TEACHING SPEAKING THROUGH CHAIN STORYTELLING TO THE TENTH GRADE STUDENTS OF SMA SRIJAYA NEGARA PALEMBANG

: The objectives of this study were to find out: (1) whether or not there was a significant improvement in speaking skill of students who were exposed to chain storytelling and (2) whether or not there was a significant difference in speaking skill between students who were exposed to chain storytelling and those who were not. The sample of this study was 60 tenth grade students of SMA Srijaya Negara, which were divided into control and experimental group, and each group had 30 students. To collect the data, each group was assigned pretest and post test. The data analyses used paired sample t test and independent sample t test in SPSS. The result from paired sample t test showed that the mean difference in post test and pretest of experimental group was 16.000 at the significance level of p<0.05 and since t obtained was higher than t table (9.649>2.04523), H 01 was rejected and there was a significant improvement in speaking skill of experimental group. The highest contribution was given by story elaboration aspect (content), and then followed by pronunciation, vocabulary, fluency and grammar. The result from independent sample t test showed that the mean difference between post test of experimental group and control group was 2.4000 at the significance level of p>0.05 with t obtained<t table (0.942<2.00171), since t obtained was lower than t table, H 02 was accepted and there was no significant difference in speaking skill between students who were taught by using chain storytelling and those who were not. Nevertheless, the mean difference from the experimental group was higher than that of control group (16.600>7.133).

Language has played a very important role in human being civilization as a means of communication. As a language, English has been claimed as an international language and widely used by many countries, including Indonesia, which establishes English as foreign language. Kurniati (2011) in her study highlighted that speaking skill plays an important role for language learners who use it for instruction, discussion, presenting arguments, expressing ideas, thinking and learning. When people want to apply for scholarship, job interview or join international forum, being able to speak English is main requirement to be involved in those activities. Along with the strengthening position of English as a language for international communication, the teaching of speaking skill has become increasingly important in the English as a second or foreign language (ESL/EFL) context (Widiati & Cahyono, 2006).
The teaching of speaking skill itself is evidently acknowledged in the curriculum along with the other three skills such as listening, reading and writing. Based on School Based Curriculum (Departemen Pendidikan Nasional, 2006), the scope of learning English in high school includes the students' competence to comprehend and produce spoken and written text through the four language skills. However, during the teaching in the classroom, teacher seldom pay attention to students' need in speaking skill. Thus, students face some difficulties when they should speak English. Gebhard (1996) as cited in Aprilia (2012) revealed that some EFL students, including advanced students, were too shy or had such high levels of anxiety over speaking that they would not speak in class. Some students are likely to become overly concerned about accuracy, possibly berating themselves for the mistakes they make and demanding constant corrections for every slip-up (Brown, 2007, p.125). Tutyandari (2005) and Zulfikri (2010) also mention that students keep silent because they lack of self-confidence, and prior knowledge about topics. Another reason is the lack of exposure on speaking practice itself (Resty, 2012). These are the problems which faced by students of SMA Srijaya Negara.
In order to help and encourage students to speak, many methods and techniques are developed to fulfill the needs in teaching speaking and one of which is chain storytelling. Chain storytelling is one of the ways in which the students will continue the story made by other students. In chain storytelling, the teachers start telling a story, break off after a few sentences, and call on a student to continue (Klippel, 1984). Thus, each student will have equal chance to speak and render their ideas to elaborate the story first told by teachers. It is very interesting, because they can continue the story and provide imaginative yet unpredictable ending (McCarthy and O'Keeffe, 2004). Chain storytelling can also help learners to improve their list of vocabulary and how to express their idea into good sentence along with their creativity in telling the story (Bintz, 2011). This also is proven by Akhyak and Indramawan (2013) in their study that teaching speaking to ESL using storytelling can improve sensitivity in recognizing pronunciation errors, enlarge vocabulary, and understand to make sentence in good structure and speak fluently in concrete content. In addition, Bailey and Savage (1994, p.8) proposes that chain storytelling can be used to develop fluency and the security of the structure of a story. By doing this, teacher can help to reduce students' fear of making mistake by maintaining a friendly atmosphere in the classroom and providing opportunities for students to practice with other students (Kurniati, 2011).
