ENGLISH REVIEW: Journal of English Education Volume 7, Issue 2, June 2019 p-ISSN 2301-7554, e-ISSN 2541-3643 https://journal.uniku.ac.id/index.php/ERJEE 27 NATIVE ENGLISH SPEAKERS VERSUS NON-NATIVE ENGLISH SPEAKERS: THE IMPACT OF LANGUAGE TEACHERS ON EFL LEARNER'S ENGLISH PROFICIENCY Tariq Elyas European Languages Department, Faculty of Arts & Humanities, King Abdul-Aziz University, Saudi Arabia E-mail: telyas@kau.edu.sa Noor Motlaq Alghofaili English Language Institute, King Abdul-Aziz University, Saudi Arabia E-mail: noor.alghofaili@gmail.com APA Citation: Elyas, T., & Alghofaili, N. M. (2019). Native English speakers versus non-native English speakers: The impact of language teachers on EFL learner’s English proficiency. English Review: Journal of English Education, 7(2), 27-38. doi: 10.25134/erjee.v7i2.1773. Received: 23-01-2019 Accepted: 25-03-2019 Published: 01-06-2019 Abstract: In the field of TESOL, the perception that Native English Speaking Teachers (NESTs) are better than Non-Native English Speaking Teachers (NNESTs) has influenced language schools, recruitment policies and institutional leadership practices. The tendency to recruit more NESTs and achieve improved learning outcomes can be seen in English as a Foreign Language (EFL) and English as a Second Language (ESL) contexts. This paper aims to investigate whether NESTs or NNESTs have any impact on the EFL learners’ language proficiency in Saudi EFL context. This quantitative study adopts pretest-posttest experimental and ex post facto designs to determine students’ achievement in two language skills, namely speaking and listening. The two groups of participants are EFL students in a foundation year program at a Saudi Arabian University. One group was taught by a NEST and the other by a NNEST. The quantitative data were analyzed by using SPSS. The findings indicated that teachers’ nativeness and backgrounds have no significant effects on the Saudi EFL learners’ speaking and listening skills. Here, Saudi EFL learners can equally perform in classes taught by NESTs or NNESTs. In the light of the findings, the study suggests that recruitment policy should not be influenced by the employers’ belief that NESTs possess better teaching skills than NNESTs. Keywords: EFL learners; experimental research; ex post facto design; language proficiency; NESTs; NNESTs. INTRODUCTION There has been a debate over the past few decades to determine whether native English speakers or non-native English speakers are better language teachers. Teaching of English in different contexts have been linked to the teacher native or native speaking abilities, and it is believed that teachers who speak English as their first language can be more effective classroom practitioners than those whose first language is not English. However, research on the phenomenon of native English-speaking teacher (NEST) and non-native English speaking teacher (NNEST) indicates the fact that the latter can be more effective owing to their experience of being a language learner. As the debate continues in the field of TESOL, this paper considers the matter from the EFL learners’ perspective in the Saudi context. The last two decades have seen exponential growth in the number of research studies that aimed to determine the characteristics of an ideal language teacher (Al-Nawrasy, 2013; Alghofaili & Elyas, 2017). It is commonly believed that NESTs are ideal language teachers (Moussu, 2010). Phillipson (1992) even labels the notion of idealizing NESTs as the ‘native speaker fallacy’ (cited in Braine, 2013). This belief has shaped the idea of hiring NESTs in schools, language centers, and even universities regardless of their teaching qualifications and experience (Alseweed, 2012). Moreover, the employers prefer NESTs over NNESTs believing the formers have knowledge of how language works (Larsen-Freeman & Anderson, 2011). This preference for NESTs is evident in most job advertisements around the world, especially in non-English speaking context, such as Arabian Gulf, Korea and China (Selvi, 2010). According to Li-Yi (2011), in Taiwan, parents consider NNESTs incompetent due to their lack of overseas experience, regardless of their qualifications. The impact of the employers’ Tariq Elyas & Noor Motlaq Alghofaili Native English speakers versus non-native English speakers: The impact of language teachers on EFL learner’s English proficiency 28 choice of NESTs can be seen in fewer job opportunities for NNESTs, whereas there is a big demand for NESTs around the world. For example, in Thailand in 2010, there were 500 native-speaker teaching job vacancies (Grubbs, Jantarach, & Kettem 2010). Similarly, Anya, Avineri, Carris, and Valencia (2010) argue that hiring decisions might be influenced by the perception that some people speak the language “correctly” while others do not. These evidences raise a question of whether native speakers are employed for the sake of their nativeness, irrespective of their teaching experience or training. In this case, Chun (2014) suggests that native speakers should not be employed for simply being native and should not be preferred over NNESTs. As there is no solid empirical evidence to show whether NESTs