The Influence of Topic Management on the Speaking Ability of Iranian Intermediate EFL Learners

In English as a Foreign Language (EFL) settings where English is not the spoken language, Topic Management has been recognised as one of the factors that influence the speaking ability of learners (Du-Babcock, 1999). In Iran, many intermediate EFL learners find it very difficult to manage the speech aspect of the English language. The present study examines the influence of Topic Management on the speaking ability of Iranian intermediate EFL learners. It provides useful insights on teaching speaking skills through Topic Management to improve the speaking ability of learners. In this study, quasi-experimental design was used. 40 intermediate learners were selected and divided into two groups, the Control Group and the Experimental Group. Three tests were administered to obtain data from the participants. These were a proficiency test to check the English proficiency level of the participants, a speaking pre-test to check their speaking ability before the treatment, and a speaking post-test to check their speaking ability after the treatment. The results indicate that Topic Management has an influence on the speaking ability of Iranian EFL learners. That is, the more the student is confident in managing the topic the more he/she is able to sustain the conversation. The study concludes that Topic Management can be a facilitative approach to improving Iranian intermediate EFL learners’ speaking ability.


INTRODUCTION
Of all four key language skills, speaking is deemed to be the most important in learning a second or foreign language as it includes all other skills of knowing a language (Ur, 1996). It is an interaction process of constructing meaning that involves producing, receiving and processing of information (Brown, 1994;Burns & Joyce, 1997). Its form and meaning are dependent on the context in which it occurs including the participants themselves, their purpose for speaking and collective experiences (Burns & Joyce, 1997). Riverse (1987) believes that language teachers should provide learners with opportunities for meaningful communicative behavior about relevant topics by using learners in interaction as a key to the teaching of language for communication, because communication derives essentially from interaction. According to Richard (1990) teachers and textbooks make use of a variety of approaches, ranging from direct approaches focusing on specific features of speaking (e.g., topic management, turn taking, and questioning strategies) to indirect approaches that create condition for oral interaction through group work, and other strategies. Researchers assume that power relation(s) somehow exist and determine the course of actual encounters, by focusing on the management of talk-in interaction. Published  This power may be viewed in terms of different distributions of lengthy resources. These lengthy resources enable certain participants to reach interactional effects that are not available to all, or are differentially available to others in the setting, (Yieke, 2007). Topic Management and topic control within the turn taking process are examples of discursive resources that may place constraints on the discourse options that are available to speakers in a discourse situation. The more powerful speakers in a workplace situation in terms of professional status may employ the use of topic management and control, which may suppress or oppress their less powerful interlocutors. With this in mind, Topic Management has been a field of inquiry for so many years considering the relationship between the learners' speaking practice and their control over their topic's selection. In Iran, and despite the long-time of learning and practicing English language, the majority of students still face difficulties in their speaking practice. To the best of the writer's knowledge, few studies have been conducted to investigate the influence of Topic Management on Iranian EFL learners' speaking ability, the present study attempts to focus on the influence of Topic Management on the Iranian EFL learners' speaking ability. It tries to answer the question: To what extent does Topic Management affect the Iranian EFL learners' speaking ability?

