Action Research for Implementing Task-Based Language Teaching and Enhancing Speaking Skills in a Class of 12th Grade English Students in an Angolan School ()
1. Introduction
English has become a global language, fostering communication across cultures and playing a central role in education, business, and social interactions (Jenkins, 2007). This global prominence has driven a shift in teaching methodologies, moving away from rote memorisation of grammatical structures toward communicative and learner-centred approaches. At the heart of this paradigm is the concept of communicative competence, introduced by Hymes (1972), which comprises grammatical, sociolinguistic, discourse, and strategic competencies. These components emphasise the need to integrate linguistic accuracy with contextual understanding, cultural appropriateness, and the ability to navigate real-world communication challenges.
Building on this foundation, Communicative Language Teaching (CLT) advocates for an integrated approach that combines various language subskills—speaking, listening, reading, writing, vocabulary, and grammar—within meaningful and interactive contexts (Canale & Swain, 1980; Richards & Schmidt, 2010). Task-Based Language Teaching (TBLT), an evolution of CLT, places practical, goal-oriented tasks at the core of instruction, enabling learners to engage in authentic communication while addressing real-world scenarios (Ellis, 2009; Nunan, 2004). Such tasks often involve collaborative activities, such as group discussions, problem-solving, and role-playing, which foster fluency, coherence, and confidence.
Despite the recognised importance of oral communication, speaking skills have historically been neglected in English language teaching, particularly in contexts where reading and writing dominate curricula (Richards & Rodgers, 2014). Early approaches, such as grammar-translation and audio-lingual methods, prioritised accuracy over fluency, neglecting essential communicative aspects. Although CLT and TBLT have shifted this focus, challenges persist in English as a Foreign Language (EFL) setting, where learners often struggle with fluency, confidence, and limited opportunities to practice outside the classroom (Tuan & Mai, 2015; Leong & Ahmadi, 2017).
For Angolan students, these challenges are compounded by factors such as socio-economic barriers, resource limitations, and minimal access to authentic English input. Cultural nuances, anxiety, and the influence of learners’ L1 further impede oral proficiency. Research consistently highlights that active engagement in meaningful communication is critical for developing speaking skills (Ellis, 2003; Long, 2015). TBLT offers a practical and effective solution by creating low-anxiety, learner-centred environments that encourage interaction and build confidence. Tasks designed to mimic real-world communication enable learners to improve fluency and accuracy while reducing anxiety (Jeon & Han, 2006; ).
Empirical studies underscore the benefits of TBLT across various educational contexts. For instance, Mayo and Agirre (2012) and Farvardin (2017) report significant improvements in speaking fluency and confidence among EFL learners engaged in task-based activities. Similarly, research in Angola (; Graciano, 2019) highlights TBLT’s effectiveness in enhancing oral proficiency and fostering greater learner engagement. Notably, these studies reveal a strong desire among both learners and teachers to implement task-based methodologies, despite challenges such as resource constraints and institutional readiness.
In TBLT, the teacher’s role shifts to that of a facilitator, guiding learners through well-structured tasks and providing constructive feedback. These dynamic fosters a collaborative environment, enabling learners to experiment with language use and develop communicative competence. By addressing learners’ needs and tailoring tasks to specific contexts, TBLT promotes sustainable language acquisition and long-term skill retention (Ellis, 2009; Robinson, 2001).
This study explores the effectiveness of a TBLT intervention in enhancing the speaking skills of Grade 12 learners in Angola. By addressing the unique challenges faced in this context and leveraging task-based strategies, the research contributes to a growing body of evidence supporting TBLT as a transformative approach to language teaching. It seeks to demonstrate how pedagogical tasks can foster communicative competence, build confidence, and prepare learners for academic and professional success in an increasingly globalised world.
2. Methodology
This study adopted a mixed-methods approach, integrating qualitative and quantitative methods to evaluate the effectiveness of Task-Based Language Teaching (TBLT) in enhancing Grade 12 students’ speaking skills in Angola. A descriptive focus identified learners’ speaking difficulties, while triangulation of data collection methods—pre- and post-tests, questionnaires, and observations—ensured reliability and validity (Dörnyei, 2007). The research followed a Classroom Action Research (CAR) framework, enabling iterative cycles of planning, action, observation, and reflection (Mertler, 2012; Creswell, 2012). This cyclical process allowed the teacher-researcher to adapt strategies based on classroom dynamics and learner needs. A quasi-experimental design compared a TBLT-based experimental group with a control group following traditional methods, examining the causal relationship between teaching approaches and speaking performance (Campbell & Stanley, 1963).
The control group was exposed to conventional teacher-centred methods, primarily involving grammar-translation and rote memorisation. These instructional strategies were systematically compared with the TBLT intervention in terms of their intensity and authenticity. While both groups received an equal amount of instructional time, the key distinction lay in the nature of the tasks. The TBLT approach prioritised communicative, meaning-focused tasks, whereas the control group’s instruction placed greater emphasis on accuracy and form without integrating real-world communicative practice.
2.1. Participants
The population for this study were Grade 12 students enrolled in Escola de Magistério Secundário No. 135 “Comandante Liberdade” in Lubango, Angola. A total of 58 students participated, with the experimental group receiving TBLT instruction designed by the researcher, while the control group followed conventional teaching methods. Given the logistical constraints of the research setting and participant availability, convenience sampling was employed. As Dörnyei (2007) observes, this is “the most common sample type in L2 research,” particularly when working with “captive audiences such as students in the researcher’s own institution” (pp. 98-99). Convenience sampling was deemed appropriate due to proximity, accessibility, and the need to minimise disruption to institutional routines, as the study relied on pre-existing class structures (Cacumba, 2014).
The study included all Grade 12 English students enrolled in the selected institution, with participation being voluntary. No specific exclusion criteria were applied, as the aim was to reflect the diversity of the student population in order to examine the impact of task-based language teaching (TBLT) in a real-world educational environment.
To ensure ethical compliance, informed consent was obtained from all participants, ensuring their voluntary participation in the study.
