Evaluating the Efficacy of Differentiated Instruction in EFL Speaking Classes: A Classroom Action Research Study

,


INTRODUCTION
Speaking skills are essential in learning English as a Foreign Language (EFL) because it facilitates effective communication.It enables EFL learners to interact clearly, exchange and understand different viewpoints, and participate actively in social contexts.Moreover, proficient speaking skills boost confidence in various activities, foster relationships with people from diverse backgrounds, and enhance career opportunities globally (Devi, 2022).Consequently, enhancing English speaking skills has become a priority, with substantial research focusing on achieving speaking proficiency (Dincer, 2017).In many EFL classrooms, speaking is often emphasized, with numerous teachers adopting Communicative Language Teaching (CLT) methods (Teh, 2021).
However, learning to speak English well is complex and challenging.Effective speaking requires managing verbal communication and non-verbal cues like body language and tone of voice simultaneously (Cendra & Sulindra, 2022).Common challenges in EFL speaking classes include inhibition, having nothing to say, low participation, and frequent use of the mother tongue (Shen & Chiu, 2019).Additionally, many EFL classes are comprised of learners of mixed proficiency levels.Researchers like Adhikary (2023) and Al-Subaiei (2017) observe that classes with mixed proficiency levels are increasingly common.Al-Shammakhi and Al-Humaidi (2015) note that teaching students with varying learning abilities is a widespread challenge.Such classes include students with different proficiency levels and learning approaches, posing unique challenges for students and teachers.
One notable challenge in mixed-proficiency EFL speaking classes is the limited interaction among students, which can impede the development of communicative competence (Lap & Thy, 2017).Students' varying levels of participation indicate that they learn at a different pace.Students eager to speak English may achieve fluency more quickly than their less willing peers.Similarly, students with more ideas to express tend to have more opportunities to practice speaking, while those struggling to find ideas may practice less.This disparity can lead to significant differences in achievement and fluency levels among learners.Additional challenges in mixed-proficiency classes include a hostile learning environment, inadequate materials, and ineffective teaching methods (Luspa, 2018).Consequently, traditional one-size-fits-all approaches typically cater to mid-range proficiency levels are no longer effective.
Addressing these issues requires Differentiated Instruction (DI), which focuses on meeting students' needs (Ur, 2012).DI is a learner-centered strategy designed to accommodate learners' unique differences, ensuring every student can succeed (Abramova & Mashoshina, 2021;Neuvirthova & Gadusova, 2021).It involves tailoring classroom activities to offer various options for absorbing information, understanding concepts, and expressing knowledge.Effective DI requires teachers to adapt their instructional methods to match students' readiness levels, learning profiles, and interests, using various strategies and tools to meet diverse learning needs (Neuvirthova & Gadusova, 2021;Tomlinson, 2017).
According to Tomlinson and Moon (2013), effective implementation of Differentiated Instruction (DI) involves considering five key components: content, process, product, affect, and learning environment.These are guided by three fundamental principles: readiness, interest, and learning profile (Tomlinson, 2001).First, instructional content should be engaging and tailored to learners' diverse needs, considering their proficiency levels, interests, and learning preferences (Tomlinson, 2014).Second, the learning process should be flexible, accommodating each student's initial learning profile and allowing variations in learning strategies and pace.Third, the learning products-often linked to summative assessments-should offer diverse and alternative ways of demonstrating learning outcomes.Fourth, addressing students' affective needs involves fostering effective communication, collaboration, and interaction between students and teachers, thus nurturing empathetic and dynamic relationships within a supportive learning community.Lastly, the physical and social learning environment should be inclusive and conducive to the needs of learners with varying preferences and behaviors.
Several studies have underscored the benefits of DI in the context of Indonesianspeaking classes.Arianto et al. (2023) emphasized how DI can enhance students' confidence in speaking.Rumkoda and Alinda (2022) observed improvements in students' speaking skills through DI-based material selection tailored to students' interests.However, several studies have documented strategies employed by EFL teachers in managing mixedproficiency classes, mainly drawing on the principles of Differentiated Instruction (DI).Sun (2023) explored teachers' practices across the five DI components-content, process, product, affect, and learning environment-in emergency online classrooms focusing on reading skills.The study found that by differentiating reading tasks with varied material levels, student autonomy in reading L2 texts improved, although it also widened the achievement gap between higher and lower achievers.Mehany (2022) observed that students engaged in DI-based writing instruction showed superior essay writing proficiency compared to non-DI settings.Additionally, a study by Sapan and Mede (2022) in Turkish secondary-level English classrooms reported enhancements in foreign language achievement, motivation, and learner autonomy through DI strategies like tiered activities, learning centers tailored to different styles, and modified post-achievement tests.
Research in the ELT field on DI has predominantly focused on addressing learners' differences in learning preferences and interests.However, more research is needed concerning the practical application of DI to support learners' varying English proficiency levels specifically.Furthermore, much DI research in ELT has been conducted in school settings with young and secondary school-aged learners.There has also been limited focus on improving speaking skills, with more studies concentrating on reading and writing skills or general English proficiency.Furthermore, the discussion highlights a significant gap in research concerning the application of Differentiated Instruction (DI) to enhance EFL speaking skills among adult learners.Given these gaps, this study aims to address challenges in mixed-proficiency adult speaking classes by exploring two research questions: 1) To what extent does DI improve speaking skills among mixed-proficiency adult learners?2) How do students respond to the implementation of DI?This research seeks to contribute to the field by focusing on these underexplored areas, thus potentially offering new insights into practical strategies for teaching speaking skills in diverse adult learner settings.

