Exploring the Effectiveness of Native and Non-Native English Teachers on EFL Learners ’ Accuracy, Fluency, and Complexity in Speaking

. This study aimed to investigate the impact of teacher type (native vs. non-native) on the oral performance of male intermediate English as a foreign language learners between the ages of 14 and 20 who possessed similar language pro ﬁ ciency levels. The participants were selected based on the results of a written test and an oral interview and were divided into two classes: one taught by a native speaker and the other by a non-native speaker. Following a -month treatment period, each learner was interviewed and their speech was recorded. The researchers analyzed the data for accuracy, ﬂ uency, and lexical complexity. The statistical analyses revealed that learners taught by the native-speaking teacher demonstrated greater ﬂ uency and lexical complexity in their oral production, while those taught by the non-native teacher showed greater accuracy in their spoken language production. These ﬁ ndings suggest that the teacher type has different effects on language learning outcomes. The implications of these results extend to policymakers, administrators, and those involved in the employment and training of language teachers in Iran.


Introduction
The teacher's role has always received considerable attention in the language teaching field.This is due to the paramount role a teacher plays in enabling learners to communicate by providing them with "teacher talk" [1].Research has shown that comprehensible input provided by teachers [2] and the techniques used by them are among the main sources of language produced by learners [1,[3][4][5].As a result, language institutes have made it a primary concern to employ qualified teachers who can efficiently equip learners with speaking ability in a relatively short amount of time.
In English as a foreign language (EFL) contexts, learners are primarily exposed to the English language in class due to limited opportunities for exposure in society.Consequently, having native teachers is often preferred [6,7], as they can provide learners with exposure to standard English [8].
As an EFL setting, native and non-native English instructors are engaged in teaching in language institutes in Iran, each providing students with language knowledge in different aspects, resulting in different outcomes in the learners' language learning [9][10][11].Native teachers possess a deep understanding of the target language, including discourse, pragmatics, and pronunciation [12][13][14][15], and are thus expected to be more successful in teaching learners to communicate fluently.Conversely, some researchers [7,11,[16][17][18] argue that non-native teachers may communicate better with learners and help them overcome problems related to negative transfer from their mother tongue.
Surprisingly, there is some research in the literature that shows that non-native teachers might even outperform their native counterparts in teaching learners to speak the target language [13,[19][20][21][22]. Therefore, this controversy remains a matter of concern and needs to be studied more.Moreover, to the researchers' best knowledge, little if any, has been conducted in Iran to distinguish the effectiveness of native and non-native teachers on learners' speaking skill.Therefore, research is needed before we can decide which type of teacher can influence learners' speaking ability in Iran.This is a significant issue, as native teachers in Iran are often preferred over non-native teachers, regardless of their training or experience, even when they do not possess high academic degrees in the English language teaching field.According to Thornbury [23], this preference is questionable, as an English teacher should possess professional competence, which is usually gained through education and training followed by experience.In this study, the accuracy, fluency, and lexical complexity of spoken language produced by two groups of EFL learners taught by either a native or a non-native teacher are compared and contrasted.

Research Questions
The following questions formed the inquiry domain of the current study: (1) Is there any difference between the speaking abilities development of learners taught by a native teacher and that of those taught by a non-native teacher?(2) In what ways can each type of teacher (native or nonnative) enhance EFL learners' speaking skills concerning accuracy, fluency, and lexical complexity?

