The effectiveness of task-based language teaching to improve students’ speaking skills

- This study aims to find out the effectiveness of Task-Based English Teaching (TBLT) to improve college students’ speaking skills. Teaching speaking is considered to be difficult as there are various factors involved, both from the teachers’ part and the students’ as well. TBLT uses authentic sources of teaching materials with tasks as the essence of the teaching instructions. The research was done in a private college in Bogor. Sixty of the 1st-semester students of the non-English department are the subjects of the research. They were divided into two groups: the control group and the experimental group. Each group consists of 30 students. The control group was taught using a more conventional method however the experimental group was taught using Task-based instructions. A pre-and post-test was administered to both groups to find out the effectiveness of TBLT to improve their speaking skills. The mean scores of pre-and post-tests of English speaking were tabulated using SPSS to identify whether there was a difference in students’ achievement of both groups and whether the difference was significant or not. Questionnaires were also distributed to find out their perspectives towards the English lessons. The results of the tests were supported by the students’ answers in the questionnaires in which all students of the experimental group showed significant improvement in their speaking skills in terms of accuracy and fluency. Their perspective after experiencing TBLT in the English lessons also demonstrates a significant change resulting in students’ higher self-confidence in speaking. speaking ability. This study aims (1) to examine the effectiveness of using TBLT on non-English major college students. (2) to examine the effectiveness of using TBLT on the students’ perception of task-based instructions in speaking class.


Introduction
Teaching speaking has a distinctive role in any kind of language teaching. In the past, as studying speaking was complicated, teachers and language scholars put more focus on written than spoken language. Two decades ago speaking began to be included in teaching, learning, and testing. However, the emphasis was not on the production of spoken discourse (Cater & Nunan, 2001).
Indonesian students, including college students, are facing the same difficulties in learning English. This can be seen in their achievement in the English examination. Their scores show that they have very low proficiency in all English skills, particularly in oral English. They are usually uncertain to use oral English because they find it very difficult in using accurate, fluent, and complex expressions. Moreover, they seem to lack interest in learning English. Speaking as a basic skill of communication is an important element in determining whether a student's English skills have reached success, particularly in college. The students who attend a speaking course are expected to communicate well in English, practice doing it many times, and are not afraid of making any kinds of mistakes or errors. The success of students' speaking skills is not only influenced by their styles of learning but also methods or strategies organized by the lecturers in the classroom.
Among the ways to create a supportive language learning context, Task-based Language Teaching (TBLT) offers a lot of opportunities for students to practice and improve their speaking skills. A lot of research shows that TBLT is a refreshing approach in language teaching. It is considered refreshing as students are given opportunities to experience the language uses in pairs and group work which let them share ideas in the classroom (Nunan, 2004).
The practice of TBLT has been flourishing around the world (Murad, 2009). Numerous researches on this topic have been reported in the scope of Asian countries. Unfortunately, it is identified that the study regarding TBLT in the Indonesian context is still limited (Ismail, 2017). This research is interested in finding out the effects of TBLT to improve non-English major college students' speaking ability. This study aims (1) to examine the effectiveness of using TBLT on non-English major college students.
(2) to examine the effectiveness of using TBLT on the students' perception of task-based instructions in speaking class. The Nature of TBLT Task-based Language Teaching (TBLT) was proposed by Prabhu in 1987. He used task-based instruction with secondary school classes in Bengalore, India, in his communicational project which began in 1979. TBLT is basically an approach that is based on the assumption that accuracy is acquired after fluency or after successful communication (Willis, 1996). This approach provides an environment that best promotes natural language learning (Foster, 1996). Students are given more opportunities to use the target language to express their ideas in which their attentions focus on meaning. This is executed through learning activities or tasks that are designed to encourage students to practice the functional use of language in a meaningful context. Hence, students not only learn grammar but also other language features which will help them to communicate. When students are asked to describe their daily activities, for example, they learn to produce good sentences with accurate tenses. In this case, task-based learning acts as a stimulator to motivate students to improve their language skills by performing meaningful tasks. Defining Tasks in TBL A task is an activity that requires learners to arrive at an outcome from a piece of given information through some process of thought, and which allows teachers to control and regulate that process (Prabhu, 1987). Tasks are always activities in which learners use the target language for communicative purposes. Lee (2000) defines a task in a more specific way. It is a classroom activity or exercise that has an objective obtained by meaningful interactions among participants with a mechanism for structuring and sequencing interaction. A task is not only an activity, but also a language learning effort which requires learners to comprehend, manipulate, and produce the target language while performing a set of task involving real-world language (Richard, 1986).
A task does not necessarily mean producing a language utterance. It can also be in the form of an activity or action which is performed as a result of processing or comprehending the language, for example drawing a map or pictures while listening to a story from a tape, listening to instruction, and performing it (Richards, Platt, and Weber, 1986). Nunan (2004) divides tasks into two kinds: real-world or target tasks and pedagogical tasks. Real-world tasks are the use of language beyond the classroom. On the other hand, pedagogical tasks are the use of language that occurs in the classroom. In applying TBLT, teachers give students communicative tasks which reflect the language used in real situations. Nunan also defines a pedagogical task as a piece of classroom work that involves different activities that students have to perform such as to comprehend, manipulate, produce on interaction in the target language. In doing so, their focus lies on utilizing their grammatical knowledge to express or communicate meaning. A framework of Task-based Lesson Prabhu (1987), Skehan (1996), and Willis (1996) propose three principal phases in a task-based lesson. They are (1) pre-task which includes a list of various activities that teachers and students can utilize before they start performing the task, (2) during-task in which students do the task through language, and (3) post-task which emphasizes two goalsthe language analysis and extended practice.