or NNESTs can contribute to the learning outcomes of EFL learners, more investigation is required. Saudi government invests a large amount of budget in EFL education, as English language has a significant role in the development of business, science and technology. According to Mahboob and Elyas (2014), English fluency has economic value in Saudi Arabia due to the large number of foreign companies that contribute to the economy of the Kingdom. As a result, the number of English language learners has rapidly increased the demands for qualified EFL teachers. Due to its importance in various fields and mandatory consideration at school and university levels, both NESTs and NNESTs are recruited by public and private sector universities for the Preparatory Year Programme (PYP). Although most of the Saudi EFL institutes prefer to employ NESTs, NNESTs from different Arab and Asian countries also teach at PYP. The large number of language teachers in Saudi higher education institutions gives rise to a question whether instructional practices of NESTs or NNESTs have any impact on the language profanely of Saudi EFL learners. Since learning English has become a requirement in higher education, Saudi students’ attitudes about NESTs have evolved to the point where many consider learning from a NEST an ideal way to attain English language proficiency (Alseweed & Daif-Allah, 2012). Despite its significance, there is dearth of empirical evidence on the issue of NESTs and NESTs in the Saudi EFL context. Therefore, this study aims to bridge that research gap. Listening skill Listening is one the most important skills that language learners acquire in a classroom environment which also contributes to their speaking abilities. For language teachers to help their students become effective listeners, it is essential to apply listening strategies and provide listening practice in authentic situations. Research shows that language learners generally prefer NESTs to teach listening, considering them an ideal model of the authentic language in a classroom setting (Al-Omrani, 2008; Lasagabaster & Sierra, 2005; Mahboob, 2004; Nam, 2010). In Mahboob’s (2004) view, the possible reason behind this preference could be that NESTs acquire a “natural” pronunciation that might help students improve their listening and contribute to their speaking ability. In a similar way, Lasagabaster and Sierra (2005) believe that the NESTs’ ability to speak better English than NNESTs may offer learners opportunities to acquire authentic language and improve their listening skills. Speaking and pronunciation skills Many language learners regard speaking ability as an indicator of knowing a language. These learners consider fluency more important than the ability to read, write, or comprehend oral language. As they regard speaking the most important skill that needs to be acquired, their preference is always to have a fluent teacher of English. Given these presuppositions, it would not be surprising if learners show interest in NESTs owing to their accents and backgrounds. In fact, this could be one major reason for idealizing the native-speaking teacher. Research in this direction shows that language learners prefer NESTs over NNESTs in teaching speaking skills (e.g. Al-Omrani, 2008; Kelch & Santana-Williamson, 2002; Lasagabaster & Sierra, 2005; Mahboob, 2004). According to Al- Omrani (2008), NESTs are rated higher in teaching speaking skills due to their accurate pronunciation and fluency. Similarly, Benke and Medgyes (2005) reveal that NESTs are good sources of inspiration for language learners to use English well; they are considered as perfect models for imitation. Nevertheless, such claims of idealizing NESTs in teaching speaking are largely based on students’ perceptions, the reliability of which can be difficult to assess. Since students prefer NESTs as speaking teachers, the impact of NESTs on learners’ speaking skills is an important consideration for researchers. Al-Nawrasy (2013) used an ex post ENGLISH REVIEW: Journal of English Education Volume 7, Issue 2, June 2019 p-ISSN 2301-7554, e-ISSN 2541-3643 https://journal.uniku.ac.id/index.php/ERJEE 29 facto design to investigate the impacts of a teacher’s nativeness on students’ achievement in speaking skills. The results show that there is no significant difference in speaking skill achievements between students taught by the two types of teachers; NEST and NNEST. Although the NESTs’ learners display better performances with respect to pronunciation, the NNESTs’ students were better in terms of accuracy. Research in this direction indicates that both NESTs and NNESTs can have impact on the English language learners’ pronunciation. Levis, Sonsaat, Link and Barriuso (2016) investigated the effects of teachers’ L1 on learners’ pronunciation. Even though many learners stated their preference for native teachers, their actual results offer encouragement to nonnative teachers in teaching pronunciation. This suggests that instruction on pronunciation skills is dependent on knowledge and teaching techniques than on the native pronunciation of teachers. These two experimental studies have led to new findings that run counter to the