EXAMINING THE EFFECT OF TASK-BASED LANGUAGE TEACHING ON UNIVERSITY BUSINESS STUDENTS IN BANGKOK

The purpose of this study was to introduce Task-Based Language Teaching in a bid to energise stagnant English language performances among second-year business students (N = 81), studying a private university in the peripheral areas of Bangkok. Students followed a TBLT learning environment for one semester (16 weeks), and subsequent end-of-term performances were compared with prior attainments achieved under tradition forms of instruction using t-tests (0.05). Overall, TBTL proved to positively influence performance outcomes when compared to the traditional method (TBLT: 60.9 = Grade C+; TRAD: 54.93 = Grade C; p [0.0195] = sig <0.05). Nonetheless, the majority of progress derived from enhancements in speaking skills, as assessments in this domain improved significantly, whereas no significant difference was observed in formal examinations. Further to this however, variability analyses highlighted that upper quartile students significantly improved in both speaking and formal examinations, while the lower quartile cluster failed to show noteworthy forms of progress in speaking, and, formal examination scores exacerbated entirely; concluding therefore that learners’ response to TBLT is governed by their linguistic potential. The recommendation therefore would be to arrange two separate groupings based on ability for two reasons: (1) assist accelerated acquisition of more proficient students who thrive in a TBLT environment, and, (2) to provide curricular support for struggling students for whom TBLT is not (yet) developmentally appropriate.


INTRODUCTION
The present study examines the instructional impact of task-based language teaching on a sample of 81 university business students.Given the timid level of descriptive details provided in prior literature, this paper comprehensively presents details of TBLT research design and lesson structures, and successively, the results from are analysed by a multitude of cross-sectional analyses, supported by in-depth discussion to determine the effectiveness of TBLT on learners of all proficiency levels.
The fundamental research questions this paper aimed to address the impact of TBLT on students" overall English attainments, and whether TBLT can contribute to improving both speaking skills and formal assessments among both male and female students of all ability levels.

Task-Based Language Teaching
TBLT has its origins in Dewey"s view that learning occurs through experiences that relate to the interests of the learner (Samuda & Bygate, 2008), as an alternative to traditional educational settings, in which most of the opportunities for language use are taken by the teacher (Willis, 1996).Learners feel the need "to add physical action to their learning … to experience the new knowledge in ways that involve them better (Lightbown & Spada, 1993), and the concept of task-based language teaching has been touted to maximise student involvement and language use, as tasks themselves also remove teacher domination (Willis, 1996).Nunan (1989) defines a task as a "a piece of classroom work which involves learners in comprehending, manipulating, producing or interacting in the target language while their attention is principally focused on meaning rather than form".For which, Ellis (2003) established a set of criteria that must be met for an activity to constitute a task, where language serves as a tool for communicating and students move away from being language learners and are positioned to become "language users" (Ellis 2001).
First of all, as mentioned by Nunan (1989), the focus must be centred on meaning, as opposed to specific linguistic forms.Secondly, acquisition is optimised when learners encounter a "gap" (Ellis, 2009), which is a situation that necessitates conveying meaning in the form of an opinion, a contribution, an argument or to impart information.This nature of negotiation forces learners to stretch their interlanguage (Nunan, 1989), and incites individuals to take risks in experimenting with new language structures; increasing pushed  (Ellis, 2009).Thirdly, students must resort to their own linguistic and non-linguistic resources to accomplish the task; and (fourthly), there is a specific outcome separate from the use of language, as the language serves as a means for achieving that outcome, rather than an end product.
Educators may misinterpret a task as a form of situational grammar exercise given the degree of overlap, although both concepts are clearly separable and the latter does not require the learner to negotiate meaning, but rather demonstrate a correct understanding of a given linguistic feature.Furthermore, the learners are explicitly made aware of the language-related objective of the situational-grammar exercise, whereas in a task they are not (Ellis, 2009).This does not stipulate that tasks cannot focus on certain grammatical structures; they can, provided that all four afore-mentioned criteria are satisfied.Tasks that are more grammar orientated are referred to as "focused" task, the clear distinction however is that learners are not explicitly informed of the linguistic feature that will form the basis of the task, this remains "hidden" (Ellis, 2009).
A syllabus comprising unfocused tasks however constitutes a range of tasks to be completed (Prabhu, 1987), rather than forms to mastered.An additional distinctive feature is that tasks can be "input-providing", involving reading and listening skills, or, tasks can be "output-prompting", which engages students in speaking and writing (Ellis, 2009); although most tasks tend to be integrative, involving two or more skills.With respect to the structure of tasks, there is no single rigid paradigm to which TBLT practitioners must adhere.The vital component of a task-based lesson is the main-task phase, although additional phases may be included in the form of pre-task and post-task activities.