Literature Review
3.1.Native or Non-Native Teachers?Thomas [24] points out that "students do not come so far to be taught by someone who does not speak English" (p. 6).The belief that nativespeaking teachers contribute more to good language learning prevails in many EFL countries to the point that without a careful examination, parents and school administrators perceive them as more qualified and efficient language teachers than their non-native counterparts.While some scholars have considered native-speaking teachers pedagogically superior to non-native-speaking teachers and favored them for their cultural knowledge and oral abilities [12,18,25], others have given priority to non-native teachers over their native-speaker colleagues in EFL contexts [7,13,22,26].In this regard, some have criticized native teachers for their poor grammar (e.g., [27]) and their pedagogical knowledge.In the same token, Phillipson [20] does not believe in a native speaker for his proficiency in English as an ideal teacher and to serve as a model for the students.Similarly, Widdowson [22] strongly opposes the common belief that a native speaker always teaches better than a teacher who himself learnt English in an EFL context.Ellis [28] also advocates for non-native English teachers, citing their ability to provide effective structural grammar lessons and empathy for student learning difficulties.According to Schenck [29,30], Seidlhofer [7], and Tarnopolsky [26], nonnative teachers can leverage their shared language with their students to facilitate and accelerate the English learning process by utilizing the students' mother tongue when necessary.
Critics of non-native teachers, however, argue that they possess poor oral skills and cultural knowledge [26].Tarnopolsky suggests that non-native teachers have a foreign accent, which limits their ability to teach learners native-like pronunciation, and that their lack of cultural awareness hinders their ability to train learners to become effective communicators in Englishspeaking countries.
3.2.The Three Components of Speaking.Skehan [31,32] suggests that promoting language produced by learners is a primary goal in the world of EFL learning and teaching.According to Skehan [33], accuracy, fluency, and complexity are the aspects of produced language that are distinguished by what they measure in the linguistic output.As the purpose of this study is to focus on these notions in the learners' produced language, a brief definition of them provided in the literature is required.
3.3.Accuracy.Accuracy refers to error-free speech that conforms to the rule systems of the language being learned [31,32,34].Ellis [35] suggests that produced language is accurate when it conforms to the target language norms.Skehan [33] further suggests that a learner's produced language can be defined as accurate as long as it conforms to the interlanguage level in which they are operating.
Skehan [33] argues that learners' speaking fluency, accuracy, and complexity are adversely affected by their limited attentional resources, particularly when task demands are high.Demanding tasks require more attentional resources for the sake of meaningful communication, resulting in less attention being available to secure accuracy.
Ellis and Barkhuizen [36] present two types of accuracy measures: specific and general.Specific measures of accuracy focus on certain forms, such as native-like plurals and verbal morphology.In contrast, general accuracy measures take into account the overall accuracy of speech and provide a more realistic measure of learners' speaking accuracy [37].Measures of accuracy include correct T-units percentage [38], error percentage [39], and the percentage of correct vocabulary use [40].
3.4.Fluency.The development of fluency is a critical issue in language pedagogy.According to Chambers [41], there is no indisputable agreement as to the definition of fluency.Johnson and Johnson [42] suggest that it is concerned with functional appropriateness and the effortless flow of the second language (L2).In other words, fluency is achieved by communicating language in an appropriate timeframe and without unreasonable pauses or hesitations [31,32,36].Ellis [43] suggests that fluency can be facilitated by acquiring automatic and analyzed L2 knowledge, which requires learners to focus on forms as well as meanings.
Lennon [44] proposes that fluency can be measured based on the rate of delivery and the number of hesitation markers, such as repetitions and false starts.In other words, fluency is characterized by the speed with which language is produced and the number of disfluencies made.

Complexity. The concept of complexity has been referred
to as having a multifaceted nature in some studies on the acquisition and use of a foreign language.Skehan [31,32] defines it as "the elaboration or ambition of the language that is produced" (p.22) and as "the capacity to use more advanced language" (p.45).

Education Research International
Housen and Kuiken [34] interpret complexity in terms of the cognitive and linguistic properties of language subsystems.Cognitive complexity refers to the level of difficulty involved in processing linguistic features while performing in L2, while linguistic complexity pertains to the objective factors that make a task challenging in nature [31,32].
According to Ellis [35], complexity can relate to various aspects of language and language use.In this regard, lexical diversity has been used in many educational and linguistic studies.Read [45] defines lexical diversity as the range of active vocabulary correctly used by speakers or writers.In this article, we focus on the learners' lexical complexity in speaking.