results typically found when comparing NESTs and NNESTs in teaching speaking and pronunciation skills and using students’ perceptions as the basis of study. A plethora of research shows that NESTs are perceived as being superior in teaching English language and helping learners to improve their pronunciation skills (e.g. Alseweed, 2012; Chun, 2014; Grubbs, Janatra, & Kettem 2010; Ma, 2012a; Walkinshaw & Dungo, 2012; Walkinshaw & Oanh, 2014). On the other hand, some studies argue that NESTs are difficult to understand since they speak too fast and use words that are often difficult for the level of their students (e.g. Ma, 2012; Sung, 2010). Although learners show preference for NESTs, interestingly, the teachers’ background and nativeness have no significant impact on students’ achievements in speaking and pronunciation skills (e.g. Al-Nawrasy, 2013; Levis, et al. 2016). METHOD When investigating the impact of NESTs and NNESTs on students, using different methods can offer useful insights to provide a comprehensive picture of the research problem. Two instruments were used in this study: pretest- posttest and ex post facto. The quantitative methods—pretest/posttest and ex post facto designs—were used to investigate the influence of the teachers’ nativeness, background, and accent on their students’ achievement. Students were given a pretest to ensure comparability of the participant groups prior to their exposure to NESTs or NNESTs and a posttest to measure the effects of that exposure. To ensure comparability, the same test was given in both pretest and posttest, since the time elapsed between the two tests was long enough (7 weeks) that the students were unlikely to remember their answers in pretest. The reading part was an achievement test taken from the teacher’s version of the English Unlimited Special Edition B1, which is assigned to intermediate level students. The reading part included two reading passages: the first was a message posted on a website that offers job and career advice to young people and the other was an email posted on a website about hotels. Each passage included five truth-value judgment sentences, which have been extensively used with second-language learners to eliminate fatigue and avoid compromising the reliability of the study (Mackey & Gass, 2005). As for listening part, it was also an achievement test taken from the teacher’s version of English Unlimited Special Edition B1. Due to the time limitation and to avoid students’ boredom and fatigue, the listening part was a radio interview with two people. Similar to the reading questions, it included five true judgment sentences. Further, in ex post facto design, also called causal-comparative design, the independent variable or variables have already happened and the investigator starts with the observation of a dependent variable or variables (Kerlinger, 1970) cited in (Cohen, Manion & Morrison, 2007). Here, the researcher investigates possible relationships to and effects on the dependent variables by studying the independent variable, which in this case is teacher nativeness (Cohen, Manion & Morrison, 2007). In other words, the researcher is examining retrospectively the effects of a naturally occurring event on a subsequent outcome with a view to establishing a causal link between them. This design focused first on the effect and attempted to determine what caused the observed effect. Due to the difficulty of using pretest/posttest design to measure the impact of NEST and NNEST on students’ writing and speaking skills, ex post facto design was used. As stated by Cohen, Manion and Morrison (2007), it can be useful to employ an ex post facto design in Tariq Elyas & Noor Motlaq Alghofaili Native English speakers versus non-native English speakers: The impact of language teachers on EFL learner’s English proficiency 30 situations where conducting experimental research is not possible. This design was chosen for various reasons. First, the pretest/posttest approach is time-consuming when evaluating such skills. Second, since the ELI offers an intensive English language course in six weeks, it would be difficult to test the students individually, particularly in speaking skills. Moreover, the ELI students are obliged to take highly valid and reliable final speaking and writing tests at the end of the module to pass the course. Finally, students might be unwilling, anxious or too tired to participate in taking tests in these two skills. These factors might affect the results of the test. This study was conducted at The English Language Institute (ELI) of a Saudi Arabian university. At the ELI, more than 8000 students are enrolled in a preparatory year program (PYP) every year, during which they have to complete four levels of English courses as follows: Table 1. English language courses at the ELI ELI COURSE CODE COURSE LEVEL CEFR LEVEL CREDITS 101 Beginner A1 0 102 Elementary A2 2 103 Pre-intermediate B1 2 104 Intermediate B1+ 2 ELI helps students in their PYP to achieve an intermediate level of English proficiency, equivalent to the Common European Framework Reference of B1 threshold level (CEFR B1). All participating students were homogeneous in terms of