METHOD
This study utilized a mixed-methods participatory action research design (Kemmis et al., 2014), focusing on enhancing EFL speaking skills in a mixed-proficiency class through the principles of Differentiated Instruction (DI) (Tomlinson, 2014).The participatory approach was designed to be empowering, with teachers and participants collaboratively addressing identified teaching challenges.The research process followed four phases: planning, acting, observing, and reflecting (Burns, 2010).During the planning phase, strategies were devised to differentiate Instruction by process, content, and product to meet diverse learner needs.Materials were also tailored for all proficiency levels.Each 120-minute class session aimed to develop general communication skills.In the acting phase, DI activities included flexible grouping to foster collaboration across different proficiency groups, tiered materials, exercises to suit varied knowledge levels and interests, and diverse open-ended tasks to assess learning outcomes.These steps ensured each learner's needs were addressed, enhancing the educational experience and outcomes.
In addition, the observation phase involved assessing speaking improvements through students' reflection.The subsequent reflection phase reviewed the data collected to evaluate the effectiveness of the interventions.The schedule included a pre-test, two progress tests and reflections, and a final test.The study featured nine adult male EFL learners in a nonformal classroom setting in Yogyakarta, Indonesia, aged 18 to 31.These participants came from diverse regions and educational backgrounds-two had college degrees, while seven had completed high school.Their varying levels of English proficiency and different educational experiences presented unique challenges, particularly in meeting the course objective of effective general communication in English, covering both formal and informal contexts.
To address the first research question regarding improving speaking skills, the study utilized a series of assessments: a pre-test at the initial meeting, progress tests at the end of the first and second cycles, and a post-test after the third cycle.The pre-test gauged learners' initial speaking proficiency, while progress tests provided feedback for planning subsequent cycles and tracked the development of speaking skills.The final post-test assessed the learners' overall improvement.The assessment scale followed the Common European Framework of Reference (CEFR), focusing on five criteria: range, accuracy, fluency, interaction, and coherence.The speaking test, modeled on the Cambridge Key English Speaking Test CEFR Level A2, involved two students and two examiners-one facilitating the conversation and another assessing.
The test had three parts: guided questions, a paraphrasing exercise, and a direct interaction on course topics.To ensure reliability, performances were evaluated by both examiners, with an inter-rater reliability score of 0.86 (p < 0.05), confirming high reliability.
The study used reflection journals and semi-structured interviews to address the second research question on students' responses to Differentiated Instruction (DI).Reflection journals provided insights into students' experiences, while interviews offered deeper perspectives on the DI learning environment, both critical for understanding DI's impact on learner engagement and satisfaction.