2.2. Instruments
Data collection tools included pre- and post-tests, adapted from Swan & Walter’s The New Cambridge English Course 3, with tests based on Brown’s (2001) rubric covering grammar, vocabulary, comprehension, fluency, pronunciation, and task performance. Pre- and post-questionnaires (Shen, 2019) explored linguistic, affective, and instructional factors influencing speaking skills. Observation sheets and field notes documented student engagement and classroom activities, supplementing quantitative findings.
All students in the experimental group were given codes (anonymous names) such as TBL1, TBL2, TBL3, TBL4, TBL5, etc., while the control group were given codes such as CG1, CG2, CG3, CG4, CG5, and so forth, so that anonymity was guaranteed and the freedom to speak was assured. Every test was audio-recorded using a Huawei Y8 smartphone recorder for further transcription and careful analysis. Side by side, a scoring rubric adapted from Brown (2001) was used for each student, apart from that, important notes were taken as a fail-safe in case of a recording or reading/listening problem. Then, information gathered was carefully transcribed. Additionally, observation notes were taken during the implementation phase of the research to triangulate the data from questionnaires and tests. Furthermore, to administer the tests, in both the control and experimental group the researcher introduced himself and explained the purpose of working with them. The guidelines for the test were read to students.
2.3. Data Analysis
Audio recordings, speaking rubrics, and direct observations were used to collect and analyse data. Pre- and post-test results were transcribed and analysed using SPSS (v. 29.0). Normality and homogeneity were tested using the Shapiro-Wilk Test. Independent samples t-tests compared post-test scores between groups, with Levene’s Test assessing equality of variances. Content validity was ensured through alignment with the syllabus and expert review, while inter-rater reliability strengthened score accuracy.
3. Findings
This section presents the findings from action research on implementing Task-Based Language Teaching (TBLT) to enhance 12th-grade English students’ speaking skills at Escola de Magistério Secundário No. 135 Cmdte. Liberdade do Lubango-Angola. The study aimed to evaluate TBLT’s impact on improving speaking skills, addressing three research questions:
1) What speaking problems do 12th-grade English students encounter when speaking English?
2) What are the key phases of the intervention plan that address speaking difficulties through TBLT?
3) How can the implementation of task-based language teaching (TBLT) improve the speaking skills of 12th-grade English students at Escola de Magistério Secundário No. 135 “Comandante Liberdade” do Lubango-Angola?
The first research question was “what speaking problems a class of 12th grade English students encounter when speaking English?” This research question aimed to explore linguistic, affective, and instructional factors that may hinder students’ ability to speak English fluently. In order to answer this research question, the qualitative data from the pre– and post-questionnaire were analysed in the form of frequency and percentage of choices to reveal different factors that cause difficulty for students’ oral performance.
The pre-test results demonstrated a notable difference in the mean scores between the control and experimental groups. The control group (N = 31) achieved a mean score of 2.39 (SD = 0.268), whereas the experimental group (N = 27) obtained a higher mean score of 2.81 (SD = 0.306). While some variation in initial mean scores was observed, statistical measures, including effect size calculations, were employed to determine whether these differences were significant. Furthermore, efforts were made to balance the groups in terms of key learner characteristics, such as prior exposure to English and self-reported confidence in speaking.
Questionnaire data revealed that the primary linguistic challenges experienced by the students included vocabulary limitations, reported by 61.76% of the control group and 59.26% of the experimental group. Insufficient grammar knowledge was a more significant issue for the control group (70.59%) compared to the experimental group (48.15%). Pronunciation difficulties were noted by 44.12% of the control group and 37.04% of the experimental group, while the influence of Portuguese on English-speaking abilities was more pronounced in the control group (82.35%) than in the experimental group (70.37%).
In relation to affective factors, the control group showed moderate enjoyment of class discussions (67.65%), but only 41.18% reported motivation and confidence in speaking. Fear of criticism was evident in 41.18% of students, and classroom anxiety affected 32.35%. Despite higher enjoyment of English TV shows (70.59%), shyness was a concern for 44.12% of students. In contrast, the experimental group displayed stronger positive affective outcomes, with 85.19% expressing enjoyment of class discussions and 59.26% reporting confidence in speaking. Fear of criticism was reported by 37.04%, and classroom anxiety affected 40.74% of students. Additionally, 81.48% of students in the experimental group enjoyed watching English TV shows.
The findings for instructional factors revealed that class size was not considered problematic in either group, with 100% of students in both groups affirming this view. However, the control group faced significant challenges, including limited access to audio activities (91.18%) and the negative impact of classroom noise (67.65%). Peer support was deemed inadequate by 32.35% of students, and 76.47% reported a lack of access to technology. The experimental group, however, showed improvements in instructional factors, with 51.85% reporting the use of CDs for speaking activities. Classroom noise was perceived as an issue by only 25.93% of students, and 81.48% identified the classroom atmosphere as supportive of learning. Nonetheless, the experimental group highlighted limitations in the use of technology for watching videos or playing lesson-related games (0%), and only 22.22% reported opportunities to choose motivating activities or games.
The intervention plan consisted of two main stages: Cycle I and Cycle II, aimed at improving 12th grade students’ speaking skills through Task-Based Language Teaching (TBLT).
3.1. Cycle I: Improving Students’ Speaking Skills
After the pre-questionnaire and pre-test results, the implementation plan was applied, tasks were designed around real-life scenarios, using resources like projectors and Bluetooth speakers for videos, audios, realia and flashcards, and were carried out over 7 weeks with 12 lessons, focusing on grammar, vocabulary, comprehension, fluency, pronunciation and task Brown (2001), and communicative tasks such as role-plays, creative tasks, comparing things and/or people and problem-solving, among many. The selection of tasks was informed by several interrelated factors, including curricular requirements, students’ proficiency levels, and their communicative needs. Tasks were designed in accordance with established frameworks, particularly those proposed by Nunan (2004) and Willis and Willis (2007), which emphasise task complexity, cognitive engagement, and real-world applicability. In addition, student feedback and classroom observations were incorporated to refine task difficulty and ensure alignment with pedagogical objectives.