TOPIC MANAGEMENT AND L2 LEARNERS' SPEAKING ABILITY
English language has four basic skills, which are divided into productive and receptive skills. Productive skills seem to be more complicated. Speaking and Writing skills are considered as the productive skills. Speaking is defined as a process of building and sharing meaning through the use of verbal and oral form (Chaney, 1998;Gebhard, 1996). As Nunan (2003) puts it, speaking includes the process of producing systematic verbal utterances to convey meaning. many L2 learners see Speaking as the hardest skill among the other three skills, Writing, Reading, and Listening to be acquired when learning another language. Unlike Reading and Writing, Speaking happens in a real situation where the speaker produces a discourse that he/she cannot edit or revise what he/she says (Van Lier, 1995).
Speaking is considered as a skill by which people are judged while their first impression is being shaped, according to Hedge (2000). Taking this in consideration, speaking skill deserves more attention in language learning because as it reflects the learners' personalities, ideas, feelings, and thoughts. Furthermore, Speaking can aid L2 learners to develop their vocabulary, grammar and then improve their writing skill. Baker and Westrup (2003) state that learners who can speak English appropriately would have greater chance for further education, finding employment, gaining promotion, and have a good communication in social communities. Since speaking happens in real time and most of the times it has not been planned before, it can be called one of the most difficult aspects for learners to master. So, it requires the simultaneous use of a number of abilities, which often develop at different rates.
Topic in conversation is an ambiguous concept, and it is a complex term to be defined (Brown & Yule, 1983;Bedrosian, 1993). According to Brown topic is usually the thing and the situation what is being talked about. Cook (1990) further adds that topic is information that carry a message in a conversation. Another meaning states that topic is a set of propositions, which a speaker either wants to give information about or seeks, and information (Keenan, 1976. Topic plays a very important role in practicing the speaking skill in any learned language because one cannot control a communication without managing the related topic. Topic introduction, shifting, and maintenance are essential elements for providing conversational coherence and continuity (Mentis, 1994). Learners have to be able to manage the topics in their communication, in order to manage longer turns in their conversations. Managing topics in speaking requires the language learners to be able to initiate a topic, continue with the topic after it has been initiated, shift a topic to another topic if necessary, and conclude the topic when it comes to its end. A brief summary of each element is briefly explained based on Wong and Waring (2010) criterion.
Mentis (1994) conducted a research on topic shading. The results of his study showed that topic shading seems to be a clear way of shifting topics requires more complicated language skills than the initiation of new topics. Therefore, topic shading has affected the speaking ability much more than topic initiation. According to Erber (1996) topic maintenance and/or development do not create any complicated difficulties in the speaking ability of the students. He adds that the results are not that surprising since maintaining a topic is not that complicated and does not need new information. Moreover, he says that topic management generally in the class will have a positive effect on the speaking ability of the students.
Du-Babcock (1999) conducted a research on the difference of topic management and turn taking strategies in a professional communication. The results of his study showed that using the topic management strategies have influenced the speaking ability of students, which their mother tongue was Thai, more than turn taking strategies. Caissie (2002) designed a research on changing topics and communication breakdowns. He investigated the effect of conversational topic shifting by partners on the occurrence of communication breakdowns in people with hearing loss. This study investigated the communication breakdowns following conversational topic changes by normally hearing partners during communicative interactions with people with a hearing loss. Eleven adults with an acquired hearing loss were videotaped engaged in 15-minute conversation. In each conversation, teacher coded the partner's speaking turns into four topic activities: topic initiation, topic shading, topic maintenance, and topic termination. The results indicated that few communication breakdowns happened when partners used speaking turns that continued or increased conversational topics. However, there was an increase in the number of communication breakdowns, and in the number of speaking turns attempted to repairing the breakdowns, when partners shifted topics, either through topic shading or topic initiation. Moreover, when topics were shifted through shading, breakdowns were more likely to occur than through the initiation of a brand-new topic. The results of this study suggested some clinical solutions for by considering topic changes for people with hearing loss. A recent study by Jones (2008) showed that controlling over Topic Management can have a positive and effective influence not only on the speaking ability of the students, but also on their self-esteem when they are talking into another language

Design
The current study adopted the quasi-experimental design in terms of using an Experimental and Control Groups. One Experimental Group and one Control Group were employed in this study. These groups were chosen randomly from intermediate level at the Shokouh English language institute in Iran. The Experimental Group was taught using the topic management designed by the researcher and the Control Group was taught using the conventional method of teaching used by the EFL teachers at the institute.

Participants
The overall participants of this study were 60 English language learners who are (at the time) doing the course "American English file", intermediate level, at Shokouh English language institute. The students aged from 14 to 17 years old. All of them were females. The text-book used in this course is one of the Pearson Longman series', which have been designed for young and adult English learners in five books from beginner to intermediate. A proficiency pretest was administered to select the students who meet the intermediate level measures. Based on the test results, only 40 students were found to be at the intermediate level and chosen to be part of the study. The students were then divided into two groups, Control Group and Experimental Group (20 students per group). See Table 1 below.