Task Methodology (Pre-Task, Main Task And Post-Task):
Pre-tasks typically take one of four forms: (1) performing a task similar to that of the main task; (2) observing a model to help understand how to perform the task; (3) creating non-task activities to prepare learners for the task; or (4) setting a time limit allowing participants to strategically plan for the main task.
In the first instance, students may participate in the initial task as an interactive class-based activity (Phrabu, 1987) (Skehan, 1996;Willis, 1996).
And thirdly, non-task activities contribute to activating learners" content schemata, and when familiar with the requirements of the task, more processing space becomes available for formulating the language needed to express ideas, also contributing to enhancing fluency and complexity of language used.This is manifested in Newton"s ( 2001) argument that such activities will "prevent the struggle with new words overtaking other important goals such as fluency or content learning"; such activities may constitute brainstorming or mind-maps (Willis, 1996).Lastly, teachers may allow students to strategically plan their approach to a given task, for which Skehan (1998) proposes that 10 minutes is optimal.
Main tasks entail both taskperformance and process options to optimally manage task implementation.The former comprises options relating to how the task is planned by the teacher, who may wish to impose a time pressure placing the emphasis on fluency, whereas, the absence of time restrictions will help enhance accuracy.The second task performance option may involve allowing students to access the input designated for task-related activities, which will naturally reflect in greater accuracy than if such input was not made available (Ellis, 2009).Thirdly, the teacher may contemplate adding a surprise element into the lesson; such spontaneity will force students to reformulate ideas and language use.
On the other hand, process options involve live decision making in the handling of tasks, especially with respect to corrective feedback.This could be achieved by addressing errors during task time where the teacher pauses the task momentarily to attend to form (Ellis, Basturkmen and Loewen 2001), or in-task correction may occur incidentally (Prabhu, 1987), and is addressed when appropriate, intervening to support a process initiated by the learner, possibly in the form of a recast or metalingual comments, a technique referred to as nudging (Lynch, 1997).
Post-tasks encompass three major pedagogic goals.Firstly, to provide an opportunity to repeat the task either under identical or modified conditions to reinforce mastery.Or secondly, students are invited to reflect on task performance to discuss communication related difficulties and problem-solving issues in order to enrich subsequent performances, for this Allwright (1984) coined the term uptake.Thirdly, the teacher may allocate attention to forms that were problematic during in-task performance that they failed to use naturally (Loshcky & Bley Vroman, 1993).

TBLT methodology in practice:
Skehan (1998) introduces a pre-task phase prior to the main task, which is allocated to more explicit instruction and form correction.However, in this model tasks are entirely unfocused and do not integrate post-task activities.Long (1985) on the other hand incorporates both focused and nonfocused tasks and also provides corrective feedback when required, but does not include a pre-task or post-task phase; whereas Willis (1996) proposes that attention to form best reserved for the post-task phase.
Ellis (2003) alludes to the inclusion of all three components and paying attention to form during all phases of the task.Similar to Long, Ellis introduces both focused and unfocused tasks, but dissimilar to both Long and Skehan, Ellis does not necessarily reject the ideology of traditional forms of instruction to supplement understanding (2009).Nonetheless, despite this variability, all three versions of TBLT bear five essential similarities: (1) all three variations provide natural opportunities for language use, (2) they are learner centred, (3) tasks are either focused or unfocused, (4) attention is paid to form is when considered appropriate, and (5), all approaches reject traditional forms of instruction (as a central ideology).The structure of TBLT utilised in this paper will be detailed in the methodology section.