The following twelve topics were taught using Wills’ (1996) TBLT lesson framework:
- Greeting and Welcoming
- Asking for Confirmation
- Asking for Agreement
- Reporting and Responding to Emergency
- Describing Images
- Predicting a Storyline
- Storytelling
- A Cinema Trip
- Interviewing a Teacher (The World of Work)
- Sorting Paragraphs to Make Up a Story
- Expressing Opinions: Agreeing and Disagreeing
- Describing a Picture
- Planning a Dinner Party
- Describing a Process
As depicted in Table 1 below, after the intervention, in Cycle I, the experimental group (N = 27) had a mean score 3.78 (SD = 0.471). During the Observation and Reflection Phase, improvements in students’ confidence, vocabulary, and engagement were observed, despite facing challenges with technical aspects (electricity breakouts) and motivation. Although issues with grammar and pronunciation remained, they were not considered critical for task completion. Overall, TBLT seemed to be effective in enhancing students’ speaking skills, although it was demanding from the teacher/researcher. Additionally, students gradually seemed to adapt to a more communicative approach, demonstrating noticeable progress in their speaking abilities. Students showed notable progress in confidence, vocabulary, and task engagement. Challenges included adapting to TBLT, and technical disruptions. While some students preferred traditional methods, most achieved expected outcomes, demonstrating the approach’s potential for speaking skill development.
Table 1. Improvement of students’ speaking skills (Cycle 1).
|
N |
Mean |
Std. Deviation |
Std. Error Mean |
Experimental group |
Pre-test |
27 |
2.81 |
0.306 |
0.059 |
Post-test 1 |
27 |
3.78 |
0.471 |
0.055 |
3.2. Cycle II: Improvement of Students’ Speaking Skill
The second cycle unfolded over eight weeks, with adjustments made in both task content, complexity and delivery based on the mixed-ability classroom setting. Lessons were conducted every Monday and Tuesday from April 15th to June 11th, 2024, with a total of 12 lessons delivered, including the final post-test and post-questionnaire sessions.
The following topics were facilitated, with adjustments made to ensure students were given tasks appropriate to their proficiency level:
Love and other problems
Writing and giving a political speech
Writing a diary
Making and responding to complaints
Advertisements
Interviewing a classmate about families
Exploring holiday destinations in Angola
Visiting a doctor
My first short story (creative writing)
What do you think is the best way of travelling?
During cycle II it was observed that pronunciation and grammatical accuracy improved, supported by multimedia resources, task complexity adjustments increased engagement for advanced learners and provided accessible scaffolding for beginners and backup plans effectively minimized disruptions from technology failures. The TBLT intervention in Cycle II marked notable progress, particularly in enhancing speaking skills. Adjustments in task design, multimedia integration, and contingency planning created an inclusive, supportive environment. This underscores TBLT’s potential in fostering fluency and accuracy, contributing to broader educational advancements in Angola.
3.3. Post-Questionnaire Results
3.3.1. Linguistic Factors
The control group exhibited limited progress in vocabulary, grammar, and pronunciation, with a notable increase in participants reporting insufficient vocabulary (61.76% to 79.41%). There was greater awareness of Portuguese’s influence on errors, rising from 82.35% to 91.18%, but only minor gains in understanding English expressions and pronunciation issues. Conversely, the experimental group achieved substantial improvements, particularly in vocabulary usage, with 81.48% post-intervention disagreeing with difficulty in word selection, compared to 33.33% pre-intervention. Pronunciation and grammar knowledge also showed marked advancements, alongside reduced L1 interference, with only 51.85% acknowledging its impact post-study, down from 70.37%.
3.3.2. Affective Factors
In the control group, enjoyment of speaking activities and engagement with English media improved slightly (67.65% to 73.53% and 70.59% to 82.35%, respectively). However, confidence deteriorated, with an increase in fear of criticism (41.18% to 47.06%) and heightened classroom anxiety (58.82% to 73.53%). The experimental group, by contrast, saw a dramatic decline in fear of criticism (37.04% to 7.41%) and anxiety (40.74% to 7.41%). Confidence and motivation rose significantly (59.26% to 85.19%), supported by better preparation and family involvement. Issues with grammar-related sentence building also dropped drastically from 48.15% to 3.70%.
3.3.3. Instructional Factors
The control group saw no change in technological resource availability, with 100% still lacking access to audio materials. Marginal improvements were noted in classroom atmosphere and learner autonomy, with 35.29% feeling they had task choices compared to none initially. However, persistent challenges included classroom noise and insufficient peer support. In contrast, the experimental group benefited from increased access to audio resources (51.85% to 96.30%) and reduced classroom noise concerns (25.93% to 7.41%). Peer support significantly improved, enabling a collaborative environment, while the small class size and learner-centred approaches continued to foster effective learning outcomes.
In a nutshell, the findings of the study suggest that the linguistic, affective, and instructional factors that impact students’ speaking abilities are as follows:
In terms of linguistic problems, the study revealed that the primary speaking difficulties encountered by the 12th grade students included vocabulary limitations (61.76% to 79.41%), insufficient grammar knowledge (70.59%), pronunciation challenges (85.29% to 94.12%), and the influence of Portuguese on their English-speaking abilities (82.35% to 91.18%). However, after the intervention, the experimental group might have showed marked improvements in these areas, particularly in vocabulary usage, pronunciation, and mitigating the influence of their native language. The comparison of the two groups demonstrates the effectiveness of the intervention in addressing affective factors, as it seems that the experimental group exhibited significant improvements in confidence, motivation, and anxiety management compared to the control group, which saw an increase in anxiety (32.35% to 41.185%) and fear of criticism (41.18% to 47.06%). While improvements were observed in areas such as classroom atmosphere and student choice, challenges remain, particularly in terms of limited access to technological resources, specifically the lack of listening activities such as CDs before speaking exercises (91.18% to 100%), peer support (32.35% to 38.28%) and learner-centred activities (35.29%).
4. Post-test Results
Following the intervention, in Cycle I, the experimental group (N = 27) had a mean score 3.78 (SD = 0.471). Additionally, in the final cycle (cycle II), following the improvement of the intervention plan, the post-test results of the control group (N = 31) had a mean score of 2.54 (SD = 0.214), while the experimental group (N = 27) had mean score 4.27 (SD = 0.283).
