Attrition of Oral Communicative Ability among English Language Graduates in Turkey

This study explored the relationship between the period of Turkish EFL (English as a Foreign Language) graduates’ not using English as well as their ages with the attrition of their oral communicative ability. The study also indicated the language maintenance strategies the Turkish EFL graduates pursue to prevent the attrition of their oral communicative ability. To investigate the issues, the present study used a quantitative design and the convenience sampling. The participants included 153 Turkish graduates majoring in English fields. A test of oral communicative ability as well as a checklist including 20 ‘Language Maintenance Strategies’ were used to collect the required data. The results showed that the longer the period of language non-use, the more likely the attrition of oral communicative abilities of Turkish EFL graduates will occur. As for the language maintenance strategies, the item ‘watching movies with its corresponding subtitle in English’ was the most frequent, followed by ‘listening to English songs’, whereas ‘speaking in English with Turkish friends’ was the least frequent one, and this was followed by ‘speaking in English with foreigners’. The study highly recommends enhancing the conditions and quality of the EFL teacher education programs in Turkey.

Considering the mentioned context of Turkey, English will probably be forgotten once it is not used or studied anymore by the EFL graduates because literature has indicated that the most common causes of FL attrition are the strength and quality of FL learning, motivational factors causing acquisition, and the way the learned language is used (Lambert & Freed, 1982;Weltens & Cohen, 1989). Further, productive skills are more prone to attrition than receptive ones (Al-Sulaiman, 2020; Hakuta & D'Andrea, 1992;Weltens & Grendel, 1993). However, the Turkish researchers have been more concerned with English language use than language loss so far. In Turkey, English language attrition is an under-investigated area of research.
Although some scholars believe in the occurrence of language loss or attrition, De Bot and Weltens (1995) suggested that when a language is acquired, it is never lost. What happens is that when a foreign language is not studied or used for a period of time, the foreign language elements in knowledge may not be readily available (De Bot & Weltens, 1995). There can be some remedies for language attrition (Neisser, 1984;Szupica-Pyrzanowska, 2016).
Therefore, Kopke (2007) argued that "the issue of language use and contact certainly needs to be investigated further in the context of attrition" (p. 25). Furthermore, as Schmid and Mehotcheva (2012) argued, in a society which people spend time, money, and effort on learning a FL, understanding about what causes and contribute to FL attrition is really important. Moreover, as Schmid and Mehotcheva (2012) put it, "research on FL attrition can have implications not only directly for FL teaching and learning but also for more theoretical linguistic aspects." (p. 103). Additionally, as Cohen (2018) pinpointed, more energy is spent helping learners learn language than to maintain what they have learned, and he recommended that teachers and learners be alerted to strategies for preventing attrition.
Some researchers have already explored the problem of first language (L1) (i.e., Turkish) attrition among the Turkish people (Karayayla & Schmid, 2019;Kasap, 2015;Yildiz & Koyuncuoglu, 2017). Nonetheless, to the best knowledge of the researcher of the current study, no study has been published exploring the English language attrition among the EFL graduates in Turkey. As a result, this study attempts to partially fill in this research gap in Turkey. The present study addresses the following questions: 1. Is there a statistically significant correlation between the period of English non-use and the attrition of the oral communicative ability of Turkish EFL graduates? 2. Is there a statistically significant correlation between the age and the attrition of the oral communicative ability of Turkish EFL graduates? 3. What are the most used techniques for preventing the attrition of the oral communicative abilities of Turkish EFL graduates?

Research Design
This research has a quantitative correlational design. The convenience sampling was utilized in this study.

Participants
The 153 participants in this study included 96 female and 57 male Turkish EFL graduates who lived in Turkey. Their native language was Turkish, and their L2 was English. They all held bachelor's degree in Applied Linguistics, Translation Studies, English Literature, Linguistics, and other majors. The mean age of the subjects was (33.75), ranging from 24 to 51 years. The time lapse between their graduation and participation in this study ranged between 1 to 29 years. None of them had been teaching English during the mentioned time lapse. Table 1 indicates a summary of participants' background characteristics

Instruments
A test of oral communicative ability was used in this study.
The questions were compiled from Levels 2, 3, and 4 of the book Real Listening and Speaking by Logan and Craven (2008) as well as (Craven, 2008a(Craven, 2008b respectively. For more information on the used test, see Appendix A. The Audioscript of the questions are also provided in Appendix B. Moreover, among the "Sample Assessment Rubrics" of New York State Education (n.d.), the Spontaneous Conversation Rubric was utilized to assess the participants' performance on the oral communicative ability test. However, the rubric was modified by omitting the Vocabulary section to match the requirements of the test. For more information on the used rubric, see Appendix C. Additionally, a checklist, which included 20 'Language Maintenance Strategies' was prepared by the Google Forms; the list was prepared using the ideas gotten from Ostovar-namaghi and Rahmanian (2017) as well as Al-Sulaiman (2020). The mentioned checklist is available in Appendix D. Further, the Skype computer application was used to administer and record the oral communicative ability test.