Criticisms of TBLT
Widdowson ( 2003) claimed to identify the structural weakness of TBLT on the grounds that criteria is too loosely formulated and that tasks are not distinguishable from more traditional classroom activities, and, that tasks neglect semantic meaning.Seedhouse (1999) argues more explicitly that a task is not a valid construct for language learning, asserting that TBLT will only result in the production of impoverished language samples of minimal acquisitional value (pidginised language), and that the over-reliance of context will promote fossilization.Seedhouse (2005) also added that the actual production of language that arises from a task is very unpredictable and disparate to the intended language.Sheen ( 2003) accused the TBLT of not possessing a grammar syllabus, and Swan (2005) was more condemning with his comments, claiming that TBLT "outlaws" grammar in its syllabus.
However, in response to these claims, Ellis (2009;2005)  not accounted for the fact that TBLT can comprise both input-prompting and output-providing tasks, which also exposes students to high quality input (through text or audio).In addition, as discussed a priori, not all tasks are unfocussed, a misconception on which these criticisms must have been founded.
With careful planning, focused tasks can ensure a close match between the intended language features selected, and those that learners process when performing the task.A task-based syllabus is not mandated to solely comprise a pure task structure, a hybrid of both focused and unfocused tasks can be considered.Ellis (2009) acknowledges the use of traditional forms of teaching if implemented carefully and sparingly alongside TBLT, primarily in the form of conscious raising tasks which target and address confusing structures to rectify misunderstandings and reinforce accuracy.
Below is a compilation of studies published relatively recently and derived from more truthful sources, emanating from three separate countries active in EFL; Cyprus, Albania and Indonesia.

Eastern Mediterranean University -Cyprus (analysed students' reactions to TBLT)
The leading study for discussion was published in the Asian EFL journal (volume 9, issue 4, 2007) and was conducted at the Eastern Mediterranean University in the Turkish Republic of Northern Cyprus.The study looked to evaluate the effectiveness of implementing TBLT to ascertain the students" reaction to TBLT and their respective opinions in comparison to traditional forms of instruction.The study included 54 first-year students from two separate classes at the ELT department of the university.The groups were demographically diverse and came from six nations, including Turkey, Turkish Republic of Northern Cyprus, Jordan, Israel, Kuwait and Pakistan; and were aged between 17-23 years.To generate data, a mixed methods technique was used consisting of questionnaires, interviews and diaries.While questionnaires were distributed to all students, interviews and diaries however only involved four select students.Questionnaires comprised a total of 26 questions utilizing a 5-point scale.The first ten questions related to traditional syllabus and the following 16 questions asked students to rate the experiences of TBLT.The means of the two sets of questions were cross-compared via t- The results indicated greater levels of engagement and enjoyment during the TBLT phase compared to that of a traditional setting.The diaries, recorded by students A, B, C and D (held in anonymity), helped to identify common perceptions and supported reasoning.The diaries apparently revealed "great satisfaction" on the grounds that students enjoyed receiving and giving presentations, as well as the variety of tasks introduced to them, offering more opportunities to speak.The same four participants in the interviews expressed their dissatisfaction with traditional style lessons due to their limited role as a listener, and the teacher was hijacking the vast majority of the talk-time.They also stated that the course book was not pertinent to their interests and the exercises were of disinterest, as most of the content constituted lengthy passages followed by a repeatedly identical nature of exercises.

South East European University -Albania (evaluated the impact of TBLT on speaking skills)
An alike study was conducted at the South East European University (Albania) and examined the effectiveness of task-based learning in developing students" speaking skills in an EFL setting.In line with the study above, the paper also investigated students" and teachers" reaction to the implementation of TBLT style lessons.The research projects comprised 60 undergraduate students between the ages 18-25 who were working at preintermediate / intermediate level; six members of the teaching staff also took part.
All participants, teachers and students, received questionnaires with a Likert scale format ranging from 1-5 (strongly disagree to strongly agree).The students were divided into two groups, a control group was taught conventionally, and the experiential group received TBLT instruction for a total of 8 weeks in the winter semester of 2012.
A pre-test was conducted to ascertain the students" current level of spoken English, consisting of a 3minute presentation in which they introduced themselves to the class and discussed their interests.Subsequently, after the task-based programme, students were subject to a post-test to ascertain the effectiveness of the taskbased approach.The findings indicated that the students responded positively to talk-based learning on the grounds that their attention is focused on tasks, which optimises their potential to learn.The study also adopted a mixed methods research approach which included the collation of quantitative data through speaking test scores (post-test), and qualitative data was acquired via interviews and field notes.

