely on students' reading and vocabulary development abilities, affective and other educational factors were not taken into account for appraising and evaluating the overall effectiveness of DI. Nevertheless, it was opportune for the research team to see such factors reveal in the overall process, for which relevant data were therefore recorded accordingly, paving the way for the next study the team plans to undertake.

Impact of this research study on the participating teachers' professional development
Obviously the pressure of extra work on the teachers collaborating in this study was great, however, we are happy to report that they considered it a rare and opportune moment for them to advance academically and professionally, perceiving it as a chance for them to sharpen their practice and their thinking about English language teaching and learning. The research team was very happy to detect it and found it a very positive feature in the differentiating process.

Limitations
Although students' English skills have been demonstrated to improve as a result of the differentiating intervention, a control group was lacking in the study to examine the effect of the course content on verbal communication. On this score, more research is deemed useful and timely to examine the impact of differentiated instruction for tertiary students in Hong Kong. Therefore, it is our plan for the next research study that a control group will be set up in parallel with the subject group, so as to better verify and confirm the positive impact of differentiated instruction on students to raise their English abilities.

Conclusion
Based on the discussion of the data collected from this research throughout the entire process in the 13-week course duration, and right afterwards in one-on-one interviews and focus groups, while at the same time taking reference of inferences from the analysis of the data, it has demonstrated that not only has differentiating the instruction for the participating students in all the sections/classes made a difference in students' academic performances, (which have thus led themselves to at least three different categories of improvement beginning from the first intervention point all the way to the end of the course), but more substantially the study has identified different students' learning styles, psychological factors impacting their learning, their preferences for group work and other realted factors such as readiness for tackling tiered tasks, and overall interest and enhanced motivation in the learning process. This outcome corresponds closely to the classic model for differentiated instruction (notably presented on page 2 in Tomlinson & Moon, 2013).
To echo the remark made above, this research study has indeed revealed much more than merely the effectiveness of differentiating the instruction to enable and facilitate students to improve on their proficiency levels. The research team has also reaped from collecting and analysing the data insights and issues relating to both the students' and the teachers' dimensions, which warrants an in-depth study to generate a MODEL for optimally differentiating the ESL/ESP curriculum in the near future.