age (they were either 18 or 19 years old), native language (Arabic), nationality (Saudi), and cultural background. Two participant groups were included in this study. The first group included the participants who participated in the quantitative approaches. In this group, stratified sampling was used, in which, as Creswell and Clark (2017) indicate, researchers divide (stratify) the population on some specific characteristic (e.g., gender) and then, using simple random sampling, sample from each subgroup (stratum) of the population (e.g., females and males). This guarantees that the sample will include specific characteristics that the researcher wants to include in the sample.” In this study, two classes were chosen to take pretest and posttest of reading and listening skills: the first one was taught by a British NEST and the other one a Saudi NNEST. These two classes were also used for the ex post facto portion of the study. Both classes were at intermediate level, which falls under the B1 Threshold level according to the CEFR scale. The second group had 18 student participants, eight in intermediate and 10 in upper intermediate. All students were taught by both types of teachers, native- and non-native- speaking. The participants’ names were given pseudonyms to protect their privacy. Background information of participated students is presented in Table 2. Table 2. Students' background information Participants’ Pseudonym Level Number of NESTs taken courses with Nationalities of NESTs taken courses with Number of NNESTs taken courses with Nationalities of NNESTs taken courses with Wajd 103 3 American, British, Canadian 2 Saudi, Malaysian Maryam 103 2 American 4 Saudi, Indian, Pakistani Ahlam 103 2 American, British 2 Saudi, Egyptian Basmah 103 1 American 4 Saudi, Pakistani, Jordanian Mona 103 2 American 4 Saudi, India Maha 103 1 American 2 Saudi Laila 103 1 British 2 Saudi, Indian Iman 103 2 British, Canadian 2 Saudi, Malaysian Samirah 104 1 American 4 Saudi, Indian Sanaa 104 1 American 3 Saudi Boshra 104 1 American 4 Saudi, Pakistani, ENGLISH REVIEW: Journal of English Education Volume 7, Issue 2, June 2019 p-ISSN 2301-7554, e-ISSN 2541-3643 https://journal.uniku.ac.id/index.php/ERJEE 31 Jordanian Bodour 104 1 American 4 Saudi Syrian Shayma 104 1 American 4 Saudi, Malaysian, Egyptian Hana 104 1 American 3 Pakistani, Saudi Kholoud 104 1 American 4 Saudi, Syrian, Turkish, Jordanian Wafaa 104 1 American 5 Saudi, Jordanian Amal 104 2 American 5 Saudi, Jordanian, Syrian Asmaa 104 1 American 3 Jordanian, Egyptian, Syrian The ELI uses a modular system of four quarters per academic year. The quantitative data for this study were collected during the third module of 2015–2016 academic year. Approval to conduct the study was obtained from the Head of the Research Unit at the ELI. Consent forms were signed by all students participated in the study. The participants knew their participation was voluntary and they could withdraw from the study at any time. Confidentially and anonymity were assured to all participants. As noted above, the pretest and posttest were conducted during the third module of intermediate level students. The test was piloted on the first day of the module with three ELI students at that level to determine the time needed for answering the questions. During the first week of the third module, the pretest was given to 18 students who attended a NEST class and 20 students who were present in an NNEST class. The test had to be completed within 40 minutes. Six weeks later, with a high absenteeism range, the posttest was given to 14 students who took the pretest and attended both the classes. The pretest and the posttest in both classes were printed, distributed, invigilated, and collected by the teacher to ensure the safety of the data. As for the ex post facto design, the students’ grades of the same NEST and NNEST classes in speaking and writing skills from module two were collected. The ELI grading sheets were provided by the academic coordinator, which were compared with their current third-module grades of speaking and writing skills. The number of students whose grades were accessible was 18 students in NNEST class and 17 in NEST class. The Statistical Package for Social Studies version 20.0 (SPSS, IBM) was used to calculate and find any significant statistical difference between the mean scores of the students who were taught by NESTs and those who were taught by NNESTs. The questionnaire data were translated by the researchers and coded to identify common themes. To achieve the highest degree of accuracy in comparing the two groups’ grades, several statistical analyses were applied, including descriptive statistics and tests of normality. Paired sample t-test analysis was also applied to determine whether there was statistical evidence that the mean difference between paired observations pretest and posttest for each group was significantly different from zero (i.e., that both means were not equal). Additionally, Pearson’s correlation was used to measure the strength and direction of the association that existed between two variables measured (pretest, posttest) for each group in the study. For further investigation, an