Based on the explanation above, the writer carried out an experimental research about teaching speaking skill by using chain storytelling entitled "Teaching Speaking through Chain Storytelling to the Tenth Grade Students of SMA Srijaya Negara Palembang". The writer wanted to find out whether or not chain storytelling is effective to be used in teaching speaking and show a significant difference in speaking skill between students who are taught by using chain storytelling technique and those who are not taught by using chain storytelling.
Thus, based on the explanation above, the problem of this study can be formulated as follows.
1. Is there any significant difference in speaking skill between pre test and post test of experimental group? 2. Is there any significant contribution of each aspect of speaking skill to the improvement of speaking skill in experimental group? 3. Is there any significant difference in speaking skill between students who are exposed to chain storytelling and those who are not?

LITERATURE REVIEW
Speaking is the most important and essential skill (Oradee, 2012). It is one of productive skills that can be measured directly and gives clear reflection of people understanding. Despite its importance, for many years, teaching speaking has been undervalued and English language teachers have continued to teach speaking just as repetition of drills or memorization of dialogues (Kayi, 2006).
In fact, many students cannot communicate each other in English either in the classroom or out side the classroom, they are frequently vacuum and passive in English communication (Tahir, 2011). Their anxiety grows larger along with the absence of practice in class. This leads to learners' lack of self-confidence and avoidance when communicating with native English. In particular, EFL learners often stammer when speaking English and this results from learners' lack of exposure to authentic English language environments that allow them to use English for communication and expression (Oradee, 2012). Consequently, the aim of English as a means of communication is failed to be reached.
In order to serve English as a means of communication, teachers need to familiarize the students with English speaking environment which arise no threat to students' confidence to speak. This can be done by doing fun learning activity which attracts students' attention and encourage them to speak. Kayi (2006) proposes that many linguistics and teachers agree on that students learn to speak by "interaction". Interaction involves both social and personal input, emotions, creativity, agreement, and disagreement (Counihan, 1998). However, before students are assigned to a whole class interaction, let them work together in a smaller group or in pair. Teachers need to include many student-to-student interactions in order to ease their feeling and help them build their confidence to speak. Umayah (2012) also proposes that students should have much time to practice a language with their friends in a group or in pairs to build up their fluency and accuracy. One of the ways is by applying a communicative and collaborative learning environment.
Communicative language teaching and collaborative learning serve best for this aim (Kayi, 2006), in which the learners' participations in class are actively needed. Collaborative learning has been further defined as the learning that occurs as a result of interaction between peers engaged in the completion of a common task (Faulin, 2013), while in communicative approach, teacher are concerned with the students not only practice speaking in a controlled way in order to produce features of pronunciation, vocabulary, and accurately, but also practice using these features more freely in purposeful communication (Hedge, 2000, p. 261).
In order to create an interactive learning environment, teachers need to do collaborative and communicative learning which allows students to express their ideas and share their ideas with their friends. Sharing ideas can be done through stories. Story is a powerful means of language teaching (Celce-Muria, 2001, p. 144). Stories are very motivating, challenging and great fun for students in which they can exercise their imagination through stories (Urbancová, 2006). Stories can be taught by using chain storytelling. According to Klippel (1984), this technique allows students to create their own story and let the other students to continue the first story based on their version. Chain storytelling offers students a choice where they are encouraged to use their imagination and make choices in content within the framework based on topic given by teacher (Putra, 2013). Chain storytelling can be used to develop fluency by balancing the risk of free conversation with the security of a structure (Bailey & Savage, 1994, p. 8).
In fact, not all students in the classroom take part in speaking during the class; some may produce one or two sentences, few dominate the speaking activity and the rest just keep silent and listen to what others say as highlighted by Tahir (2011). Many students, especially low-level learners, either will not volunteer to speak or will have difficulty choosing a subject (Bailey & Savage, 1994, p. 8). Hence, chain storytelling should be applied in classroom.