The Improvement of Speaking Skills
The study was conducted in three systematic cycles with the central intervention through DI.The intervention consists of differentiating the lesson content through tiered inputs and exercises, differentiation of process through flexible grouping, and differentiation of learning products through open-ended tasks across the cycles (Tomlinson, 2017).
To examine students' progress over the cycles, scores from the pre-test, two progress tests, and a post-test were examined through a descriptive statistical analysis.The pre-test showed students' baseline proficiency in speaking skills, while the post-test showed students' final learning results in the DI-based learning environment.Students' scores obtained from the tests are presented in Figure 1.The study employed the Friedman Test to assess the statistical significance of these score improvements.This non-parametric test is appropriate for analyzing small sample sizes across multiple time points without assuming a normal distribution (Pallant, 2016).This allowed for direct analysis of the score data using the Friedman Test.The outcomes of this test are presented in Table 1..000 The Friedman Test results revealed a statistically significant improvement in student speaking test scores over four assessments: the pre-test, two progress tests, and the posttest (χ²(3, n = 9) = 25.93,p < .05).This signifies that the student's speaking abilities significantly increased after completing three cycles of the DI-based program.The data reveals that Differentiated Instruction (DI) strategies have notably enhanced students' speaking skills, corroborating findings from other research on DI's effectiveness in EFL speaking classes.Meşe and Mede (2023) found significant improvements in EFL speaking proficiency and self-regulated learning among students in a Turkish higher education institution when DI was applied during online sessions.Similarly, Arianto et al. (2023) observed that DI elements boosted secondary students' confidence in English speaking.Additionally, Adawiyah (2018) demonstrated positive effects on speaking skills through content-based Instruction, a DI approach.These studies underscore the importance of teachers' commitment to the consistent use of DI, ensuring inclusive education that maximizes learning outcomes for all students (Naka, 2018).

Students' Responses to the Implementation of DI
The qualitative data obtained from the two research instruments reveal various responses from the participants.The data were categorized into four overarching themes, including (1) responses to the tiered contents, (2) responses to the flexible grouping strategy, (3) responses to the open-ended tasks, and (4) perceived speaking skill improvement.

Responses to the Tiered Contents
In this study on speaking skill development, multimodal texts were introduced at the beginning of each new topic to provide diverse language inputs.These texts-videos, written texts, and audio recordings-served as differentiated content, adjusted for three levels of student proficiency to meet their varying degrees of readiness throughout the action research cycles.
In the first cycle, for the topic "Nurturing the Nature," videos of different lengths and linguistic complexities were presented.Students, organized into homogeneous proficiency groups, received tailored inputs on environmental care.Student 2 reflected, "Through watching the video, I can understand well how to explain a procedure."(ST 02-Reflection), while Student 8 remarked on the relevance to their personal experience, "...From that video, I can exercise my listening.I also like watching the video about cleaning a fish pond.It is closely related to my life."(ST08-Reflection).Both students, from different proficiency levels, found the content appropriate to their abilities and interests.
In subsequent cycles, all students received texts on the same topics, such as "Natural Phenomena" and "Barack Obama," but these were modified to suit different proficiency tiers.Lower-tier students received shorter, more straightforward texts, whereas upper-tier students received longer, more complex versions.Students responded positively, with comments like, "I love the activity of reading a famous person's life story.It helps me practice retelling someone's life story."(ST03-Reflection) and, "I guess the activities and forms of exercises are suitable with my ability."(ST01-Reflection).Another affirmed, "Yes, we were provided with sufficient text about natural phenomena."(ST02-Reflection).
Most students felt the materials matched their proficiency level, underlining the significance of teachers tailoring content to learners' comprehension abilities.This echoes research by Dack et al. (2022), Magableh and Abdullah (2020), and Mardhatillah and Suharyadi (2023), which stresses the importance of modifying content to align with learners' developmental range and learning progress.These adaptations are critical in facilitating learners' understanding and engagement, enhancing the overall learning experience.