independent samples t-test was applied to check if there was a significant difference between the pretests of both groups and if there was a significant difference between the posttests of both groups. RESULTS AND DISCUSSION This section which reports on the data obtained from the two quantitative methods used is divided into two subsections; listening skill and speaking skill. Listening skill Table 3 shows descriptive statistics for students’ achievements in listening skills of both NEST and NNEST groups. In addition, Figure 1 shows the histograms (pre-test and post-test) of the grade distributions of each listening group. Tariq Elyas & Noor Motlaq Alghofaili Native English speakers versus non-native English speakers: The impact of language teachers on EFL learner’s English proficiency 32 Table 3. Descriptive statistic for students’ achievement in the listening skill Listening Group pre-test post-test NEST N Valid 14 14 Missing 0 0 Mean 6.1429 4.7143 Std. Error of Mean .64499 .77338 Std. Deviation 2.41333 2.89372 Variance 5.824 8.374 Range 6.00 8.00 Minimum 2.00 .00 Maximum 8.00 8.00 NNEST N Valid 14 14 Missing 0 0 Mean 3.0000 5.7143 Std. Error of Mean .58366 .65824 Std. Deviation 2.18386 2.46291 Variance 4.769 6.066 Range 8.00 8.00 Minimum .00 2.00 Maximum 8.00 10.00 Figure 1. Grade distributions of NEST and NNEST groups in the listening pre/post tests The above curves display the normal distribution, and the Shapiro-Wilk test showed that the data is normally distributed for most listening groups (p>0.05). The pre-test of the NEST group is not normally distributed (p>0.05) (Table 4) but that would be accepted as the data displayed above in the histograms is accepted. Table 4. Test of normality for the listening skill test Listening Group Kolmogorov-Smirnova Shapiro-Wilk Statistic Df Sig. Statistic Df Sig. NEST pre-test .351 14 .000 .741 14 .001 post-test .229 14 .045 .853 14 .025 NNEST pre-test .248 14 .020 .892 14 .087 post-test .257 14 .013 .882 14 .061 a. Lilliefors Significance Correction ENGLISH REVIEW: Journal of English Education Volume 7, Issue 2, June 2019 p-ISSN 2301-7554, e-ISSN 2541-3643 https://journal.uniku.ac.id/index.php/ERJEE 33 This part of the study is based on the following null hypotheses:  NH3: NEST has no impact on student achievement in listening skill.  NH4: NNEST has no impact on student achievement in listening skill. To test the above hypotheses, a dependent t- test paired sample was applied to determine whether there is any statistical evidence that the mean difference between paired observations pre-test and post-test for each group is significantly different from zero. Table 5, Table 6 and Figure 2 show that for the NEST group, the difference between the pre- test and the post-test (1.43) is not significantly different from zero. That is, both grades are approximately equal in pre-test (6.14±2.41) and post-test (4.71±2.89). Therefore, the difference between the mean of the two grades is not large enough to reach the significant level (t(13)=1.408, P=0.183). However, the post-test mean is slightly lower. The third null hypothesis is accepted. As for the NNEST group, the difference between the pre-test and post-test (- 2.714) is significantly different from zero. That is, both grades are not equal pre-test (3.00±2.18) and post-test (5.71±2.46). Therefore, the difference between the mean of the two grades is significant (t(13)=-2.723, P=0.017), and the post-test mean is significantly higher. Therefore, the fourth null hypothesis is rejected. Table 5. Paired sample statistics for students’ achievement the listening skill Listening Group Mean N Std. Deviation Std. Error Mean NEST Pair 1 pre-test 6.1429 14 2.41333 .64499 post-test 4.7143 14 2.89372 .77338 NNEST Pair 1 pre-test 3.0000 14 2.18386 .58366 post-test 5.7143 14 2.46291 .65824 Table 6. Paired sample test for students’ achievement the listening skill Listening Group Paired Differences T Df Sig. (2- tailed) Mean Std. Deviation Std. Error Mean 95% Confidence Interval of the Difference Lower Upper NEST Pair 1 pre-test - post-test 1.42857 3.79705 1.01480 -.76378 3.62092 1.408 13 .183 NNEST Pair 1 pre-test - post-test -2.71429 3.72989 .99686 -4.86786 -.56071 -2.723 13 .017 Figure 2. Paired sample test for students’ achievement in the listening pre/post tests Pearson’s correlation (Table 7) was used to measure the strength and direction of the association between the two variables measured (pre-test, post-test). The following table shows that r=-0.016 and p=0.957 for the NEST group, and r=-0.286 and p=0.322 for the NNEST group, indicating that the correlations are weak and insignificant. Tariq Elyas & Noor Motlaq Alghofaili Native English speakers versus non-native English speakers: The impact of language teachers on EFL learner’s English proficiency 34 Table 7. Paired samples correlations for students’ achievements in the listening skill Listening Group N Correlation Sig. NEST Pair 1 pre-test & post-test 14 -.016 .957 NNEST Pair 1 pre-test & post-test 14 -.286 .322 For further investigation, the independent samples t-test was applied (Table 8) to check whether there is a significant difference between the pre-tests and the post-tests of two groups. The