Proficiency test
As discussed earlier, this study is an experimental one. In order to fulfill the requirements of an experimental study, all the subjects of the study must be homogenous. A pre-test, namely a TOEFL test (Appendix A) is administered before the beginning of the study. This test was taken from the book Longman Preparation Course for the TOEFL Test: The Paper Test (Philips, 2004). All the students have to achieve the required score to participate in this study. This test includes three parts, listening, vocabulary, and grammar. PBT TOEFL test is divided into three sections: Section 1 -Listening comprehension (40 minutes -50 items), Section 2 -Structure and written expression (25 minutes -40 items) and Section 3 -Vocabulary and reading comprehension (45 minutes -60 items). TOEFL scores for each section are reported on a scale that can range from 31 to 68. The scores are then scaled to give the total score. TOEFL total scores are reported on a scale that can range from 310 -677. The following Table 2 shows the TOFEL scale ranking.

Speaking pre-test
The Speaking test is 16 minutes long and consists of three parts. The standard test format is two candidates and two examiners. One examiner (the interlocutor) conducts the test, providing the student with the necessary materials and explaining what the student has to do. The other examiner (the assessor) will be introduced to the student, but then takes no further part in the interaction (Cambridge, 2013).
Part 1 (2 minutes) -The interlocutor first asks you and your partner a few questions, which focus on information about yourselves. Part 2 (4 minutes) -In this part of the test you and your partner are asked to talk together. The interlocutor places a set of pictures on the table in front of you. There may be only one picture in the set or as many as seven pictures. This stimulus provides the basis for a discussion. The interlocutor first asks an introductory question which focuses on two of the pictures (or in the case of a single picture, on aspects of the picture). After about 1 minute, the interlocutor gives you both a decision-making task based on the same set of pictures. Part 3 (10 minutes) -the students are each given the opportunity to talk for 2 minutes, to comment after his partner has spoken and to take part in a more general discussion.
The interlocutor gives the student a card with a question written on it and asks him to talk about it for 2 minutes. After the student has spoken, the interlocutor asks them both another question related to the topic on the card, addressing the partner first. This procedure is repeated, so that his partner receives a card and speaks for 2 minutes and a follow-up question is asked. Finally, the interlocutor asks some further questions, which leads to a discussion on a general theme related to the subjects already covered in Part 3.

Treatment
Topic Management Strategies were applied to the Experimental Group to explore the influence of Topic Management on the speaking ability of the learners while the other group (Control Group) received no treatment (traditional methods of speaking teaching were applied). The participants in the Experimental Group received training on Topic management Strategies over 10 sessions. The treatment included teaching speaking skill using ordinary teaching materials in speaking as well as topic managing skills. The learners in the Control Group, on the other hand were taught normal speaking lessons over the same period (10 sessions) but with no treatment. At the end of the 10 weeks, both groups took a speaking post-test, the same speaking test that was applied to them before they start the 10 weeks learning. There was a pattern used to conduct the research, which is as follows.
G 1 (Experimental Group) O1 X O2 G 2 (Control Group) O3 O4 In this formula O1 and O3 are the tests before applying the treatment for the experimental and control groups respectively, and O2 and O4 are the tests after manipulating the variable again for experimental and control groups respectively, and X is the treatment.

Speaking post-test
This test was administered to all 40 students at the end of the 10 weeks learning period. The only difference between the two groups in regard to the test is that the Experimental Group took the test after they received the treatment on Topic management while the Control Group took it with no treatment. The format and structure of the post-test is the same of the pre-test that was administered to the students at the begging of the course. This is to see whether the treatment has any effects on the students' speaking ability.

Data Collection and Analysis
The Control Group (20 students) received normal speaking teaching for 10 weeks, twice a week, one hour and a quarter each session. They received conventional classroom instruction and met in a regular classroom the entire time.
The lessons used for the conventional classroom speaking instruction were originally compiled by the teaching staff of Shokouh institute and have been used for years. The second group, which is the Experimental Group, included 20 students. To uncover the influence of Topic Management on the speaking ability of the students, topic managing strategies were taught in addition to the normal speaking teaching to the students over the same period (10 weeks). During intervention, which lasted 10 sessions, the students of this group were taught different kinds of Topic Management strategies such as topic shift, topic initiation, topic termination, and turn taking. All the instructions and interactions during the treatment time (10 weeks) were in English and the students were exposed to different types of Topic Management Strategies and taught skills on how to manage the topic when speaking L2. After 10 sessions of training, the post-test was administered. Two groups of scores were obtained from the learners, the pre-test scores before the treatment and the post-test scores after the treatment. In order to answer the research question, these two groups of scores were analyzed by the SPSS software version 22 th . Two paired sample t-tests and an independent sample t-test were used to analyze the obtained scores. The purpose behind this was to unveil the effect of the independent variable "using Topic Management in conversations" on the dependent variable "speaking ability" of the Iranian intermediate EFL learners.