Responses towards flexible grouping strategies
Flexible grouping in Differentiated Instruction (DI) involves strategically assigning students to groups based on their proficiency levels for some sessions.In contrast, in others, they are mixed or allowed to choose their groups.This approach is designed to optimize knowledge delivery and peer interactions.For instance, homogeneous groupings are utilized during input-reception sessions to ease knowledge delivery tailored to students' capabilities.
In contrast, heterogeneous groupings are favored during task discussions to facilitate peer support and foster teacher-student relationships (Lai, 2011).
Students' responses underline the benefits of this method.One student appreciated peer support in group discussions, noting, "I love the group discussion activity because I can get help from my friends, and I can help other friends too" (ST06-Interview).Another student highlighted the learning advantage of working in diverse groups, stating, "What I enjoy most is working in a discussion group because when I do not know particular words, my friends can explain them to me" (ST01-Interview).These comments confirm that flexible grouping can significantly enhance student engagement and learning performance, particularly by supporting underperforming students.
The positive impact of collaborative learning in these settings is further supported by students who find the mixed-proficiency groups beneficial for providing and receiving assistance.For example, a student mentioned, "For me, group work is beneficial.I can contribute my knowledge and also seek help from more knowledgeable peers" (ST04-Interview), and another added, "In our groups, I assist my friends with difficult words" (ST07-Reflection).Such interactions improve language skills and boost confidence and engagement among students.
However, the approach has its challenges.Some students have voiced concerns about homogeneous groupings, feeling they might benefit more from being mixed with higherproficiency peers.One student shared, "I am not comfortable being grouped with peers of similar proficiency; having more competent friends in the group could help us more" (ST09-Interview), and another suggested, "More capable peers should be distributed among groups to aid those who are less proficient" (ST06-Interview).This feedback highlights the need for careful consideration in group composition to ensure all students benefit optimally from the collaborative learning environment.

Responses to the open-ended tasks
In Differentiated Instruction (DI), "products" are the outputs students create as evidence of their learning, tailored to each student's readiness, interests, and learning profiles.This differentiation ensures that tasks match students' capabilities and encourage engagement at various difficulty levels, often involving different amounts of teacher involvement (Gangi, 2011;Sun, 2023;Magableh & Abdullah, 2020).Open-ended tasks enable students to produce multiple correct answers and to engage according to their ability levels.
Tomlinson (2017) emphasizes using performance assessments, which allow students to demonstrate their proficiency through various forms of expression and alternative methods.These assessments typically offer varying degrees of difficulty and diverse types of evaluation and scoring.In our study, most tasks incorporated into each action research cycle were open-ended and concluded with a performance-based assessment.Students' responses to these methods have been overwhelmingly positive.For instance, one student reflected, "I can perform the group role-play very well.I acted as 'the neighbor,' and I think it suits my ability so that I can do my part in the role-play."(ST09-Reflection).Another student shared, "I did the interview task with my friend.I am happy that I could write my questions and try to write a short report from the interview."(ST07-Reflection).
These reflections show that students value the ability to create learning products through role-play and interviews, which align with their skills and interests.The flexibility and creativity inherent in these tasks enhance students' engagement and instill a sense of pride and enjoyment in their achievements.Such positive educational experiences support the findings of Suwastini et al. (2021), who noted that product differentiation not only encourages but delights students as they tackle assessments.Furthermore, accommodating diverse proficiency levels and preferences through product differentiation optimizes each student's potential, demonstrating the critical role of open-ended tasks in allowing students to express essential understandings uniquely and effectively (Arianto et al., 2023;Tomlinson, 2014).