The basic concept of chain storytelling is to continue the story composed by other people. As formulated by Bray (1994) in Bailey and Savage (1994, p. 119): 1. Teacher divide students into some groups 2. Nominates a topic and say something about it, like a story starter. 3. Let the students continue from the story starter the teacher had given. 4. To help students elaborate the topic, teacher may provide some cues, such as word cards, or temporal connectors. Furthermore, Eggins and Slade (1997) suggests that before coming into the main activity, there are some steps to prepare students' readiness in making a story: (1) Brainstorming idea or topic, Brainstorming is a simple and effective skill that used for creating of thinking and ideas. Cullen (1998) also suggests that brainstorming is an activity used to generate ideas in small groups. Students can use brainstorming in class activities, but it takes a short time of doing brainstorming.
(2) Presenting an introductory lesson to raise their awareness of the story's generic structure. Teacher can tell a common story as warmup activities.
Students can create their own story by elaborating the existing story or simply create their own.
The story used in chain storytelling is in the form of narrative, which its basic purpose is to entertain and hold readers' interest. The language features of this kind of story as proposed by Putra (2013) are the use of past tense, time conjunction, and specific characters. It also has three aspects which basically compose structure of narrative story (Putra, 2013).
(1) Orientation Introduce the character, setting, and time of the story.
(2) Complication Present the problem, conflicts faced by the characters of the story.

(3) Resolution
Describe the problem solution of the conflicts.

RESEARCH METHOD
This study belonged to experimental method and used quasi experimental design. This design consisted of two groups, and they were control and experimental groups. The population of this study was 216 tenth grade students of SMA Srijaya Negara, and the sample was 60 tenth grade students, in which each group had 30 students.
The technique used to choose the sample was purposive sampling, in which the researcher used personal judgment to select the sample for the specific purpose. The X 1 and X 2 class were taken because they were taught by the same teacher, and they shared similar number of students. Besides, based on the interview from the teacher of both classes, their speaking was considered low among the other three skills.
Between the two groups, only experimental group (X1) was given treatment while the control group (X2) was not. During the treatment, the experimental group did chain storytelling in classroom. Chain storytelling was done in group, each group was given the same story starter but each of them had different ending. Each member of the group continued the story, so each of them had contribution upon the completion of the story. Before they did chain storytelling in group, they also did chain storytelling between group. It was done to familiarize them with the technique.
To collect the data, both experimental and control group were assigned a speaking test in the form of pretest and posttest. The test was constructed based on content validity and the test content was also consulted with two experts who were also the advisors of this study. To check the reliability of the study, inter rater reliability was used. The raters scored the test by using rubric, and the data gathered was analyzed by using Pearson Product Moment. Based on the analyses, the reliability coefficient of pretest and post test of experimental group and pretest and post test of control group consecutively were 0.776, 0.989, 0.986, and 0.953. All of the reliability coefficients were higher than 0.7, thus the test was reliable.
After collecting the data, the data were analyzed by using paired sample t test and independent sample t test. Paired sample t-test was used to analyze data gathered from pre test and post test of experimental group, whereas independent sample t-test was used to analyze data gathered from experimental group and control group.
From paired sample t-test and independent sample t-test analyses, t obtained, the degree of freedom and the significance level (in two-tailed test) were found. To know whether the null hypothesis was rejected or accepted, there are two ways, based on the ratio of t obtained and t table; and the ratio of the probability value. Hence, t table should also be found. If the t obtained < t table and the p value > 0.05, then the null hypothesis is accepted and if the t obtained > t table and the p value < 0.05, thus the null hypothesis is rejected.

FINDINGS AND INTERPRETATION
Based on the data obtained, the mean of experimental group in speaking skill was 41.133 (very poor) in pretest and 57.133 (average) in post test, while the mean difference was 16.000. The data distribution from experimental group during their pretest and posttest is presented in Table 1. From the table, students' speaking skill in pretest concentrates on very poor and poor level, while only four of them were in average level. In the pretest, half of the students were in very poor category. After they were exposed to the treatment, one student was in very good level, eight students were in average level and good level of speaking skill, while students who were very poor in speaking shrank and moved up to poor and average level. If the means from pretest and posttest were to compare, there was a significant improvement in their speaking skill as a whole.