Data Collection Procedure
After the oral communicative ability test was prepared, the researcher searched the Turkish EFL graduates out via LinkedIn and Facebook. Moreover, the email addresses of several EFL graduates were obtained either from their universities or via their friends. The invitation to participate in the study was sent to 201 Turkish EFL graduates. A hundred and sixty-three people consented to take part and the researcher guaranteed their anonymity. Ten out of 163 subjects agreed to participate in the test-retest process to validate the oral communicative ability test, and the rest (153 subjects) participated in the study. The prepared test of oral communicative ability was administered twice (after a lapse of 12 days) to the group of ten individuals (Hatch & Lazaraton, 1991); subsequently, the reliability of the test was measured running the Pearson correlation. The time lapse of 12 days was decided based on Henning (1987, as cited in Hatch & Lazaraton, 1991), which recommended that "the time lapse should be less than two weeks" (p. 532).
Then, the date and time of doing the test were scheduled at each participant's convenience. Afterwards, the researcher connected the participants individually via Skype and administered the test. The duration of the whole process varied among the participants, but the maximum duration was nearly 25 minutes. The variation was mainly because of the differences in participants' fluency in spoken English. Each subject's participation was recorded for further analyses.
Following that, the link of the form, entitled 'Language Maintenance Strategies', was sent to the participants either by email or their WhatsApp accounts at each participant's convenience. They checked the language maintenance strategies which they either pursued or would like to adopt and submitted the list to the researcher.
In order to prevent the possibility of the researcher's bias and considering the rater reliability, each oral communicative ability test was scored independently by two raters: the researcher and a colleague who held master's degree in Applied Linguistics and was a professional researcher. Each rater used the mentioned rubric and graded the participants' performance out of 20. The final score consisted of the average score of the two raters.

Estimating the Reliability
The Cronbach alpha coefficient, utilized to measure the reliability of the oral communicative test as well as the interrater reliability, were .991 and .993 respectively.

Correlation between the Period of English Non-use and the Attrition of the Oral Communicative Ability
To investigate the probable relationship between the Turkish EFL graduates' period of English non-use and their attrition of the oral communicative ability, first the preliminary analyses were performed to ensure no violation of the assumptions of normality, linearity, and homoscedasticity.; the value of kurtosis statistics was -1.128, so it was not within +/-1; in addition, the result of the Shapiro-Wilk test was .000, so it was lower than .05. It was concluded that the assumption of normality was violated based on Thode (2002), Larson-Hall (2010), Phakiti (2010), and Pallant (2016). As a result, Spearman's rho was utilized to explore the correlation between the period of English non-use and the attrition of the oral communicative ability.
There was a strong negative correlation between the period of English non-use and the Turkish EFL graduates' oral communicative ability total score, r = -.753, n = 153, p = .000 < .01 (Cohen, 1988). In other words, the more years passed since the participants' graduation, the more impaired their level of oral communicative ability was.

Correlation between the Age and the Attrition of the Oral Communicative Ability
The Spearman's rho was conducted to seek for the probable existence of the correlation between the age and the attrition of the oral communicative ability among the Turkish EFL graduates.
There was a strong negative correlation between the Turkish EFL graduates' age and their oral communicative ability total score, r = -.725, n = 153, p = .000 < .01 (Cohen, 1988). In other words, the older they were, the more impaired their level of oral communicative ability was.

The language maintenance strategies used by the Turkish EFL graduates
To find out what techniques the Turkish EFL graduates use or would like to adopt to prevent the attrition of their language ability, the frequencies and percentages of the items related to language maintenance strategies were calculated. Table 2 shows the results. As Table 2 shows, the item 'watching movies with its corresponding subtitle in English' was the most frequent (66.0%), followed by "listening to English Songs (58.2%). 'Speaking in English with my friends in my country' was the least frequent one (26.8%), and this was followed by 'speaking in English with foreigners' (29.4%).