test reveals that there is a significant difference between the pre-test means of both groups ((6.14 and 3.00 for the NEST and NNEST group, respectively), p=0.001). Moreover, there is no significant difference between the post-test means of both groups ((4.71 and 5.71 for the NEST and NNEST group, respectively), p=0.334). Table 8. Independent samples test for the listening skill Levene's Test for Equality of Variances t-test for Equality of Means F Sig. T df Sig. (2- tailed) Mean Difference Std. Error Difference 95% Confidence Interval of the Difference Lower Upper pre- test Equal variances assumed .905 .350 3.613 26 .001 3.14286 .86987 1.35482 4.93090 Equal variances not assumed 3.613 25.745 .001 3.14286 .86987 1.35395 4.93176 post- test Equal variances assumed 1.218 .280 -.985 26 .334 -1.00000 1.01558 -3.08755 1.08755 Equal variances not assumed -.985 25.352 .334 -1.00000 1.01558 -3.09015 1.09015 According to the analysis above, student achievement in the NEST group did not show any improvement in the listening part so that the third null hypotheses is accepted. On the other hand, studentd’ achievement in the NNEST group showed significant development in listening skill. Therefore, the fourth null hypothesis is rejected and NNEST appears to have a positive effect on students’ achievement in listening skill. Speaking skill Table 9 shows descriptive statistics for the NEST group and the NNEST group in speaking skill. In addition, Figure 3 shows the histograms for Module 2 and Module 3 grade distributions of each group. Table 9. Descriptive statistic for students’ achievement in the speaking skill Speaking Group Module 2 Module 3 NEST N Valid 15 15 Missing 0 0 Mean 9.1067 8.6133 Std. Error of Mean .30402 .36172 Std. Deviation 1.17745 1.40095 Variance 1.386 1.963 Range 3.10 4.00 Minimum 6.90 6.00 Maximum 10.00 10.00 NNEST N Valid 16 16 Missing 0 0 Mean 8.8250 9.1563 Std. Error of Mean .31563 .33726 Std. Deviation 1.26254 1.34906 Variance 1.594 1.820 Range 3.50 4.50 Minimum 6.50 5.50 Maximum 10.00 10.00 ENGLISH REVIEW: Journal of English Education Volume 7, Issue 2, June 2019 p-ISSN 2301-7554, e-ISSN 2541-3643 https://journal.uniku.ac.id/index.php/ERJEE 35 Figure 3. Grade distributions of NEST and NNEST groups in the speaking pre/post tests The above curves (second and third) display the normal distribution, but the Shapiro-Wilk test shows that data are not normally distributed for each group (p<0.05) (Table 10). Table 10. Test of normality for the speaking skill Speaking Group Kolmogorov-Smirnova Shapiro-Wilk Statistic Df Sig. Statistic df Sig. NEST Module 2 .297 15 .001 .751 15 .001 Module 3 .196 15 .127 .871 15 .035 NNEST Module 2 .305 16 .000 .780 16 .001 Module 3 .297 16 .001 .703 16 .000 a. Lilliefors Significance Correction The study null hypotheses are:  H5: NEST has no impact on student achievement in speaking skill.  H6: NNEST has no impact on student achievement in speaking skill. To test these hypotheses, a dependent t-test paired sample was applied to determine whether there is statistical evidence that the mean difference between paired observations for Module 2 and Module 3 grades for each group is significantly different from zero. Table 11, Table 12 and Figure 4 show that for the NEST group, the difference between Module 2 and Module 3 grades (0.493) is not significantly different from zero. That is, both grades are approximately equal (Module 2: 9.11±1.18, Module 3: 8.61±1.4). Therefore, the difference between the mean of the two grades is not large enough to reach the significant level (t(14)=2.145, P=0.05). However, the Module 3 mean is slightly lower. The fifth null hypothesis is accepted. As, for the NNEST group, the difference between Module 2 and Module 3 grades (-0.331) is not significantly different from zero. That is, both grades are approximately equal (Module 2: (8.83±1.26), Module 3: (9.16±1.35)). Therefore, the difference between the mean of the two grades is not large enough to reach the significant level (t(15)=-0.975, P=0.345). However, the Module 3 mean was slightly lower so that the sixth null hypothesis is accepted. Tariq Elyas & Noor Motlaq Alghofaili Native English speakers versus non-native English speakers: The impact of language teachers on EFL learner’s English proficiency 36 Table 11. Paired sample statistics for students’ achievement speaking skill Speaking Group Mean N Std. Deviation Std. Error Mean NEST Pair 1 Module 2 9.1067 15 1.17745 .30402 Module 3 8.6133 15 1.40095 .36172 NNEST Pair 1 Module 2 8.8250 16 1.26254 .31563 Module 3 9.1563 16 1.34906 .33726 Table 12. Paired sample test for students’ achievement the speaking skill Speaking Group Paired Differences t Df Sig. (2- tailed) Mean Std. Deviation Std. Error Mean 95% Confidence Interval of the Difference Lower Upper NEST Pair 1 Module 2 - Module 3 .49333 .89080 .23000 .00002 .98664 2.145 14 .050 NNEST Pair 1 Module 2 - Module 3 -.33125 1.35878 .33970 -1.05529 .39279 -.975 15 .345 Figure 4. Paired sample test for students’ achievement in the speaking pre/post tests Pearson’s correlation (Table 13) was used to measure the strength and direction of the association between the two variables measured (Module 2 and Module 3 grades). The