As depicted from Table 2, in cycle I, pre-test results indicated that the control group (N = 31) had a mean score of 2.39 (SD = 0.268), while the experimental group (N = 27) had a mean score of 2.81 (SD = 0.306).
Table 2. Pre-test, Cycle I and Cycle II scores.
|
N |
Mean |
Std. Deviation |
Std. Error Mean |
Control group |
Pre-test |
31 |
2.39 |
0.268 |
0.048 |
Post-test |
31 |
2.54 |
0.214 |
0.041 |
Experimental group |
Pre-test |
27 |
2.81 |
0.306 |
0.059 |
Cycle I |
27 |
3.78 |
0.471 |
0.055 |
Cycle II |
27 |
4.27 |
0.283 |
0.054 |
Normality Test
The normality of post-test scores was assessed using Shapiro-Wilk tests (see Table 3). The results of the control group was: W(31) = 0.838, p = 0.028 while experimental group was: W(27) = 0.913, p < 0.001.
Table 3. Test of normality.
|
Kolmogorov-Smirnova |
Shapiro-Wilk |
Statistic |
df |
Sig. |
Statistic |
df |
Sig. |
Control group |
0.178 |
31 |
0.027 |
0.838 |
31 |
0.028 |
Experimental group |
0.283 |
27 |
<0.001 |
0.913 |
27 |
<0.001 |
The test of normality indicated significant deviations from normality for both groups (p < 0.05), suggesting that the post-test data were not normally distributed in either group. Given the non-normal distribution of the data, a Mann-Whitney U test was conducted to compare the post-test scores.
As displayed in Table 4, the ranks for the groups for the post test score revealed that the control group Mean Rank = 16.00 (Sum of Ranks = 496.00) while experimental group Mean Rank = 45.00 (Sum of Ranks = 1215.00). In order to make sure the difference between the two means was significant or not, the U test was run to measure the rank differences between the two sets of scores.
Table 4. Cycle II post-test score (Ranks: Mann-Whitney U test).
|
N |
Mean Rank |
Sum of Ranks |
Control group |
31 |
16.00 |
496.00 |
Experimental group |
27 |
45.00 |
1215.00 |
Total |
58 |
|
|
The U statistic is a measure of the rank differences between the two groups. A U value of 0.000 suggests a complete separation in ranks between the groups, indicating that the ranks of all observations in one group are consistently higher or lower than the ranks in the other group. Additionally, the Z value represents the standardized test statistic. This is the standardized test statistic (Z-score), which indicates how far the observed U value is from the mean of the U distribution and a negative Z value indicates that the first group (control group) has a lower mean rank compared to the second group (experimental group). Last, but not least, Asymp. Sig. (2-tailed), indicates that the p-value is less than 0.001, which is highly significant (Table 5).
Additionally, an independent samples t-test was conducted to compare the post-test total scores between the control group and the experimental group. Before performing the t-test, Levene’s Test for Equality of Variances was used to assess whether the assumption of equal variances was met. Levene’s Test for Equality of Variances indicated a significant result (F = 6.249, p = 0.015), suggesting that the assumption of equal variances was violated. Therefore, the t-test results with equal variances not assumed were used. The results of the t-test showed a significant difference between the mean scores of the control group (M = 2.5, SD = 0.199) and the experimental group (M = 4.27, SD = 0.283), t(45.884) = −27.174, p < 0.001, with a mean difference of −1.770 (95% CI: −1.901 to −1.639).
Table 5. Test statistics of control and experimental groups (Cycle II).
Test Statistics |
Mann-Whitney U |
0.000 |
Wilcoxon W |
496.000 |
Z |
−6.611 |
Asymp. Sig. (2-tailed) |
<0.001 |
5. Key Phases of the TBLT Intervention Plan for Addressing
Speaking Difficulties
The second research question was “what are the important phases of the intervention plan that address speaking difficulties through TBLT in this class?” This question aimed to design TBLT strategies and tasks to increase grade 12 English major speaking skills at “Escola de Magistério Secundário no. 135 ‘Comandante Liber-dade’ do Lubango-Angola”. It aimed to identify the key phases of the TBLT intervention to improve speaking skills in Grade 12 English students at “Escola de Magistério Secundário no. 135 ‘Comandante Liberdade’”. Data from two cycles and observations were analysed.
Cycle I: Pre-test results showed that the experimental group (mean = 2.81) had higher speaking skills than the control group (mean = 2.39). After the first TBLT intervention (Cycle I), the experimental group’s mean score rose to 3.78, while the control group improved slightly to 2.54.
Cycle II: Following refinements to the intervention, the experimental group’s mean score increased to 4.27, demonstrating further improvement. In comparison, the control group had a mean score of 2.54 after Cycle I and a lower mean rank in post-tests (Mann-Whitney U = 0.000, Z = −6.611, p < 0.001). Statistical tests, including the Mann-Whitney U test and an independent samples t-test, confirmed a significant improvement in the experimental group’s performance (t (45.884) = −27.174, p < 0.001), with a mean difference of −1.770.
Phases of the Intervention Plan:
1) Pre-Test Phase: Assessed baseline speaking skills. The experimental group showed slightly better initial skills.
2) Cycle I (First Intervention): Introduced TBLT strategies, leading to a significant improvement in the experimental group’s speaking abilities (mean = 3.78).
3) Cycle II (Refined Intervention): Further refined the TBLT strategies, resulting in a substantial improvement in the experimental group (mean = 4.27).
These phases show a clear progression, indicating that TBLT effectively addressed speaking difficulties in the experimental group.
6. Impact of Implementing TBLT on the Speaking Skills of
12th Grade English Students
The third, and main, research question of the current study was “how can the implementation of task-based language teaching (TBLT) improve the speaking skills of a class of 12th grade English students at Escola de Magistério Secundário no. 135 ‘Comandante Liberdade’ do Lubango-Angola”? It sought to evaluate how Task-Based Language Teaching could improve the speaking skills of Grade 12 English students at Escola de Magistério Secundário no. 135 “Comandante Liber-dade”. Data from pre- and post-tests across two intervention cycles and post-questionnaires were analysed.
