Disertasi
Driving students along data-driven learning in efl writing classrooms: educational applications and implications / Humairah Fauziah
Abstrak
This study aimed at investigating the effectiveness of DDL (direct and indirect corpus access) on EFL students rsquo English writing achievement. Prior to this research a preliminary pilot study was conducted to gauge the effects of direct corpus access specifically through concordancing tools on students writing performance. The findings from this pilot study indicated that relying solely on direct corpus access did not yield significant enhancements in the students writing outcomes. Three primary shortcomings were identified in the pilot study that likely contributed to these results. First students faced difficulties in effectively selecting Keywords in Context (KWIC) and analyzing concordance lines. These challenges were attributed to the students English proficiency levels and a lack of sufficient training in both KWIC analysis and the operation of concordance tools. Second the students might exhibit a reluctance to participate in DDL activities which was partly due to the unfamiliar nature of the learning environment as it differed significantly from their traditional educational experiences. This unfamiliarity with the DDL approach likely hindered their engagement and overall effectiveness. Furthermore many studies have shown that DDL can improve students English writing skills. However much of the existing research treats DDL activities in isolation without integrating them into the curriculum and regular classroom activities. Additionally these studies often neglect other tools that could further enhance students language learning. There is a significant gap in research on the comprehensive integration of DDL particularly in integrating it into students rsquo curriculum including the teaching materials and activities. Besides there has been a lack of evidence regarding the long-term impact of DDL on language retention especially when compared to other teaching methods. To address the shortcomings identified in the pilot study I adopted a new approach for the current research. This involved incorporating both direct and indirect corpus access into Data-Driven Learning (DDL) activities extending the duration of these activities and blending DDL with traditional teaching methods. The goal was to enhance student engagement and acceptance without disrupting the regular classroom routine. Direct corpus access was facilitated through concordancing activities which allowed students to interact with authentic language data and discover language patterns on their own. Indirect corpus access was implemented via printed concordance line analysis and reading text observations providing a more guided and structured learning experience. To ensure a comprehensive integration of DDL activities in the English as a Foreign Language (EFL) writing classroom I utilized the EODI (Explore Observe Discover and Internalize) language learning pattern. This study spanned an entire semester encompassing 16 meetings of 90-minute learning activities. To integrate DDL with students regular routines a textbook aligned with the curriculum and learning objectives was provided. To accommodate students familiar tools students were initially surveyed about the tools they commonly used for learning English. The study also integrated direct and indirect corpus access into the teaching materials included in the textbook. The study utilized practical action research involving 59 non-English department students across three different writing classes. Data collection methods included essay writing tests sentence completion exercises and questionnaires on students perceptions of the DDL activities and teaching materials. The essay writing test data were analyzed using the Wilcoxon test sentence completion exercise data were analyzed using descriptive statistics and one-way ANOVA and questionnaire data were analyzed thematically. The results of the essay writing tests revealed a marked improvement in the quality of the students writing post-DDL involvement. Specifically the essays composed after the DDL activities were not only longer but also demonstrated a higher degree of organization and coherence. In addition the effectiveness of direct corpus access through concordancing was measured by comparing it with the other tools that the students had used for learning English before participating in this study. These other tools like concordance tools functioned as reference aids for correcting language errors. In this case I assigned the students to do a sentence completion exercise focusing on four key language aspects subject-verb agreement tenses prepositions and incorrect word usage. These aspects were chosen based on the common errors identified in the students initial essays. The results of this comparison revealed that while concordancing showed the lowest accuracy rates initially it had a significant impact on the students retention of correct answers from the first exercise to the second. At the end of the semester the students perceptions of both direct and indirect corpus access methods were evaluated through a questionnaire. The feedback revealed a nuanced understanding of the strengths and weaknesses of each approach. The students felt that direct corpus access particularly through concordancing significantly enhanced their awareness of how specific language uses vary across different contexts. This method was noted for improving their analytical and critical thinking skills as it encouraged them to scrutinize KWIC and form hypotheses about language usage based on their observations. Conversely indirect corpus access which included tasks such as analyzing provided concordance lines and reading texts was valued for its ability to present language data in a more structured and manageable format. This approach helped mitigate the risk of students feeling overwhelmed by the data as it focused their analysis on pre-selected examples. Additionally students appreciated that activities involving reading texts contributed to better organization in their essays. They found these exercises beneficial in structuring their thoughts and improving the coherence of their writing. The students faced two main challenges with direct corpus access through concordancing. The first challenge was the complexity of analyzing KWIC. This complexity was linked to their foundational understanding of language elements such as parts of speech and grammar. Many students perceived that they struggled with KWIC analysis because they did not have a firm grasp of how different parts of speech function within sentences. This lack of grammatical knowledge hindered their ability to interpret KWIC data effectively making it difficult to understand and apply language patterns. The second challenge was the difficulty of analyzing concordance lines. The vast amount of language data students was required to process often felt overwhelming. The sheer volume of information made it hard for students to identify relevant patterns and draw useful conclusions. Additionally students reported that the context of the language data did not always align with their academic interests or educational background. For example they found some reading texts unengaging because they did not relate to their field of study which was education. This misalignment affected their motivation and learning experience indicating a need for more relevant and contextually appropriate materials in their studies. Based on the implementation of both direct and indirect corpus access in this study I have identified several key reflections for improving future DDL (Data-Driven Learning) integration. These reflections address the previously used EODI (Explore Observe Discover Internalize) teaching pattern. Specifically I suggested incorporating an additional stage after the ldquo Discover rdquo phase termed the ldquo Confirm rdquo stage represented as C which is essential for EFL students to validate their insights about language use. In the ldquo Expose rdquo phase involving the students in choosing reading materials aligned with their interests can enhance relevance and engagement. The combined use of corpora and other reference resources is recommended to address potential challenges faced by EFL learners. Furthermore expanding the use of the Sketch Engine to include features beyond concordancing and integrating tools like Google search and GenAI during the ldquo Observe rdquo stage can aid in overcoming difficulties with language data comprehension. Enhancing the ldquo Discover rdquo stage through peer collaboration can improve analysis and understanding while the ldquo Confirm rdquo stage facilitates active engagement and discussion of hypotheses. Finally making the ldquo Internalize rdquo stage more dynamic through peer reviews and multimedia tools can foster collaborative learning and deeper comprehension. Integrating these stages and employing various teaching methods and resources can make DDL activities more effective and relevant thereby improving students English writing skills.