Perceived improvements in speaking skills
In the action research cycles focused on Differentiated Instruction (DI), students' reflections revealed significant perceived improvements in their speaking skills, providing valuable insights into the DI-based actions' effectiveness from their perspectives.Their reflections underscore the dual aspects of learning outcomes and processes, highlighting their personal development and increased engagement in learning activities.One student shared, "I feel like I experience progress in myself.I can understand what my friend is talking about, and I can reply a little in English," indicating better comprehension and increased ability to participate in English (ST01-Reflection).Another noted, "I made improvements in group discussions.I stated some ideas.I can already participate in group discussions and try to correct friends' works," reflecting their active contribution and confidence in using English in a group setting (ST09-Reflection).
Additionally, some students reported increased confidence in more public settings: "Now I am brave enough to speak in public, especially to my seniors" (ST02-Reflection), and "I have been more confident to speak in public, although my English is not good.I start to speak in English during the English Day" (ST03-Reflection).These comments highlight the shift from anxiety to confidence, allowing greater participation in communicative activities.Another student candidly expressed overcoming their fear of making grammatical mistakes: "I used to speak to people a little or quietly because I think about English grammar.I am afraid to be wrong.Now I do not care about that" (ST07-Reflection) illustrating a significant reduction in language anxiety.These qualitative insights complement the quantitative findings from the study, illustrating not only the improvements in speaking abilities-as students became more proficient interlocutors and active participants in discussions-but also how they perceived their learning processes.Students reported feeling more empowered and confident, which is echoed in previous DI research.Studies suggest that DI improves comprehension of critical concepts and enhances students' interest and intellectual engagement (Girma, 2022;Kotob & Ali Abadi, 2019;Magableh & Abdullah, 2020).Furthermore, the process improvements noted-increased confidence and decreased anxiety-are likely due to DI's emphasis on collaborative learning, which fosters a supportive and engaging learning environment (Celik, 2019;Meşe & Mede, 2023;Suwastini et al., 2021).This supportive setting helps mitigate the pressures associated with language learning, facilitating a more conducive atmosphere for active and confident participation.

CONCLUSION
This study employed Differentiated Instruction (DI) techniques focused on learning content, process, and product over a three-cycle action research framework.The findings demonstrate that DI has significantly contributed to students' success in developing speaking skills, as evidenced by consistent progress in test results throughout the implementation of the action research program.Students' positive perceptions and responses to DI strategies highlight the effectiveness of pedagogical approaches that are tailored to individual needs and are flexible.Using tiered learning content, varied tasks with flexible grouping options, and incorporating performance assessments and open-ended tasks have enhanced student engagement, motivation, and support in their language learning journey.This suggests that DI effectively addresses specific challenges in the EFLspeaking classroom and promotes a more inclusive and supportive learning environment.Celik (2019) noted that an inclusive classroom fosters a productive atmosphere where students feel comfortable and valued.
Given the positive impact observed, it is recommended that EFL teachers consistently apply DI strategies in their teaching practices.Even with the constraints of fixed textbooks, teachers can differentiate Instruction through tiered activities and by varying the products students create to demonstrate their learning.Where content differentiation is limited, teachers can still adjust the learning process by utilizing flexible grouping strategies, which are particularly useful in large classroom settings.Furthermore, since DI is not a rigid formula but a flexible set of strategies, it allows teachers to select the most appropriate methods to meet their students' diverse needs.

Figure 1 .
Figure 1.Students' speaking test scoresThe bar charts for each student (S1 -S9) show increased speaking test scores over the cycles with the implementation of Differentiated Instruction (DI).Each student's progression is indicated by their respective bar charts, with most students demonstrating improvements throughout the cycles, although the extent of progress varies.On average, scores rose from the pre-test (mean = 60.14, SD = 7.23), through the progress tests (mean = 62.33,SD = 7.3 and mean = 66.7,SD = 5.18), to the post-test (mean = 69.72,SD = 5.97).The study employed the Friedman Test to assess the statistical significance of these score improvements.This non-parametric test is appropriate for analyzing small sample sizes across multiple time points without assuming a normal distribution(Pallant, 2016).This allowed for direct analysis of the score data using the Friedman Test.The outcomes of this test are presented in Table1.

Table 1 .
The Output from the Friedman Test