The implementation of TBLT might have resulted in significant improvements in speaking skills. The following graphic summarizes the mean scores in the Pre-test, Cycle I, and Cycle II of the experimental group. As previously mentioned, (see Table 2) the mean score of the pre-test is 2.81, with a standard deviation of 0.306. The mean score of post-test I (Cycle I) is 3.78, with a standard deviation of 0.471. The mean score of Post-test II (Cycle II) is 4.27, with a standard deviation of 0.283.
Post-questionnaire results also indicate that linguistically, students showed substantial improvements in vocabulary, grammar, and pronunciation, with 81.48% reporting no difficulties in word selection post-intervention (up from 33.33%) and a reduction in the influence of Portuguese errors from 70.37% to 51.85%. Affective gains included increased confidence and motivation, with confidence levels rising to 85.19% (from 59.26%), alongside significant reductions in fear of criticism (37.04% to 7.41%) and classroom anxiety (40.74% to 7.41%).
These changes were supported by active engagement in tasks and family involvement. Instructionally, access to audio resources for speaking activities increased markedly, from 51.85% to 96.30%, while classroom noise concerns decreased (25.93% to 7.41%), and peer support improved, fostering a collaborative and learner-centred environment.
7. Discussion
This section attempts to interpret the findings of this study in relation to the research questions, emphasizing their theoretical and practical significance. The focus is on speaking difficulties faced by 12th grade English students, key phases of the TBLT intervention plan, and the impact of TBLT on students’ speaking skills at Escola de Magistério Secundário no. 135 “Comandante Liberdade” in Lubango, Angola.
7.1. Speaking Problems Encountered by 12th Grade English
Students
The study identified linguistic, affective, and instructional factors influencing speaking skills. Linguistically, students showed improvements in vocabulary, pronunciation, and reduced influence from their L1 (Portuguese), which had previously significantly limited their vocabulary, grammar, and pronunciation. Affective factors, including increased confidence and reduced anxiety, were noted, especially in the experimental group. In contrast, the control group experienced heightened anxiety over time. Instructional factors, such as access to technological resources and peer support, were critical, with the experimental group benefiting more from the TBLT approach despite challenges faced by the control group.
7.2. Linguistic Improvements and the Role of Focus on Form
The improvements in students’ linguistic skills support the arguments put forward by SLA researchers and TBLT proponents such as Willis (1996), Ellis (2003, 2009, 2014), Nunan (2004), and Long (2015, 2016), who assert that explicit language instruction—whether pre-task or post-task—helps learners notice and consolidate language forms. Long (2015) highlights that TBLT’s analytic syllabus includes a focus on form, enabling learners to address linguistic gaps in real communicative contexts. This approach differs from traditional methods that often isolate grammar instruction from communication. While grammar knowledge improved through post-task activities, it is important to examine whether this progress met TBLT’s theoretical expectations. Long (2016) argues that in TBLT, grammar attention arises reactively, driven by learners’ communicative needs. Students’ ability to refine language knowledge through task-based interactions reflects TBLT’s focus on language as a dynamic medium for communication, not just a set of abstract rules.
The reduction in L1 influence aligns with SLA research, which shows that L1 can either facilitate or hinder L2 acquisition (Odlin, 1989; Ellis, 2009). Task-based learning, combined with explicit instruction on language transfer issues, promotes more target-like language use (Lightbown & Spada, 2013; Gass & Selinker, 2008). By focusing on real-world tasks, learners notice and correct L1 errors during communication, as emphasized by Ringbom (2007).
Traditional approaches, such as grammar-translation or audio-lingual methods, emphasize grammar and vocabulary mastery before communication (Ellis, 2014; Richards & Rodgers, 2014). These methods often isolate linguistic structures, limiting learners’ communicative competence. In contrast, TBLT prioritizes communication, allowing grammar and vocabulary to emerge naturally through the different stages of the TBLT framework (from pre-task to the main task). As Ellis (2003, 2014) and Long (2016) argue, tasks in TBLT encourage language use in real contexts, with linguistic properties occurred during task phase looked at closely at the post-task to address specific communicative needs.
Unlike traditional deductive grammar instruction, TBLT employs inductive learning, where learners encounter grammatical structures within tasks and reflect on them afterward (Long, 2016). This cyclical process of communication and reflection supports both fluency and accuracy, contrasting with the linear, rule-based approach of more traditional methods.
In summary, the intervention plan of this study and the respective findings might have highlighted the fact that TBLT emphasizes communication first, with grammar explored through post-task reflection, fostering a holistic approach to language learning that enhances both fluency and accuracy.
7.3. Affective Factors and the Role of Classroom Atmosphere
The study’s findings highlight that TBLT enhances students’ confidence and motivation by fostering meaningful communication and learner autonomy, as suggested by TBLT proponents such as Willis (1996), Ellis (2003, 2009), Nunan (2004), Long (2015, 2016). A supportive, low-anxiety classroom atmosphere, emphasized by Harmer (2014), reduces the fear of making mistakes and encourages risk-taking. This contrasts with the control group, which exhibited increased anxiety, aligning with Horwitz et al. (1986) who note that fear of negative evaluation in traditional, teacher-centred classrooms can hinder language use, especially in the Angolan context where the dominant language is Portuguese and for most students the only source and exposition to the target language is the teacher, in the language classroom.
In the experimental group, collaborative tasks created a less pressured environment, lowering anxiety and encouraging communicative practice (Skehan, 1998; Willis, 1996). The task-based approach to teaching and learning helped build a supportive atmosphere, allowing students to practice without fear of harsh correction, thus increasing confidence. Conversely, more traditional methods that prioritize accuracy often leads to anxiety, as discussed by Krashen (1985), who argues that high anxiety acts as an “affective filter”, limiting language acquisition.