Speaking Skills
Wilkins in Oktarina (2002) states that speaking skill is the ability to produce and arrange sentences as communication occurs through the series of sentences which reflect different behaviour from different societies. Luoma (2004) proposed six characteristics of the use of language in speaking: (1) the use of spoken language, (2) spoken language is represented by pronunciation, (3) speaking uses spoken grammar, (4) a speaker uses generic words or phrases with fillers, hesitation markers, and repetition, (5) speaking is reciprocity, and (6) speaking can be in the form of monolog or dialog.

Method
The design of this research is quasi-experimental research because it uses an empirical interventional study to estimate the causal impact of an intervention on the target population without random assignment. It aims to establish a cause-and-effect relationship between an independent and dependent variable. A quasi-experimental design allows the researcher to control the assignment to the treatment condition by using some criterion.
The research chose one experimental group and one control group. These groups were chosen from six classes of semester one students of non-English major at a local private college in Bogor. The students of group A were the experimental group while those of group B were the control group. The experimental group was taught using the TBLT planned by the researcher in which task-based procedures were applied from the beginning to the end of the experiment. The control group was taught using a combination of methods where TBLT was one of them. Both the experimental and control groups had to undergo pre-test and post-test to learn the development in their speaking skills. Teaching and learning activities for the experimental group were designed by the researcher.
The average score in all items in pre-test and post-test were tabulated and calculated to find out the influence of TBLT in developing the students' speaking skills. The difference of the scores and T-test were calculated to find out the significant difference between the pre-test and post-test scores.
The researcher also prepared a questionnaire which was distributed to the experimental group during the treatment before and after TBLT is utilized. The objective is to compare the students' responses and perceptions towards the tasks. The questionnaire contains 42 statement items and they were arranged with a five-point Likert Scale and assessed with values ranging from 1 to 5. The scoring for the positive statements are described as follows: strongly agree = 5, agree = 4, neither agree nor disagree = 3, disagree =2, and strongly disagree = 1. The raw data were collected and then analyzed by calculating the mean values and standard deviation. Table 1 shows that the mean scores of each aspect of the speaking skill pre-test and post-test between the control and experimental group are relatively the same. Moreover, it also shows that the mean scores of the pre-test of the control group are higher than those of the experimental group.