In addition, for the control group, the mean of the pretest was 47.600 (poor), while the mean of posttest was 54.733 (poor), and the mean difference was 7.133. The data distribution from control group during their pretest and posttest can be viewed in Table 2. From the table, it can be concluded that nearly half of the students from control group in pretest were in very poor level. Ten of them were in average level and the other was in poor state. This group was not being exposed to the treatment. After doing the post test, each level advanced and the overall students' speaking skill were in average level but only two students excelled to good level. If the pretest and post test were to compare, they also had a quite improvement.
The data were also analyzed statistically by using paired sample t test and independent sample t test. Before doing a statistical analysis, the normality of the data distribution needs to be checked. Each of the data from the pretest and post test from experimental and control group was analyzed. It was analyzed by using One-Sample Kolmogorov-Smirnov test. Based on the results, the significance value in two tailed testing gained from pretest and post test of experimental group were 0.584 and 0.830, while from pretest and post test of control group the value were 0.629 and 0.442. Priyatno (2008, p.28) states that the data can be said having a normal distribution if the p>0.05. The results showed that the value from both group were higher than 0.05, they were 0.830, 0.584, 0.629 and 0.442. Thus, it can be concluded that the data obtained were considered normal.
After the normality of the data distribution was ensured, t-test can be applied. In this study, the writer used paired sample t-test and independent sample t-test. Paired sample t-test was used to analyze data gained from pretest and post test of experimental group, while independent sample t-test was used to analyze data gained from both experimental and control group. The result of paired sample t-test can be viewed in table 3. From table 3, the mean difference of students' speaking skill in the experimental group was 16.000. The result of paired sample t-test showed that t obtained was 9.649. At the significance level p<0.05 in two-tailed testing and the degree of freedom (df) was 29, t table was 2.04523. Since t obtained was higher than t table (9.649>2.04523) and p value<0.05, the null hypothesis was rejected and the research hypothesis was accepted which means that there was a significant improvement in students' speaking skill between pretest and post test after being taught by using chain storytelling.
To see the correlation and the contribution of each aspect of speaking towards the improvement of speaking skill in experimental group, stepwise regression analysis was used. Based on the result of the analysis, the coefficient correlation of each aspect of speaking towards speaking can be seen in table 4.  (2011), if the correlation is 0.60-0.79, it means that it has strong correlation, and if the correlation is 0.8-1, it means that it has very strong correlation. From the table, it can be seen that pronunciation, grammar, vocabulary and content aspect had a very strong correlation to the improvement of speaking, while fluency aspect had a strong correlation to the improvement of speaking skill. Each of five aspects had correlation in the improvement of speaking skill in experimental group.
Based on the ANOVA analysis in Stepwise Regression, it was also shown that the p value of each aspect was 0.000 and p<0.05. It means that each aspect gave contribution to the improvement of speaking skill in experimental group. The contribution of each aspect was determined from the R square of each aspect (see appendices). Based on the R square, the contribution of each aspect towards the improvement of speaking skill can be seen in Table 5. Based on Table 5, the highest contribution in the improvement of speaking skill was given by content (story elaboration) aspect and consecutively followed by pronunciation, vocabulary, fluency and grammar aspect. It can be concluded that chain storytelling which was used in the teaching process of experimental group could improve students' ability in elaborating a story and developed their pronunciation, vocabulary, fluency and their grammar. The contribution of each aspect was obtained from the R Square difference of each aspect.
Whereas, to see the difference between pretest and post test score of both experimental and control group, independent sample t test was done. The result of independent sample t-test of pretest of both groups is presented in Table 5. Based on the homogeneity test by using Levene's test, the significance value was 0.113 (p>0.05). Since significance value was higher than 0.05, it means that the variance was assumed equal. From the table, it can be seen that for equal variances assumed, the significance value was 0.028 (p<0.05), degree of freedom was 58, t obtained was 2.252, and t table was 2.001. Since t obtained was higher than t table (2.252>2.001) and p value<0.05. It can be interpreted that there was significant difference in pretest of speaking skill of both experimental and control group in which the mean of pretest in control group was higher than in experimental group.