IJEE (Indonesian
The purpose of the study was founded on the observation that, despite many years of learning and demonstrating a reasonable knowledge of grammatical structures, the problem remains that students appear timid, lacking in fluency and unable to communicate in real-life scenarios.This study sought to introduce a method of teaching that would equip students with the confidence and skills needed to communicate effectively in true-tolife situations; putting more emphasis on meaning more so than form.Out of a total of 68 students on the programme, 40 participated in this study, selected through random sampling.
The teacher talk-time was mostly allocated to brainstorming ideas, and not to introducing a topic through monologue.The teacher proceeded to monitor progress and to ensure the target language was being utilised to discuss ideas within and between groups during jigsaw activities, roleplay or presentation preparation time.The qualitative data also suggested that students" higher levels of performance were owed to a more relaxed and collaborative atmosphere, and also because tasks were pertinent to practical situations; such as, greeting friends, asking / giving information, agreeing and disagreeing, asking / giving suggestions and describing people.
The 40 students taught by taskbased learning were reported to have shown a significant improvement in terms of their speaking skills compared to the 28 that did not.The mean score of the TBLT group was 79.18, which was considered to be significantly superior than those taught conventionally (74.22).
The research papers hitherto discussed tend to lack fundamental detail.First of all, in all forms of research, findings centred purely on speaking skills, which were reported to have significantly improved as a result of task-based learning, nonetheless, the issue that remains is the lack of statistical instruments, reporting and Secondly, the studies that conducted qualitative data did not appear to have any specific algorithm in place to code responses and interpret data, rather the results were imparted and concluded in anecdotal form.
Thirdly, very little data was displayed in relation to performance variations within the parameter of demographic profiles, or according to ability level.Fourthly, the descriptions regarding the tasks that were implemented in this trial were seemingly vague; a key design feature that was not sufficiently explicated.
Based on the review of related theory and prior studies, within the context of the current study, TBLT is expected to significantly improve students" performance in both forms of English Language assessment (speaking and formal examination); irrespective of gender or ability.

METHOD
This study examined the impact of task-based language teaching methodologies on 81 second-year business studies students, by comparing students" formal examination results as a result of this TBLT trial vis-à-vis attainments of their previous course, obtained under traditional instruction.

Research design
Virtually all activities prescribed in the existing curriculum focused on lower order skills, such as choosing correct answers, identifying irregularities and correcting structures; which was considered pertinent to the current level of the students (elementary / pre-intermediate).This trial revamped the syllabus to promote, develop and sustain interactive language use.Below is an example of Unit 1 from the standard format of the syllabus.

Original content for Unit 1: Introductions and holidays:
-Grammar drills for auxiliary verbs (gap fill and word select)

Pre-task phase
Within this curricular remodelling, reading activities were not omitted form the syllabus, but were converted into "read and do" tasks (Ellis, 2009), and the conversion into an interactive task involved three steps.First of all, as a warm-up, students were presented with the text and answered the conventional true, false, NG response format.When finished, answers were discussed as a whole-class activity followed by unexpected extension questions (which required full sentence structures); these were read out by the teacher.
Formed into small groups, students" listening and speaking skills were brought into use as they were required to understand the questions, search for the answers embedded in the text, and reply independently in full sentences.For struggling groups, the teacher could write the extension questions on the board and allow students time to transcribe their responses.