Evaluation of Overall Foreign Language Proficiency (TOEFL) Test for the Sampling Purpose
The first thing a researcher should pay attention to before applying any treatment to his research participants, is to be sure about their homogeneity (Field, 2009). To check the homogeneity of the research participants, a TOEFL proficiency test was administered to 60 students. This test was used not only to show the homogeneity of the students, but also to determine the students who meet the intermediate level's criteria. The results of the test were analyzed using a descriptive statistics. Table 3 shows the numerical statistics of the TOEFL proficiency test for the 60 students studying English at Shoukoh English institute. This test was administered to find the homogeneity of the students. Therefore, measures of dispersion (the variance, the range, and the standard deviation), central tendency measures (the mean, the median, and the mode), measures of stability or the standard error, and measures of variability (maximum, minimum, and range of values) have been checked and shown in the above-mentioned table. Moreover, measures of the shape of distribution namely Kurtosis and Skewness were measured, respectively. The Kurtosis value of +/-1 is considered as a good point of normality for a test, since the Kurtosis level of the scores is -0.326 it can be said that the population is normal based on the level of Kurtosis. The acceptable value for the Skewness measure is (+/-1 to +/-2), the Skewness value of these 60 students is -1.042. Although this value is not so positive, the normality of this test is acceptable based on the Skewness measure.
As the study focuses on the intermediate level, students above or under the intermediate level measures have been excluded. 40 students who score from 380 to 450 were considered in the intermediate level. 7 students scored more than 450 and 13 other scored lower than 380 points were excluded because they did not meet the intermediate level measures (380 to 450 points). That is, only 40 students who were at the intermediate level were selected to take part in the study (Figure 1).

Examining the Normality Assumption of the Parametric Tests Applied for the Research Question
Before applying this test on the scores of the students, the normality assumption of the tests must have been checked. There are some tests to examine the normality assumption of the data namely, Kolmogorov-Smirnov test, Lilliefors corrects K-S test, Shapiro-Wilk test, Anderson-Darling test, and so on.
The researcher decided to investigate the normality assumption of the students based on the Kolmogorov-Smirnov test. Table 4 shows the statistics of the pre-test and post-test of both groups (Control and Experimental). These statistics show that the distribution was normal, based on the obtained data of Skewness and Kurtosis values. The results of Skewness, which was obtained by dividing the results of Skewness by its standard error showed that the assumption of normality was met in the pre and post-tests of both groups. The Skewness of the pre-test of the control group was 0.618, and it was 0.462 for the post of the same group. The Skewness result of the pre-test of the Experimental Group was 0.459, and the result was 0.338 for the post-test of the same group. All the result fall between -2 to +2, therefore it can be said that the data met the normality assumption based on the Skewness test. The Kurtosis of the pre-test of the Control Group was -0.653, and it was -0.897 for the post-test of the Control Group.
The kurtosis results for the Experimental Group pretest were -0.873, and -0.511 for the post-test. Although, the above-mentioned statistics shows the normality of the assumption, a one-sample Kolmogorov-Smirnov test was applied in order to approve the normality assumption of the scores more accurately. The following table shows the obtained data for one sample Kolmogorov-Smirnov test (abbreviated as K-S test). Table 5 shows the normality assumption of the 40 participants speaking scores. The obtained data for all groups are higher than 0.05, which shows that they have a normal distribution. As the parameters of all four distributions assume, the normality assumption has been met.