DISCUSSION AND CONCLUSION
This study found strong negative correlations between years since graduation and the Turkish EFL graduates' oral communicative ability as well as their ages and their mentioned ability. These findings are consistent with some of the similar previously done studies worldwide in other contexts, such as Gardner et al., (1987) and Al-Sulaiman, (2020).
However, generalizing the findings from the earlier studies on FL attrition to the current one might be impossible due to the fact that the characteristics of the language pedagogy and learning method used in the context of Turkey as well as the Turkish EFL graduates' social and individual differences could affect the attrition of their English oral communicative ability. As was stated in the Introduction, in Turkey, English is taught exclusively in formal settings of classrooms where there is limited authentic input. Speaking and use of the language is also limited to the classroom and each individual can have limited opportunities for authentic real-life use of English (Coşkun, 2016;Dinçer & Yesilyurt, 2013;Gençoğlu, 2011;Güney, 2010;Karaata, 1999;Özsevik, 2010;Toköz-Göktepe, 2014).
Furthermore, some similarities have been found between the language maintenance strategies which Turkish EFL graduates adopt or would like to utilize, and the ones employed by the subjects in Ostovar-namaghi and Rahmanian's (2017) as well as Al-Sulaiman's (2020) studies. As for the context of Turkey, this study showed that speaking in English with friends in Turkey was the least frequent one (26.8%), and this was followed by 'speaking in English with foreigners' (29.4%). Based on the previously done studies, the reason can be their anxious feeling while speaking English, (Coşkun, 2016;Dinçer & Yesilyurt, 2013;Öztürk & Gürbüz, 2014;Savaşçı, 2014;Tok, 2009), and lack of knowledge about English culture (Coşkun, 2016). Moreover, the old-fashioned grammar-based teacher-centered English teaching approach can definitely cause the challenges of speaking English in Turkey (Coşkun, 2016;Dinçer & Yesilyurt, 2013;Gençoğlu, 2011;Güney, 2010;Karaata, 1999;Özsevik, 2010;Toköz-Göktepe, 2014).
Further, as the literature revealed, English teachers in Turkey are not entirely familiar with the Communicative Language Teaching (CLT) and follow the traditional teacher-centered grammar teaching method (Gençoğlu, 2011;Güney, 2010;Kirkgoz, 2007;Özsevik, 2010;Uysal & Bardakçı, 2014). Therefore, if the English teachers in Turkey perfectly learn about the CLT approach and do not follow the traditional teacher-centered grammar teaching method, the conditions can improve.
Additionally, lack of skilled fluent English teachers and quality education to train them (Sak, 2012), as well as use of Turkish language instead of English by the teachers in EFL classes (Coşkun, 2016) were mentioned as two key factors affecting the English speaking ability of the Turkish students and people. In addition, literature already revealed the weaknesses of EFL teacher education in Turkey. The weaknesses include lack of "a clear-cut philosophy of teacher education" (Karakaş, 2012, p. 8), lack of "systematic planning and scientific research on training needs of teachers due to the non-functional organizational structure and under qualified personnel of the Ministry of Education's in-service training department." (Uysal, 2012, p. 19), an absence of "a comprehensive, current, and consistent conceptual framework that is informed by current L2 learning and teaching and teacher education research … Second, there seems to be a lack of focus on a background in linguistics and SLA" (Mahalingappa & Polat, 2016, p. 8), not following a holistic and an experiential approach, in which a variety of techniques are applied (Bayrakcı, 2009;Karakaş, 2012), not supporting the Turkish teachers of English by any feedback or evaluation system (Bayrakcı, 2009;Özer, 2004;Uysal, 2012), the limited hours allocated to practically oriented courses, such as teaching practice (Coskun & Daloglu, 2010;Sanli, 2009;Seferoğlu, 2006), and not including "a reflective practice component" (Karakaş, 2012, p. 10). Consequently, if the EFL teacher education programs in Turkey improves, the quality of the EFL teachers' teaching, as well as the oral communicative ability of the Turkish EFL learners improve, which may also reduce the probability of language attrition.
Even though this study is informative about the attrition of oral communicative ability among Turkish EFL graduates, there is no denying that it has some shortcomings and further research is needed to enlighten the EFL experts in the context of Turkey on how to pursue the mentioned matter. For example, a bigger sample size will contribute to more generalizability in the results. Additionally, further research is needed to discover the various internal or external factors of attrition, other than age and language non-use.

1) Socializing a) Starting a Conversation
Imagine you meet 4 new people at a party. Listen and answer their questions. b) Maintaining a Conversation and Asking Follow-up Questions) Imagine 5 people talk to you at the party. Listen to their statements (a-e). Respond with follow-up questions.

2) Eating Out
(Talking about what you ate at a restaurant) Think about the last time you went to a restaurant. Listen and answer the 7 questions.

3) Travel a) Offering Travel Tips
Imagine you are in your hometown and a tourist asks you 6 questions (a-f). Listen and reply, using your own ideas. b) Asking for Detailed Information to Book a Trip Imagine you are on holiday in India and an Indian tour guide is trying to sell you an excursion. Using the prompts below (a-e), ask questions, and then you will hear the tour guide's response to each of your questions.