The study’s results align with Krashen’s (1985) affective filter hypothesis, suggesting that a low-anxiety environment is crucial for effective language learning. By focusing on meaningful communication rather than accuracy, the classroom atmosphere in this study encouraged students to take risks without the fear of judgment, supporting Dörnyei’s (2001) framework on motivation. However, challenges related to task difficulty, especially for less confident or lower-proficiency learners, were noted. The refinement of the initial intervention plan, in terms of task design and scaffolding, as advocated by Ellis (2003, 2009), Willis (1996) and Long (2015, 2016) helped mitigate these challenges and ensured that tasks were appropriate for learners’ skills, helping them build linguistic competence in a low-stress setting. In this study, scaffolding in the intervention took various forms, such as providing pre-task activities to activate learners’ relevant vocabulary and/or useful expressions for successful task completion, offering models of expected performance (watching or listening to a similar task), using clear instructions to guide learners through the task and assisting students when needed or required, especially after the refinement of the intervention plan.
Overall, the study demonstrates that task-based learning, when carefully designed, can create a supportive classroom atmosphere that fosters both emotional and linguistic development, confirming the importance of reduced teacher control and collaborative learning in increasing confidence and lowering anxiety. The findings suggest that with proper scaffolding, TBLT may be an effective approach for sustaining motivation and creating a conducive learning and teaching environment for both students and teachers.
7.4. Instructional Limitations and the Need for Resources
The study highlighted significant instructional challenges, particularly the lack of technological resources, which limited alignment with theoretical expectations. Research (Thomas & Reinders, 2010; Zhao, 2003) emphasizes the transformative role of technology in language education, providing learners with interactive, authentic communication opportunities and real-time feedback crucial for language acquisition. Tools like video conferencing, multimedia platforms, and online simulations enhance speaking proficiency by creating immersive, low-stress environments for practice (Shih, 2010; Golonka et al., 2014). Technology also strengthens TBLT through interactive, collaborative tasks that mimic real-life communication (González-Lloret & Ortega, 2014). However, in resource-constrained contexts, such as Angola, limited access to digital tools restricts the benefits of technology-driven TBLT, requiring reliance on traditional methods. Studies in low-resource settings (Brewster et al., 2002) suggest creative alternatives like realia, flashcards, and group work, which align with TBLT principles but lack the immediacy and authenticity of technology.
In this study, low-tech solutions such as computers, projectors, Bluetooth speakers, and smartphones, combined with peer collaboration supported vocabulary, grammar, and pronunciation development and low-stress environments for opportunities and confidence language practice. Yet, we believe this did not fully replicate the interactive benefits of technology postured above. Despite these limitations, TBLT outperformed more traditional methods, such as the PPP by prioritizing meaningful communication, fostering fluency and accuracy, and reducing anxiety to build confidence. While socio-economic variables were not explicitly controlled, their potential influence was acknowledged in this analysis. To mitigate disparities in English exposure outside the classroom, all students were provided with equal access to instructional materials and classroom-based resources. The study also recognised that variations in students’ linguistic environments could have contributed to differential outcomes, which were considered in the discussion of limitations.
In a nutshell, while technology significantly enhances TBLT, this study underscores the need for context-specific adaptations in resource-limited environments to replicate its benefits. Teachers should creatively address these constraints to optimize language learning outcomes.
7.5. Key Phases of the TBLT Intervention Plan for Addressing
Speaking Difficulties
The second research question explored the critical phases of the TBLT intervention designed to address Grade 12 English students’ speaking difficulties at Escola de Magistério Secundário no. 135 Cmdte. Liberdade do Lubango. Action research, as characterised by its cyclical and iterative nature, involves identifying problems, planning interventions, implementing actions, observing outcomes, and reflecting to inform subsequent cycles (Mertler, 2012).
7.5.1. Phase 1: Pre-Test Results and Initial Identification of Speaking
Problems
Long (2015) underscores the importance of needs analysis in TBLT as a means to address learner diversity and optimize instructional interventions. This study’s methodology and findings align closely with Long’s perspective, emphasizing the necessity of tailored approaches to meet both the linguistic and cognitive demands of learners. Research by Ellis (2006) stresses the importance to address both linguistic (e.g., vocabulary, grammar) and cognitive (e.g., task complexity) challenges in TBLT. In this study, through the pre-test, pre-questionnaire and direct observation, learners demonstrated significant challenges with linguistic, cognitive and instructional factors before the intervention. Both limited vocabulary or reliance on memorized structures, and poor pronunciation and task complexity likely impeded learners’ ability to communicate effectively. This echoes Ellis’s (2006) assertion that cognitive demands must be carefully calibrated to learners’ developmental stages and proficiency levels.
Low baseline scores (control group mean = 2.39; experimental group mean = 2.81) reflect not only linguistic deficits but also heightened affective barriers. Learners’ anxiety and reliance on rote memorization signal a restricted capacity to engage with authentic communicative tasks while reports of hesitation to speak and fear of making mistakes underscore the need for a supportive classroom environment. These findings resonate with previous studies (e.g., Tuan & Mai, 2015; Leong & Ahmadi, 2017) highlighting the interplay of emotional and linguistic readiness in second language acquisition.
Swain (2000) emphasizes the importance of providing learners with ample opportunities for both input (comprehension) and output (production) in language learning. The study reveals that learners’ restricted exposure to authentic English input impeded their ability to internalize and replicate target language structures Krashen (1985). Furthermore, learners’ reluctance to engage in spontaneous interaction may also reflect a lack of practice in producing the target language in the classroom and beyond controlled environments. Swain’s (2005) advocacy for output opportunities is critical here, as structured speaking tasks could scaffold learners toward greater fluency and confidence, especially in Angolan context where there are very little chances for students to produce the target language outside the classroom.
In short, these findings not only reaffirm established theories (e.g., Krashen 1985, Swain, 2005, Ellis, 2003, 2006) but also highlight the necessity of a tailored, multi-dimensional approach to TBLT that considers the unique linguistic, cognitive, and emotional profiles of learners.