Whereas, the difference of post test score of both group can be shown in Table  6. Based on the homogeneity test by using Levene's test, the data showed the p>0.05 (0.125>0.05), it can be concluded that the variance of the data is equal. Thus, the value which was used in independent sample t test was taken from equal variances assumed point.
At the significance level p>0.05 (0.350>0.05) and the degree of freedom was 58, t obtained was 0.942, and t table from those data was 2.001. Since the t table was higher than t obtained (2.001>0.942) and p>0.05, thus the null hypothesis (H01) was accepted that there was no significant difference in speaking skill between students who were taught by using chain storytelling and those who were not. Nevertheless, the mean difference from experimental group was higher than control group's (16.000>7.1330) although there was no significant difference between both groups.
It happened because of experimental group and control group had different level of speaking. Based on the means of pretest of both groups, the mean of experimental group was only 41.133 which were categorized as very poor level, whereas the control group's was 47.600 which were categorized as poor level. In post test, the means of experimental group was 57.133 (average), while control group was 54.733 (poor level). in post test, both group had different level of speaking, but the difference was very small which was only by 2.4 point.
Based on the findings of this study, some interpretations are drawn. The findings show that (1) there was a significant improvement in speaking skill of experimental group before and after given treatment, and (2) the highest contribution towards speaking improvement was in story elaboration aspect (content aspect) and then followed by pronunciation, vocabulary, fluency and grammar, but (3) there was no significant difference in students' speaking skill of both experimental and control group.
The first finding showed that there was significant improvement in speaking skill of experimental group before and after they were given the treatment. It can be seen from the mean difference of students' speaking test in pretest and post test. The result showed that the mean of students' speaking test in post test was higher than that of pretest. It was also proven and supported by the statistical analysis done by using paired sample t test in the previous discussion. The improvement itself could happen because after the experimental group was assigned pretest, the writer gave them the treatment in the form of chain storytelling for around a month.
There are two reasons why chain storytelling can improve students speaking skill. Firstly, chain storytelling offers an interesting way of learning in which they have to build a story completion by chaining the plot story. By chain storytelling, each student continues the story however they want because they decide the story plot themselves, it makes the ending of the story unpredictable. It is also in line with the finding of Febianti (2011) that chain storytelling offered an interesting way to involve the students to actively participate in classroom.
Secondly, chain storytelling provides opportunity for students to speak evenly and it also provides more interaction between students to students which ease them to speak. It is also supported by Febianti (2011) that chain storytelling could help students to increase their confidence to speak and ease them to speak in front of their friends. Each of them should contribute to complete the story and has the chance to speak equally, so even they really have nothing to say, it still reinforces them to give their contribution in building the story plot. Thus, the problem highlighted by Tahir (2011) about inhibition in speaking and uneven or low participation in classroom can be reduced.
The second finding showed that there was a contribution from each aspect of speaking in the improvement of students' speaking skill in experimental group. It can be proven from the statistical analysis done by using stepwise regression. The result of the analysis showed that the highest contribution was given from content aspect, followed by pronunciation, vocabulary, fluency and grammar. This could happen because during the treatment, they were exposed to many stories which helped them to know how to build a story plot. They knew how to express their idea into good sentences along with their creativity in telling the story (Bintz, 2011).
Akhyak and Indramawan (2013) also found that storytelling could improve sensitivity in recognizing pronunciation errors. In the first meeting, most of the students were having problem in pronounce words correctly. Some had good pronunciation, but most of them still needed improvement. Thus, in every meeting, the writer taught the students on how to pronounce some words correctly, but due to the time management and it was quite hard to focus on each student's pronunciation aspect, they often mispronounce the words that had been taught. They needed more practice to pronounce some words correctly, because the incorrect one had fossilized, so it was quite difficult to change their pronunciation in a day. However, after the treatment, the experimental group showed betterment in pronunciation aspect.
As for the vocabulary issue, before the students were asked to chain storytelling, the writer told them the complete version of the story chosen. Hence, they would have some references of vocabulary in creating a different version of the story. After the treatment, their vocabulary lists were developed. Thus, it was proven that chain storytelling could improve students' list of vocabulary as proposed by Bintz (2011) andAkhyak andIndramawan (2013).