Performance of the Participants on Speaking Skill Pre-test
The scores of both groups were analyzed using SPSS version 22th. The results of the independent sample t-test for analyzing the pre-test scores of the students are shown in the following tables. Table 6 illustrates the statistics of both groups clearly. The first row of the table indicates the data related to the control group. As it is obvious, both groups include 20 students. The total score of the test was 10 based on the speaking score scale. The mean score of the Control Group was 5.325, and the mean score of the Experimental Group was 5.385 as it is shown in Table 4. Although, the Experimental Group got the higher mean score, the significance of the groups is not as much different as it seems. The significance of these groups are displayed in Table 7, Figure 2 shows the statics of the table 4.4 more clearly. Table 7 shows the p value is 0.869, and F = 0.313. Since the significance of this test is 0.869 and it is obviously higher than 0.05 that is to say (p = 0.869 > 0.05), consequently it indicates that there is no significant difference between these two groups in terms of their speaking ability. That is to say, both groups are homogenous, and they can go through the requirements of this study, in other terms these groups are comparable and ready to receive the treatments.

DISCUSSION OF THE FINDINGS
The research question of this study deals with the effects of Topic Management Strategies on the speaking ability of the Iranian intermediate EFL learners. An independent sample t-test was used, to analyze the scores of post-test of both groups. The following tables clearly illustrate the results.
As it is shown in Table 8, the speaking skill post-tests of the two groups have been compared since homogeneity was proved at the beginning of the study. In the table, the mean score of the experimental group is 7.100, and the mean score of the control group is 5.880, respectively. As it was mentioned in Table 4.4, the pre-tests' mean score of the Experimental Group was 5.385, and the mean score of the Control Group was 5.325. It should not be ignored that both groups had 20 participants, and the experimental group received Topic Management Strategies training as part of the study. Obviously, the mean score of both groups have improved. The researcher analyzed both groups mean scores to see which factor has the better effect on speaking ability of the Iranian intermediate EFL learners. Since the  Experimental Group got the higher mean score, it seems that Topic Management Strategies had a better effect on the speaking ability of these participants. Therefore, it can be said that the null hypothesis of the study is rejected. However, to prove this result more accurately the p and t value should be checked as well, which is mentioned in the following table. Table 9 explains the independent samples t-test of the speaking ability post-test of both groups. The level of significance (that is 0.927) is clearly bigger than 0.05, so it obviously shows that the variances are equal. Besides, the significance level is 0.000, which is smaller than 0.05 (that is = p value 0.000 < 0.05). In the column of 95% confidence interval of the difference, zero is not included (that is -0.573 and -1.866). So, these groups do not have similar effects on speaking ability of Iranian intermediate EFL learners, and there is a meaningful difference between the populations. Based of this, it could be said that the null hypothesis of the current study is not proved, and that topic management strategies have a great influence on speaking ability of the Iranian intermediate EFL learners. Figure 3 shows the difference between the mean score of the Experimental and Control groups post-test. Obviously,  the mean score of the Experimental Group is higher than the mean score of Control Group. This means that Topic Management Strategies have a noticeable influence on the speaking ability of tested participants.

CONCLUDING REMARKS
To conclude, this study aimed to explore the effect of Topic Management on the speaking ability of Iranian intermediate EFL learners. The results show that both the Control and Experimental groups have (to some extent) improved their speaking abilities after the treatment. Nevertheless, the participants of the Experimental Group who underwent the treatment achieved more scores in their speaking post-test compared to the Control Group. (Figure 3 clearly explains the difference between the mean score of the speaking posttest of both groups.) These findings are in line with Sacks (1992) and Herman's (1998) argument that Topic Management has a remarkable influence on the speaking ability of learners. Herman (1998) argues that students with a high proficiency level are more successful in using Topic Management. The findings support this argument and further add that intermediate level learners can also handle Topic Management Strategies successfully in their conversation to improve their speaking ability. Erber (1996) posts that Topic Maintenance helps speakers handle conversations smoothly and successfully and that Topic Management Strategies have positive effects on the speaking ability of learners. According to him, Topic Maintenance is deemed to be more effective since it does not need new information. The findings of this study support this argument and further add that Topic Management Strategies not only improve the speaking ability of students, but they also improve other learning skills such as writing and listening.
The findings of this study also support the contentions of Jones (2008) and Schegloff (1992) stating that Topic Management is a social behavior that helps learners build their self-esteem. It (the study) found that controlling over Topic Management as a skill can effectively influence not only on the speaking ability of learners, but also increase their confidence when they practise speaking in the foreign language.
Finally, due to the scope and size of the study, the results obtained cannot be generalized to large populations and hence, further detailed investigation into the topic is advised.