Main task
Based on the theme of text discussed, the teacher introduced a series of questions to initiate and sustain a conversation relative to holiday experiences, eliciting answers from groups and coaching students to respond on in full sentences, encouraging peers to collaborate and help struggling classmates.
The questions were in the present simple tense the first two were discussed as a class: How often do you go on holiday?Where do you like to go? Subsequently, the following questions were introduced and based on which, students were asked to create a roleplay: Who do you usually go with?Why do you enjoy going there?How do you get there?How long do you stay there?What do you normally buy?How much does it cost?What bad experiences can you have on holiday?Students could modify, add o remove questions, whilst the teacher visited the groups individually to monitor the progress and facilitate participation.
The grammatical content of the unit (auxiliaries, connecting words, present and past simple tenses) were not explicitly presented to the students in a traditional sense, and remained "hidden" (Ellis, 2009).When completed, students were instructed to project their dialogues into the past.To clarify Where did you go?).Students were free to utilise both past simple and present perfect forms, amend or add further questions to render the conversation as free and natural as possible.During the task, the teacher checked students" progress and understanding to verify use of tenses, correct usage of auxiliaries, connecting words question tags and vocabulary items.Depending on the nature of the unit and the teacher"s discretion, this second part of the main task could be performed privately in groups, in front of the class, or as part of an informal speaking assessment.An additional section of the main task involved a discussion in pairs or groups to review several city profiles and decide which destination interests them the most, and why, while also stating reasons for discarding the less appealing locations.

Post-task phase
The grammar intensive exercises were not forsaken altogether, given potential instructional value if used appropriately and kept to a minimum (Ellis 2001).In this case, they served as a consolidation phase to reinforce the students" understanding of regular and irregular verbs and auxiliaries in present and past forms and successfully employed in the context of a full sentence.Time allocated to this phase of the lesson was typically 20 minutes, but varied depending on the completion time of the main task.

Participants
The participants included in this study consisted of 81 second-year business students, which were relatively well-balanced demographically (females = 48, males = 33).Business students were invited to partake in this trial given the importance of English to their academic discipline.The assortment of students into groups was entirely at random and were not categorised according to any particular criteria.Given the nondiscriminatory sorting process, English attainments across groups were not analysed individually.Furthermore, prior attainments were the control element of the experiment, thus the inclusion of specific control groups was not deemed appropriate.

Measures
The formal assessment criteria of the university"s English language courses consist of one speaking test (40%) and two formal assessments (60%); mid-term tests (20%) and final examinations (40%).

Data analysis
This study used t-test significance testing (0.05) to compare current performances achieved under a TBLT methodology with prior attainments achieved in a traditional setting.In addition, standard deviation was also analysed to ascertain the degree of consistency in both performances.Further to which, variability analyses were carried out to ascertain the extent to which all tiers of students respond to TBLT; (Q1: top 20 students; Q 2 & 3: 41 mid-tier students; Q4: 20 lowest performing students).

FINDINGS AND DISCUSSION
The opening set of research questions concerned the general impact of TBLT on elementary / preintermediate business students" language attainments, and, whether TBLT can help improve performances in both speaking and formal examinations.
Table 1 shows that TBTL generally had a positive impact on performance outcomes when compared to the traditional form of instruction (60.9:Grade C+ vs. 54.93:Grade C; p 0.0195 = sig <0.05).Nonetheless, upon closer inspection, it becomes clear that the majority of progress originated from developments in speaking skills, as assessments in this domain improved significantly (p = <0.01)compared to formal examinations, where no significant difference was observed (p = .324).Furthermore, higher levels of standard deviation were also noted in the TBLT results, inferring greater variance in language attainments.The second research question in this paper was to analyse performance variations on the basis of gender.
The results presented in Table 2 imply that neither gender technically responded significantly to TBLT, despite the results displayed in the previous table.
Although, male students" improvements did reach near statistical significance given that the value of "p" (.0611) almost fell below the 0.05 threshold.Furthermore, akin to Table 1, both genders showed more progress speaking skills than in formal examinations, and, higher achievements among male students in a TBLT environment contributed to narrowing the achievement gap between genders.
The analyses of the following data set will examine the variability in English language performance to ascertain whether TBLT benefits both above and below average learners.3 is that the most capable students represented in the upper quartile performed significantly better in the TBLT trial in both forms of assessment: speaking (p = <.01) and formal examinations (p = .002).Conversely, the lower achievers (Q4) performed significantly worse as a result of TBLT (overall: 35.55 vs. 40.5;p = .0139),as speaking skills did not demonstrate noteworthy improvement (p = .313),and formal examination results significantly deteriorated (p = 0 <.001).However, mid-tier students" language performance (Q2 & Q3) was more varied.The overall performances improved significantly (p = .004),especially in speaking (p = <.001), but formal examinations did not show significant signs of progress (p = .137).