Method
4.1.Participants.The data for this study were collected from an established language institute in Iran.The participants comprised 30 male learners with the same language proficiency background, ranging in age from 14 to 20 years (M = 17.63,SD = 2.18).These learners had already attended introductory, elementary, and pre-intermediate levels in the institute, all of which were taught by non-native teachers.It is worth noting that the institute has designed a classification of five levels based on the New American Headway series for learners, which includes introductory, elementary, pre-intermediate, intermediate, and upper intermediate levels.
According to the institute's standards, learners who score 80 or above out of 100 on the multiple-choice tests and oral interview exam, which are both based on the textbook series mentioned earlier, are considered to have achieved the required language proficiency and are permitted to move on to the next level.The multiple-choice tests include listening comprehension, grammar, vocabulary, and writing questions.
To begin the experiment, the students were randomly divided into two groups: one taught by a native English teacher and the other by a non-native English teacher.The native teacher was born and raised in the United States, but he left the country with his family at the age of 18 and spent a 2-year period of military service in Iran before starting to teach English at the institute at the age of 20.At the time of data collection, he had 1 year of teaching experience.He was fluent in speaking American English and exposed the students to his intelligible input using simple words quite talkatively.However, he did not have explicit knowledge of the language grammar, and he lacked fluency in speaking the students' mother tongue (Persian).
In contrast, the non-native teacher was an Iranian master's graduate majoring in Teaching English as a Foreign Language (TEFL) and had 2 years of experience in teaching English.He tended to use academic vocabulary while speaking English and could easily provide definitions of any vocabulary.Additionally, he had explicit knowledge of English grammar.

4.2.
Instruments.This research employed a predeveloped list of 10 open-ended questions (the appendix) devised by the institute and based on the textbook taught to the learners.The questions' validity and reliability had already been confirmed by the institute to ensure that the learners were qualified to attend the intermediate level.Additionally, a pilot study of the questions was conducted in another English institute on a group of learners with a similar level of language proficiency to establish the questions' reliability.Intra-and interrater reliability procedures were used, resulting in reliability coefficients of 84 and 79.8, respectively, calculated using the Pearson Product formula.The semistructured interview was conducted to encourage natural communication and elicit the learners' implicit knowledge.4.3.Measure of Analysis.In order to analyze the oral data, AS-unit, suggested by Foster et al. [46], was employed in this study.They referred to AS-unit as "a single speaker's utterance consisting of an independent clause, or subclausal unit, together with any subordinate clause(s) associated with either" (p.365).Several previous studies have reported the suitability of the As-unit for analyzing oral data, including works by Norris and Ortega [47], Plough et al. [48], and Foster et al. [46].
To arrive at the accuracy scores, the number of error-free AS-units produced by each participant in response to the questions was divided by the total number of AS-units generated by the participant, following the method used by Foster et al. [46].Additionally, following Ellis and Barkhuizen [36], self-corrected units were considered error-free.
For the fluency measure, the dysfluency principle suggested by Lambert and Engler [49] was employed.Dysfluency was defined as the pauses in which the learner is searching for the language due to a lack of automaticity or deficiency in the target language knowledge.Thus, the number of AS-units produced by each participant in the interview was divided by the number of dysfluencies produced by the participant.The ratios obtained were indicative of their fluency.
Finally, to score the lexical density of the students' oral performance, the number of lexical words produced by each participant was counted and divided by the total number of words they used in their speech, following the method used by Robinson [50].The obtained statistic was then multiplied by 100 to obtain each student's score.

Data Collection.
To ensure the homogeneity of the learners, the audio-taped interviews of the learners following the pre-intermediate level were transcribed and coded for analysis as a pretest using SPSS.As the data were found to be normal, an independent samples t-test was used to confirm that there was no statistically significant difference in the proficiency level concerning accuracy, fluency, and lexical complexity between the two groups before treatment.
After the intermediate course was completed, the researchers conducted posttests by conducting interviews with the learners (see the appendix) to determine any variations in their fluency, accuracy, and lexical complexity.The interviews lasted ∼15 min, and the learners' voices were tape-recorded.The tape recordings were transcribed independently by several transcribers.The researcher and a colleague analyzed and coded the transcriptions separately, and interrater reliability was computed.Additionally, to increase the accuracy and reliability of Education Research International the analysis, the researcher listened to them and scored them again with a time interval of about 1 month, and intrarater reliability was calculated using the Pearson Product Formula.The slight differences did not affect the results, and conformity was achieved for both intra-and interrater reliability (0.87 and 0.83, respectively).

The Results of Pretests on Students' Speaking Proficiency.
In order to confirm the learners' speaking skill homogeneity concerning the accuracy, fluency, and lexical complexity, interview scores for the pre-intermediate level were calculated using SPSS.The results are presented in Table 1.
As illustrated in Table 1, the independent sample t-test revealed that the learners in both classes were not significantly different from each other regarding accuracy (p ¼ 0:926), fluency (p ¼ 0:323), and lexical complexity (p ¼ 0:507).Thus, their speaking skill homogeneity was established, which paved the way for subsequent comparison of their progress in speaking proficiency.