following tables show that r=0.775 and p=0.001 for the NEST group and that r=0.46 and p=0.073 for the NNEST group, indicating that the NEST group displays a strong positive correlation, whereas the NNEST group displays an insignificant correlation. Table 13. Paired samples correlations for students’ achievements in the speaking skill Speaking Group N Correlation Sig. NEST Pair 1 Module 2 & Module 3 15 .775 .001 NNEST Pair 1 Module 2 & Module 3 16 .460 .073 For further investigation, the independent samples t-test was applied (Table 14) to check whether there was a significant difference between the Module 2 grades of both groups and to test whether there was a significant difference between the Module 3 grades of both groups. The test revealed that there was no significant difference between the Module 2 grade means of both groups ((9.1 and 8.61 for the NEST and NNEST group, respectively), p=0.525). In addition, there was no significant difference between the Module 3 grade means of both groups (8.83 and 9.16 for the NEST and NNEST group, respectively), p=0.281). ENGLISH REVIEW: Journal of English Education Volume 7, Issue 2, June 2019 p-ISSN 2301-7554, e-ISSN 2541-3643 https://journal.uniku.ac.id/index.php/ERJEE 37 Table 14. Independent samples test for the speaking skill Levene's Test for Equality of Variances t-test for Equality of Means F Sig. T df Sig. (2- tailed) Mean Difference Std. Error Difference 95% Confidence Interval of the Difference Lower Upper Module 2 Equal variances assumed .007 .933 .641 29 .526 .28167 .43926 -.61671 1.18004 Equal variances not assumed .643 29.000 .525 .28167 .43824 -.61463 1.17796 Module 3 Equal variances assumed .424 .520 -1.099 29 .281 -.54292 .49394 -1.55314 .46730 Equal variances not assumed -1.098 28.687 .281 -.54292 .49456 -1.55489 .46906 In conclusion, for both groups, Module 2 and Module 3 grades means were not significantly different. Neither group had a significant increase or decrease in grades. Thus, neither the NEST nor the NNEST had any effect on student achievement in the speaking skill. Therefore, the fifth and sixth null hypotheses are accepted. CONCLUSION This study concludes that teachers’ nativeness and backgrounds have no significant impact on the Saudi EFL learners’ achievement in the four skills. Overall performances in the pretest and posttest results show that NESTs and NNESTs can be equally effective as language teachers. These findings should be a convincing factor in shaping the hiring policy in the Arabian Gulf and the Native Speaker Fallacy (Phillipson, 1992) should not have any impact on EFL/ESL students, administrators, recruiters, and non- native teachers’ self-recognitions. This study shows no significant influence of NEST or NNEST on the EFL learners’ achievement in reading, listening, speaking and writing skills. The findings indicate that many students prefer NEST in teaching the reading skills, which is contrary to Al-Omrani (2008) and Mahboob (2004) who state that NNESTs are considered more effective in teaching reading skills. The data do not shed light on the learners’ preference for NEST as a reading teacher; however, NNESTs are favored for using reading strategies and techniques. This is in line with the results of Al-Omrani (2008), who reveal that NNESTs can teach students reading strategies that are required to overcome their reading challenges. Similar to Lasagabaster and Sierra (2005), few participants believe that teachers’ nativeness has no influence on their achievements in the reading skill. This point was also proven in the quantitative data indicating that being taught by either a NEST or NNEST has no effect on students’ achievements in the reading skill. Besides, the findings also show that there is no difference between NESTs and NNESTs in teaching listening. This is in contrast with the literature, which indicates that NESTs are better at teaching listening as they are the ideal models for authentic life (e.g. Al-Omrani, 2008; Kelch & Santana-Williamson, 2002; Lasagabaster & Sierra, 2005; Mahboob, 2004; Nam, 2010). However, results indicate that NESTs have no influence on students’ achievements in listening skill, whereas NNESTs have a remarkable positive influence on students’ achievements in listening skills. Further, the quantitative data of this study show that there is no significant difference among students’ achievements in speaking skills whether taught by NESTs or NNESTs. The teachers’ backgrounds and nativeness have no significant influences on students’ achievements, based on the results of this experimental study. This finding agrees with Al-Nawrasy (2013) and Levis et al., 2016), whose experimental studies yielded the same findings. Finally, it can be inferred from the findings that teachers’ nativeness has no significant effect on students’ achievements in speaking and listening skills. As part of the quantitative data of this study, the EFL students’ results on the writing skills show a negative influence of a NNEST on their achievements, whereas no influence has been noted in the learners’ progress who were taught by a NEST. This finding contradicts literature that NNESTs are favored by students in teaching writing skills as they