7.5.2. Phase 2: Implementation of TBLT and Cycle I Results
The results of Cycle I in this study underscore the effectiveness of TBLT in addressing speaking challenges while simultaneously revealing areas for refinement in subsequent cycles. The experimental group’s mean score improvement to 3.78 highlights the positive impact of TBLT, consistent with theoretical underpinnings of the approach. The use of interactive tasks such as role plays, ordering and sorting, matching, comparing, problem solving, and discussions created a classroom environment conducive to language acquisition. These tasks, as suggested by Willis and Willis (2007), align with the principles of engagement, meaningful communication, and contextualised language use. Moreover, the alignment with Interaction Hypothesis reinforces the notion that interaction-driven tasks facilitate negotiation of meaning and language development through input and output opportunities.
The dynamic and low-anxiety environment observed during Cycle I aligns with Krashen’s (1985) Affective Filter Hypothesis, which posits that reducing anxiety enhances language acquisition. Learners’ active participation in tasks suggests that TBLT effectively mitigated barriers to interaction. This finding is particularly significant in contexts where speaking apprehension is common, as learners benefited from opportunities to practise language in meaningful and authentic scenarios. However, as observed in this cycle, the reduction of anxiety and increased engagement are not solely attributable to task design but also to the teacher’s ability to facilitate a supportive learning environment.
The findings from Cycle I provided interesting evidence of TBLT’s potential to enhance speaking skills and reduce learner anxiety. However, as highlighted by the iterative nature of action research, the challenges encountered offer critical opportunities for refinement and improvement. Therefore, this phase laid a foundation for refining the intervention, further discussed in subsequent phase.
7.5.3. Phase 3: Refinement of the Intervention in Cycle II
Cycle II incorporated refinements based on Cycle I feedback, emphasizing scaffolding, peer feedback, and collaborative learning. Scaffolding, informed by the Sheltered Instruction Observation Protocol (SIOP) model (Echevarria et al., 2023), included sentence starters and vocabulary banks, aligning with Vygotsky’s Zone of Proximal Development to help learners manage task complexity and enhance fluency. These adjustments reduced cognitive load and improved learner confidence in engaging with complex tasks. Following Cycle I, specific challenges were identified in relation to grammatical accuracy and pronunciation. Students exhibited persistent difficulties with tense consistency, syntactic variety, and complex sentence formation. Pronunciation-related issues were particularly evident in vowel production and word stress. In response, Cycle II introduced targeted interventions, including form-focused task, structured peer feedback, and pronunciation drills integrated into communicative tasks during pre-task and post-task stages. These strategies aimed to enhance accuracy without compromising fluency.
Real-world task relevance, as emphasized by Liu (2018), increased engagement by focusing on meaningful applications rather than isolated structures. Collaborative learning and peer feedback, also reported by Albino (2017) and López (2021) studies, helped students to foster negotiation of meaning and interaction within a supportive environment, enhancing fluency and building a sense of community. During the intervention it was evident that learners seemed to be aware of the need for peer support for successful acquisition for all; this was more evident in the planning and reporting stage of the task phase. Furthermore, the mitigation of low-tech solutions such as computers, projectors, Bluetooth speakers, and smartphones, combined with peer collaboration led to significant learning.
Cycle II incorporated a range of multimedia resources, including recorded dialogues, instructional videos, and encouragement for language learning applications use, such as Duolingo, American English, Babbel, Rosetta Stone and Hello Talk. The inclusion of such materials was intended to provide students with exposure to authentic linguistic input, facilitate self-paced learning, and promote interaction with diverse accents and speech patterns. Video-based speaking tasks, in particular, played a key role in improving pronunciation and intonation, as students engaged in drilling exercises and interactive role-plays.
This phase revealed how responsive adjustments to the intervention plan through TBLT could address linguistic, cognitive, and affective challenges. The study reinforces the value of reflective and flexible teaching practices, contributing to discussions on effective TBLT implementation in resource-limited contexts like Angola.
7.6. Impact of TBLT on Speaking Skills
This study explored how implementing TBLT improved the speaking skills of 12th grade English students at Escola de Magistério Secundário no. 135, Comandante Liberdade. Data from pre-tests, post-tests, questionnaires, and observations revealed significant improvements in linguistic, affective, and instructional aspects of speaking. The reduction in anxiety and increase in confidence, alongside the improvements in vocabulary use and pronunciation, the inclusion of technology, especially listening materials to facilitate learners’ comprehensible input, demonstrate that task-based learning may be an effective pedagogical approach in enhancing speaking skills (Ellis, 2003; Nunan, 2004; Jeon & Han, 2006; Jasim, 2011; Long, 2015; Mahdavirad, 2017).
The study systematically measured improvements in five major components of speaking skills: fluency, pronunciation, grammar, vocabulary, and coherence. Fluency, gauged through speech rate, frequency of pauses, and hesitation phenomena, showed the most substantial progress, with participants demonstrating faster, more confident delivery after the interventions. Pronunciation, assessed at both the segmental (phoneme production) and suprasegmental (intonation, stress, rhythm) levels, improved notably, though traces of foreign accent remained detectable, as expected in second language acquisition contexts. Grammar was evaluated in terms of syntactic complexity and error rates; findings revealed a clear reduction in basic grammatical mistakes but suggested that full mastery of complex structures may require longer-term exposure. Vocabulary development, measured through lexical diversity and appropriateness, expanded modestly, with gains primarily in thematic areas addressed by the tasks, highlighting the contextual sensitivity of lexical growth. Coherence, assessed by the use of discourse markers and logical sequencing, improved significantly, supporting the argument that task-based learning fosters better discourse management. While improvements were recorded across all domains, the uneven nature of gains—particularly the slower advancement in advanced grammar and wider lexical range—points to the need for sustained, longitudinal interventions to consolidate higher-order speaking skills.
When contrasting the findings of the current study with previous research on enhancing oral fluency and speaking skills through Task-Based Language Teaching, both similarities and differences emerge, particularly in relation to how TBLT strategies were applied, measured, and the outcomes they produced.