In addition, Bailey and Savage (1994, p.8) states that chain storytelling can be used to develop fluency. At first, the students were still confused to express their story, sometimes they only produced two or three sentences, but after some practices, they could manage to produce more sentences. They could continue the story spontaneously and fluently. Moreover, the improvement of the fluency was also in accordance with the development of their vocabulary and grammar. The more vocabulary and grammar they understand, the easier they continue the story.
As in grammar matter, the students still had difficulties in using the right verb or to be, sometimes they misused the verbs or other vocabulary, but step by step, after some brief explanations and lot of exposure towards the use of language feature in the story, they began to aware the different use of verbs, and the security of the structure were developed as supported by Bailey and Savage (1994, p.8). It was quite hard to distinct it due to their pronunciation matter though.
The third finding confirmed that there was no significant difference in speaking skill between experimental and control group. It was because there was also an improvement in control group's speaking skill although it was not as much as the experimental group. It happened because the students were not given any treatment. The control group was only given pre test and post test. However, during the teaching and learning activity, the students also learned about narrative story. Mostly, the teacher gave them explanation about the generic structure of the narrative. They were barely exposed to create a story, they were only asked to answer questions related with the story and were insisted to know which was the orientation, complication and resolution.
Besides, the mean of pretest in control group was higher than that of experimental group and there was significant difference in pretest between both groups. (see Table  5, p…). From the mean of experimental group, it can be said that the level of speaking skill in experimental group was very poor, whereas from the mean of control group, the speaking skill was in poor level. It means that control group was a level higher than experimental group. Nevertheless, if the improvement of both groups were to compare, the improvement of experimental group was twice higher than the control group. This can be seen from the mean of post test from both groups. The mean of post test in experimental group showed that the group was in average level of speaking, while the control group was in poor level. However, even if the speaking level of experimental group was a level higher than control group, the mean difference of post test from both groups was not much different. This was the reason why there was no significant difference between post test of experimental and control group.
Apart from that, the using of chain storytelling still gave much contribution to the improvement of speaking skill in experimental group. Hence, teaching speaking through chain storytelling is preferable because it helps them to elaborate their story content, vocabulary, fluency, improve their pronunciation and grammar as shown in the result of this study. Moreover, chain storytelling is very interesting, because they can continue the story and provide imaginative yet unpredictable ending (McCarthy and O'Keeffe, 2004). Besides, it also provides students with meaningful interaction and communication with others without leading them in memorization (Hedge, 2000) because they need to make their own based on what proposed by the previous storyteller.

CONCLUSION AND SUGGESTIONS
The writer concluded that the use of chain storytelling is proven to be effective in teaching speaking and can improve their speaking skill apart from the absence of significance difference in speaking skill between control and experimental group.. The students' speaking result in post test of experimental group showed that there was a significant improvement in speaking before and after being taught by using chain storytelling, and also there were some contribution given by each aspect of speaking to the improvement of speaking, especially in story elaboration skill (content). Chain storytelling also helped students to be an active student because each of them needs to contribute in building the story. Moreover, it created an enjoyable learning atmosphere where they needed to guess how the story would become and what they should give to keep the story more interesting and unpredictable.
Referring to the conclusion above, the writer proposes some suggestions for the betterment of teaching English especially speaking skill to the students. For the teachers, the writer hopes that chain storytelling can be considered as one of ways to be used in teaching English especially to improve their speaking skill. Chain storytelling can be better observed when being used in a small classroom, but it also can be used in a large classroom. The point is the teacher should be able to manage the class, so the students will focus on building or chain storytelling rather than talking each other. In addition for the students, they should be active in a classroom. If the teachers' explanation is not clear enough, feel free to ask. Moreover, if you want to have better speaking skill, keep practicing, especially in pronunciation, and read a lot so you will have lot of vocabulary. Lastly, for those who want to do further research, preparation is really important. It not only covers the material and method of assessing the students, but also covers the strategy and approach to deal and manage the students. If the strategy is good but the students are not cooperative enough, then it will be a waste.