Discussion
The implementation structure of TBLT in this trial aimed to increase students" fluency, confidence (especially with risk-taking), and eventual improvements in the accuracy of language produced.To this end, the trial based lessons on textual input (reading), which was followed by communicative extension activities and consolidation exercises.Most students responded well to the introduction of tasks, especially the more proficient students.Nonetheless, lower achieving students" attainments exacerbated in a TBLT environment, which, (within the context of the lower performing group), partly validates Seedhouse"s (1999) supposition that TBLT will only result in the production of impoverished language samples, promoting fossilization.
The shortfall among the lower achievers also gainsays the claim that in collaborative environments, senior students assist struggling peers to reduce imbalances, the core construct of a theory known as the more knowledgeable other (Vygotsky, 1978).The purpose of this discussion is to identify solutions that may benefit both more capable and less responsive students in the context of EFL.
Owing to the limited levels of English proficiency of Q4 students, compounded with a lack of familiarity with TBLT, struggling learners may require more support from the instructor seeing as their level of progress declined during the TBLT trial.Therefore, the first suggestion may consist of dividing students into two separate groups (group One: Q1 & 2; Group Two: Q 3 & 4).The central justification for this proposal is based on the observation that considerable class time was allocated to the constant rectification of misunderstandings among embattled students, which Taking into account the limitation of pedagogical resources, a case could also be made that two separate classes of a shorter duration tailored to each group would be more productive than one single class comprising mixed ability students working at different paces.The division into groupings would enable higher achievers to work on tasks freely and thrive under the autonomy, while Q3 and Q4 students may profit more from task-supported learning from worked examples and greater teacher intervention.
The second element of this suggestion directly relates to Sweller et al."s findings (2012) that discoverybased learning is not productive for novice learners, resulting in the discovery often being missed.Sweller (1985) proposed that individuals learn by studied examples, (worked-example effect), and gradually transition to a more autonomous learning environment (guidance-fading effect).This also reduces the cognitive load during skill acquisition and enhances the learning process; (this) "is one of the earliest and probably the best known cognitive load reducing technique" (Paas et al., 2003).Subsequently, with gradual guidance removal, learners increase their possession of schemas or partial schemas to be applied in problem-solution based tasks introduced at a later stage (Kalyuga, Chandler, Ayres and Sweller 2003).
Furthermore, the separation of classes into separate groups would also address an additional limitation noted in this study, which was the relatively large class sizes (20 students on average), and in relation to which Ellis (2009) highlights that TBLT is not easily implemented in large classes, an unfavourable structural feature typical of many educational settings.

CONCLUSION AND SUGGESTIONS
The results in this trial largely support the idea that TBLT lessons help to promote fluency, owing to the creation of more opportunities to practice verbal output and the resulting increased levels of confidence.Inconveniently however, Sweller"s (1985) remark that novice or less-able learners often fail to make the discovery also holds true.Therefore, the primary inferential relationship highlighted in this study is that the students" linguistic potential appears to govern the likely level of progress made in TBLT learning environments.
In response to this observation, the central recommendation would suggest arranging students into two separate (1) to assist accelerated acquisition of more proficient students who thrive in a TBLT environment, and, (2) to provide curricular support for struggling students for whom TBLT is not (yet) developmentally appropriate.Successive research may wish to contemplate analysing the effect of a variety of methodologies of TBLT structures across a large number of groups as an experiential project.The groupings could be formed as follows: -Group 1: Traditional instruction or PPP -Group 2: Long"s (1985) TBLT model consisting purely of a main task (focused).-Group 3: Long"s model comprising purely unfocused tasks.
-Group 4: Skehan"s (1998) method of a pre-task to support the following main task -Group 5: Ellis"s (2003) suggestion of implementing all three phases The first cycle of the experiment (comprising all groups) could consist of mixed ability groupings, and the second cycle could test the five separate models on high performing students and lower achievers separately.This would help confirm which method of implementation (including traditional instruction) is most effective for all tiers of ability.
Extending the scope of research to encompass more universities in Thailand, and more importantly, universities in other countries active in EFL would enrich findings, also helping to identify and compare the influence of cultural settings on the productiveness of TBLT learning structures.

Table 2 :
Performance variations according to gender

Table 3 :
Analysing the variability of English language attainments