Like previous studies (e.g., Jeon & Han, 2006; Jasim, 2011; Carless, 2009, 2012; Albino, 2017), the current study also confirmed the effectiveness of Willis’ (1996) three-stage task-cycle in fostering improvements in oral fluency. The sequence of task, planning, and reporting stages played a crucial role in providing students with ample opportunities for authentic communication, reflection, and feedback whether in pairs, small groups or in front of the class. These stages were observed to improve students’ ability to formulate their thoughts more fluently as they progressed through the tasks, a finding consistent with earlier research.
Moreover, the post-task stages (Hung, 2014; Albino, 2017; López, 2021), which included analysis and practice phases, further consolidated the gains made during the task-cycle. These stages allowed students to analyse their language use and refine their output (Long, 2015, 2016), thus enhancing both fluency and accuracy—outcomes similar to those observed in previous studies (Albino, 2017; López, 2021 and Murad, 2021) where students benefited from the reflective practice incorporated into the post-task stages. Furthermore, in both the current study and previous research, the use of real-life, interactive tasks emerged as a core element in enhancing oral fluency. Studies by Jeon and Han (2006), Jasim (2011) and Carless (2012) emphasised that task-based learning provides an environment where students are encouraged to engage in meaningful language use. Similarly, in this study, learners reported that the authenticity of tasks helped reduce anxiety and encouraged active participation, leading to measurable improvements in fluency.
These findings support the idea that tasks designed around real-life scenarios promote natural language use and improve speaking skills. This study, consistent with prior research (Carless, 2012; Albino, 2017), observed significant affective improvements, including reduced anxiety and increased confidence, highlighting TBLT’s role in enhancing motivation and fluency by fostering a supportive, low-stress environment. A distinctive feature of this research was its use of pre- and post-questionnaires to identify learners’ initial concerns—fear of mistakes, low self-esteem, and limited practice opportunities—enabling a more tailored and responsive intervention to address these specific needs.
This study distinguishes itself by addressing both cognitive and affective dimensions of speaking, prioritising anxiety and confidence alongside fluency and accuracy. Unlike prior research (e.g., Jeon & Han, 2006; Jasim, 2011; Carless, 2012; Albino, 2017), which often overlooked systematic affective diagnostics, this study employed pre- and post-questionnaires to identify and address learners’ concerns, resulting in tailored TBLT tasks that targeted specific barriers like fear of judgment and low self-esteem, vocabulary limitation, pronunciation issues and the influence of students’ L1. Moreover, the study integrated quantitative and qualitative data, using pre-test/post-test scores alongside questionnaire insights to evaluate changes in learners’ confidence and anxiety levels. This holistic approach offered a broader evaluation of TBLT’s impact, contrasting with the predominantly quantitative focus of previous studies.
Contextually, the Angolan classroom added a unique socio-cultural dimension, with customised tasks addressing classroom dynamics and public speaking fears— factors less emphasised in studies conducted in more immersive and developed educational settings. Collaborative tasks and scaffolding played a significant role in reducing anxiety while fostering peer support, reflecting a stronger focus on continuous emotional development than in earlier research.
8. Conclusion
The implementation of the TBLT framework significantly improved students’ oral fluency, particularly in the experimental group. Post-test and post-questionnaire results and observations revealed increased proficiency in confidence, fluency, coherence, and grammatical accuracy. Students responded positively to the task-cycle stages, with meaningful language use observed during the task stage and enhanced organisation and presentation skills developed in the planning and reporting stages. The post-task phase, including analysis and practice stages, proved effective in fostering reflective thinking. Students analysed their performance, addressed areas for improvement, and achieved greater mastery of communicative skills and accuracy. The intervention positively impacted students’ confidence and motivation to speak English. While pre-questionnaire results indicated initial anxiety and lack of confidence, post-questionnaire findings showed a significant improvement in comfort and willingness to engage in communicative tasks. The study identified a gap between students perceived and actual challenges in speaking English, particularly regarding fluency and coherence. These challenges were addressed during the intervention, emphasising the need for ongoing diagnostic assessments to align instruction with students’ true needs.
9. Limitations of the Study
Despite the positive outcomes, several limitations were identified in the study. Firstly, the small sample size limits the generalisability of the findings to other Grade 12 English major classes in Angola. While the results offer valuable insights, caution is needed when applying these conclusions to broader contexts. Secondly, the short time frame of the intervention, coupled with disruptions such as nationwide teacher strikes, may have restricted the full potential of the TBLT approach, warranting further investigation in longer-term studies to assess its sustained impact. Thirdly, external factors, including socio-economic challenges, limited access to resources, and varying levels of exposure to English outside the classroom, were not controlled for and could have influenced student engagement and learning outcomes, suggesting the need for future research to address these variables for a more nuanced understanding of TBLT’s effectiveness. Lastly, although the study did not explicitly disaggregate findings by demographic variables such as gender or socio-economic background, observational data suggested some variation in engagement and performance. Female students, for instance, appeared to participate more actively in communicative tasks, which may be attributed to broader socialisation patterns, a factor that needs further investigation in future research.
10. Implications of the Study
To enhance English speaking skills in Angola, in secondary schools, adopting task-based language teaching (TBLT) with real-life tasks can cultivate functional speaking skills. Curricula should emphasise debates, presentations, and interactive activities to enhance communication abilities. Targeted exercises focusing on pronunciation and fluency should be integrated into lessons, and periodic speaking assessments implemented to monitor progress and refine teaching strategies. Increasing exposure to native-like English through videos, guest speakers, and exchange programmes can further strengthen students’ communicative competence and confidence in the language. Collectively, these measures aim to establish a strong foundation for English speaking skills among learners, promoting their success in both academic and real-world contexts.
Moreover, schools should foster language-rich environments through initiatives such as English clubs, debates, and events that encourage regular language use. Professional development for teachers is essential to equip them with effective task-based methods for teaching speaking. Students should have access to diverse English-speaking resources, such as podcasts and role-play materials, and are encouraged to enrol in English clubs and utilise social media to connect with new friends abroad, interacting in English on a weekly basis as a means of exposing themselves to and challenging their use of the target language.
A true TBLT course, conversely, requires an investment of resources in a needs analysis and production of materials appropriate for a particular population of learners—Long (2015) Second Language Acquisition and Task-Based Language Teaching.