Results and Discussion
Moreover, the mean scores of the post-test reveal that both groups have experienced an improvement in their speaking skills. In the aspect of pronunciation, the mean score of the control group rises from 16.6552 to 17.5517. Thus, there is an increase of 5.38%. Similarly, the mean score of the experimental group rises from 15.6667 to 17.7333 and this shows that there is an increase of 13.19%.
In the aspect of grammar, which is considered the most difficult among students, the mean score of the control group shows an increase from 15.2414 to 16.0690. The percentage of the increase is 5.43%. The experimental group's means score demonstrates an increase from 14.700 to 16.8333 and marks an additional 14.50%.
In the aspect of vocabulary, the mean score of the control group accelerates from 16.4483 to 17.4828. It shows that there is an increase of 6.29%. Likewise, the mean score of the experimental group grows from 15.5667 to 18.000 which means that there is a considerable increase of 15.63%.
In the aspect of fluency, the mean score of the control group rises from 17.5517 to 19.2069. This means that the mean score experiences an increase of 9.43%. The mean score of the control group rises as well, from 16.2000 to 19.6333. It substantially increases by 21.19%. The researcher observed that during the implementation of the TBLT instructional program, students showed high interest in the activities and the tasks. They can practice their speaking skills by relating the tasks with their real activities. The atmosphere during the class made them comfortable to get involved more actively in the activities. This influences their confidence in speaking and eventually results in their improvement in fluency. On the other hand, the control group also shows a promising development in their speaking ability.  Table 2 indicates that the increase of the mean scores of the experimental group is higher than the control group in all aspects. The average increase of the experimental group is 12.902%, whereas the control group is 5.306%. This proves that the use of task-based instruction increases the students' speaking skills. Table 2 also shows that the highest increase is in the students' fluency. This is due to the fact that after having experienced various tasks, they feel more confident with their English skills as their pronunciation, grammar and vocabulary also improves.  Table 3 shows the difference in the mean scores of each aspect in pre-test and post-test. The results indicate that the difference in the mean scores of the experimental group is higher than the control group. There is a statistically significant effect of the implementation of a taskbased approach in the English lessons as the score of sig.2-tailed < 0.005. Table 4 and 5 show the mean scores, Pearson Correlation, tStat and critical One-Tail of the speaking skill tests of the control and experimental groups. Table 4 shows the significant difference in the total mean scores of the pre-and post-test of the experimental group. There is a 16.93% increase of the total mean scores from 62 to 72.5 On the other hand, the total mean scores of the control group have also shown an increase of 6.61% from 65.172 to 69.483. This means that the experimental group which was treated by TBLT instructional program has experienced a higher increase in the total mean score of speaking skills tests than the control group. The table also shows that the tStat ˃ tCritical one-tail (7.950 ˃1.699) which means that TBLT instructional program is effective to improve the students' speaking skills. Figure 1 represents the scores of pre-test and post-test of speaking skills of the control group. The above diagram shows that most students have improved their speaking abilities during the instructional program. However, there are six (6) students whose total speaking test scores remain the same. Five of them got 60 and based on the researcher's observation, these students showed a lack of interest in involving themselves in the learning activities despite the encouragement they received from their classmates and tutor. Figure 2 represents the scores of pre-test and post-test of speaking skills of the experimental group.  The Pre-Test and Post Test of Speaking Skills of the Experimental Group Figure 2 shows that there are only 3 students whose total scores in the pre-test and pos test are the same, 2 of them had excellent performances in both tests. An interesting data reveals that the increase of the scores is very significant as shown in the diagram. Table 6 shows the correlation between the pre-test and post-test.  Table 6 shows that the Correlation Coefficient of all dimensions of the speaking test is significant. Since the Sig. Value ˂ probability value (0.000 ˂ 0.05), it can be concluded that there is a significant difference between the pre-test and post-test for both the control group and experimental group.
This means that both the control and experimental groups demonstrated an increase from the beginning of the course to the end. There is an interesting phenomenon that students in the control group experienced an improvement in all aspects of the speaking skills test as they have more opportunities to speak English in the classroom compared to what they had in their high school.