learn better from their Tariq Elyas & Noor Motlaq Alghofaili Native English speakers versus non-native English speakers: The impact of language teachers on EFL learner’s English proficiency 38 teaching strategies (Al-Omrani, 2008; Alseweed & Daif-Allah, 2012; Mahboob, 2004). REFERENCES Alghofaili, N. M., & Elyas, T. (2017). Decoding the myths of the native and non-native English speakers teachers (NESTs & NNESTs) on Saudi EFL tertiary students. English Language Teaching Journal, 10(6), 1-19. Al-Nawrasy, O. (2013). The effect of native and non- native English language teachers on secondary students’ achievement in speaking skills. Jordan Journal of Educational Sciences, 9(2), 243-254. Al-Omrani, A. H. (2008). Perceptions and attitudes of Saudi ESL and EFL students toward native and nonnative English-speaking teachers (Doctoral dissertation). Indiana University of Pennsylvania, USA. Retrieved from ProQuest Dissertations and Theses database. (UMI No. 3303340) Alseweed, M. A. (2012). University students’ perceptions of the influence of native and non- native teachers. English Language Teaching, 5(12), 42-53. doi:10.5539/elt.v5n12p42. Alseweed, M. A., & Daif-Allah, A. S. (2012). University students’ perceptions of the teaching effectiveness of native and nonnative teachers of English in the Saudi context. Language in India, 12(7), 35-60. Anya, U., Avineri, N., Carris, L. M., & Valencia, V. (2010). Languages, identities, and accents: Perspectives from the 2010 Linguistic Diversity Conference. Issues in Applied Linguistics, 18(2), 157-169. Braine, G. (2013). Non-native educators in English language teaching. London: Rutledge. Chun, S. Y. (2014). EFL learners’ beliefs about native and non-native English- speaking teachers: Perceived strengths, weaknesses, and preferences. Journal of Multilingual and Multicultural Development, 35(6), 563–579. doi: 10.1080/01434632.2014.889141. Creswell, J. W., & Clark, V. L. P. (2017). Designing and conducting mixed methods research. Thousand Oaks: Sage Publications. Cohen, L., Manion, L., & Morrison, K. (2007). Research methods in education (6th ed.). London: Rutledge. Grubbs, S. J., Jantarach, V., & Kettem, S. (2010). Studying English with Thai and native‐speaking teachers. Teachers and Teaching: Theory and Practice, 16(5), 559-576. doi: 10.1080/13540602.2010.507966. Kelch, K., & Santana-Williamson, I. E. (2002). ESL students’ attitudes toward native- and nonnative- speaking instructors’ accents. The CATESOL Journal, 14(1), 57-72. Larsen-Freeman, D., & Anderson, M. (2011). Techniques and principles in language teaching. Oxford: Oxford University Press. Lasagabaster, D. & Sierra, J. M. (2005). What do students think about the pros and cons of having a native speaker teacher? In E. Llurda (Ed.), Non- native language teachers: Perceptions, challenges and contributions to the profession (pp. 217-241). New York: Springer. Levis, J. M., Sonsaat, S., Link, S., & Barriuso, T. A. (2016). Native and nonnative teachers of L2 pronunciation: Effects on learner performance. TESOL Quarterly. doi: 10.1002/tesq.272. Li-Yi, W. (2011). Taiwanese pre-service English teachers' attitude towards native-and-non-native- English-speaking-teacher team teaching. English Teaching & Learning, 35(2), 1-46. Mackey, A., & Gass, S. M. (2005). Second language research: Methodology and design. Mahwah, NJ: Lawrence Erlbaum. Mahboob, A. (2004). Native or nonnative: What do students enrolled in an intensive English program think? In L. D. Kamhi-Stein (Ed.), Learning and teaching from experience (pp. 121-148). Ann Arbor, MI: University of Michigan Press. Mahboob, A., & Elyas, T. (2014). English in the Kingdom of Saudi Arabia. World Englishes, 33(1),128-142. doi:10.1111/weng.12073. Moussu, L. (2010). Native and non-native English- speaking as a second language teachers: Student attitudes, teacher self- perceptions, and intensive English program administrator beliefs and practices (Unpublished doctoral dissertation). Purdue University, West Lafayette. Nam, H. H. (2010). The pedagogy and its effectiveness among native and non-native English speaking teachers in the Korean EFL context (Unpublished doctoral dissertation). State University of New York at Buffalo, Buffalo, NY. Phillipson, R. (1992). Linguistic imperialism. Oxford: Oxford University Press. Selvi, A. F. (2010). All teachers are equal, but some teachers are more equal than others: Trend analysis of job advertisements in English language teaching. WATESOL NNEST Caucus Annual Review, 1, 156-181. Sung, C. C. M. (2010). Native or non-native? Exploring Hong Kong students’ perspectives. The Lancaster University Postgraduate Conference in Linguistics & Language Teaching, 4, 243. Walkinshaw, I., & Oanh, D.H. (2012). Native or non- native English speaking teachers in Vietnam: Weighing the benefits. TESL-EJ, 16(3). Walkinshaw, I., & Oanh, D. H. (2014). Native and non-native English language teachers: Student perceptions in Vietnam and Japan. SAGEOpen,4(2). doi: 10.1177/2158244014534451. https://doi.org/10.1111/weng.12073 http://sdl.summon.serialssolutions.com/search?s.dym=false&s.q=Author%3A%22Walkinshaw%2C+I%22 http://sdl.summon.serialssolutions.com/search?s.dym=false&s.q=Author%3A%22Oanh%2C+D.+H%22