The Effects of TBLT on Students' Attitudes Towards English
To measure the students' attitudes towards the English lessons, the researcher calculated the means, standard deviations and standard error means of the students' responses to the attitudinal questionnaire items before and after the implementation of TBLT program. The results of the analysis are presented in table 7. Table 7 shows that there is an observed difference between the scores of means of each orientation of pre and post-tests, namely before and after the implementation of the TBLT teaching program.
The items in the first orientation, developmental orientation, are designed to investigate the students' perception of the English lessons. The results show that the mean score rises from 15.967 to 19.167. It indicates that there is a 20.04 % increase. The participant's perception of the English lesson has changed significantly. Before the implementation of task-based instruction, many of them did not enjoy the English classes they took in their previous schools. They thought that English lessons were not interesting and dull. Some of them even considered that learning English was a waste of time and they would give up studying English when they left school. However, after experiencing the task-based instruction, their perception of the English lesson has become more positive because they were given more opportunities to explore their English practice. In the integrative orientation, the items are used to identify the students' perception of whether the English lessons will be beneficial for them to understand more about the language itself, the people, and their culture. The results indicate that the mean score increases from 49.300 to 53.400, and the percentage of the increase reaches 8.32%. Before the implementation of task-based instruction, the students showed no interest in the English or American people and culture. They did not think that English was important for them to be able to communicate with English-speaking people. After the implementation of TBLT, their point of view changed although the increase was not too high. In conclusion, this signifies that the student's interests in the English language, the people, and their culture have grown quite considerably.
The items in the instrumental orientation are applied to examine the students' perception of the importance of English lessons in helping them with their study and the advancement of their future careers. The mean score rises from 61.667 to 68.267 which shows a 10.70% increase. Before the implementation of the instruction, they had already had a positive attitude towards English as a means of study and a career boost. They realized that English was very important to help them understand their textbooks, information they received from various sources, and for their future work. It means the students knew the importance of English for their education and work. After the implementation of the approach, their attitudes towards English in this orientation increased. This indicates that more students begin to realize that English is important for their present and future needs.
In the travel orientation, the items are used to find out the students' perception of whether studying English will be valuable when they go abroad. The results show that the mean score also increases from 22.900 to 26.667 which signifies that there is a 16.45% raise. The increase is relatively high since the implementation of the task-based instruction, the students have lacked the intention to go abroad either for vacations or further education. After the carrying out of the program, more students got interested to have some experience in foreign countries. Hence, their attitudes to English became more positive. This means that the student's awareness of the importance of having English skills, particularly speaking skills, is important when they go abroad. The increased percentage of the means the subjects' responses to the Attitudinal Questionnaire Items are presented in table 8. The percentages listed in Table 4.8 show that the student's perception of the English lesson has positively increased, with an average increase of 13.88%. To measure the significant difference of means of the students' perception towards English and the English lesson, the researcher performed a Paired Sample t-Test and the results are presented in Table 9.  Table 9 shows that there is a statistically significant effect of TBLT (Sig. 2 tailed ˂ 0.05) attributed to the Task-Based program. This proves that TBLT affected positively the students' attitudes towards English on the dimensions of the questionnaire.

Conclusion
The results of the post-speaking tests of the control group show that there is an increase in the students' speaking skills. Using authentic materials is one of the reasons why students feel more comfortable learning English. Since the activities taken during the classes require them to interact not only with the teacher but also with their friends, their self-confidence has also improved. The support that comes from the teacher and friends makes each of the students feel more relaxed during the lessons and eventually, it encourages them to try to speak using English. By doing this, the students' fluency and accuracy also improved. However, there are also some barriers to implementing TBLT instructions. The textbooks used in the English lessons are not sufficient enough for a task-based program since they do not contain materials with tasks. This problem can be solved by finding additional sources that are available abundantly on the Internet. Another disadvantage is related to the curriculum which does not support the purpose and the application of TBLT.
The researcher urges that EFL lecturers use TBLT procedures in their teaching instructions because it is proved it develops the students' speaking skills in terms of accuracy and fluency as well as their attitudes towards English. As the role of the English instructors in the TBLT program is very important, the researcher recommends that the English coordinator organizes training simulations for instructors on the application of TBLT instructions in their daily classroom activities.
Curriculum and syllabus designers are suggested to supply the English textbooks which are equipped with TBLT features and procedures. The textbooks should contain well-designed tasks and activities both in the teachers' and students' books. It is highly recommended that the other researchers perform further studies to observe the effectiveness of TBLT on improving other language skills. As classroom teachers know the real situations of their students, it is also recommended that they design some of the content of textbooks and search additional sources from different applications to create a more up-to-date topic and approach based on the procedures and principles of the TBLT.