The Results of Posttests on Students' Speaking Proficiency.
Following the treatment, the speaking proficiency scores of the learners in both groups concerning accuracy, fluency, and lexical complexity were analyzed through SPSS.The results are illustrated in Table 2.
Table 2 shows that students taught by both teacher types made significant progress in accuracy and fluency while speaking.More specifically, however, the results show that students taught by the non-native teacher outperformed (M = 90.93,SD = 4.04) the students taught by the native teacher in terms of accuracy (M = 80.80, SD = 5.77, p ¼ 0:001).This is due to the fact that the non-native teacher would try to conform strictly to the rules of English grammar while speaking and would also provide students with feedback in this regard.
In contrast, regarding fluency, students instructed by the native teacher outperformed (M = 86.53,SD = 5.47) learners instructed by the non-native teacher (M = 73.33,SD = 7.22, p ¼ 0:001).This can be explained by the fact that the native teacher provided the learners with plenty of easily intelligible input (e.g., making a lot of associations and paraphrases) while raising topics for speaking practice in class, which would, in turn, motivate the learners to take part in the discussions.In other words, enjoying the class, the learners would get the feel of being in an English-speaking context and would try to get engaged in speaking and produced more output by putting forward their ideas.
As for lexical complexity, however, no meaningfully significant difference appeared between the two groups (p>0:547).
To provide an even more comprehensive perspective of speaking performance, each group was analyzed based on the difference between pretest and posttest scores utilizing paired sample t-tests.
As the results shown in Table 3, the learners taught by the native teacher improved significantly regarding fluency (p ¼ 0:001) and lexical complexity (p ¼ 0:006); however, they did not improve meaningfully regarding accuracy (p ¼ 0:341).It can be due to the fact that the native teacher teaching the students had simply picked up the language himself in the United States and had not been instructed the English language course academically.Therefore, he must have been easygoing with the errors students made while speaking and provided them with no feedback, as he mentioned this once in an informal meeting of teachers.
The learners taught by the non-native teacher, on the contrary, made meaningfully significant improvement concerning speaking accuracy (p ¼ 0:002) but did not make significant progress in lexical complexity (p ¼ 0:257).They  4).This result's interpretation can be that the non-native teacher majorly directed the learners' attention to the matters of accuracy while speaking and did not try to elevate their level of fluency by motivating them to take part in meaningful speaking activities, which, in turn, kept them from developing their fluency and practical lexical repertoire.

Discussion
This study attempted to compare the effect of teacher type (native vs. non-native) on students' speaking skills concerning accuracy, fluency, and lexical complexity.The findings in this study provide support for previous investigations, which suggest that both native and non-native teachers have specific qualities affecting the efficacy of English instruction [10,11].
The first research question attempted to demonstrate whether teacher type (native and non-native) would lead to any meaningful difference in EFL students' speaking abilities.The findings in this investigation demonstrated that the students of both groups made significant progress in their speaking abilities.The posttest results revealed that students in both groups made statistically significant progress regarding two variables, namely accuracy (p ¼ 0:001) and fluency (p ¼ 0:001) but not in lexical complexity (p ¼ 0:547).This can be claimed as a support for Skehan's [33] belief that, due to limited attentional resources, humans do not have the capacity to focus on all aspects of language (accuracy, fluency, and lexical complexity) simultaneously.Thus, learners have to meet a trade-off between different aspects; for instance, if a learner attends to accuracy when speaking, they have to compromise fluency and vice-versa.
The second research question aimed to examine which teacher type helps the learners to speak the language more accurately and effortlessly and use a more diverse lexicon.The results of paired sample t-tests revealed that the learners taught by the native teacher could improve their speaking skills concerning fluency (p<0:001) and lexical complexity (p<0:001).However, they did not significantly improve their speaking accuracy (p ¼ 0:341).In contrast, the learners taught by the non-native teacher only improved their speaking abilities concerning accuracy (p ¼ 0:002).They could not improve their lexical complexity (p ¼ 0:257) and even deteriorated their fluency as shown in the posttest results (p ¼ 0:012).These findings align with Schenck's [29,30] studies demonstrating that non-native-speaking teachers can help learners achieve higher speaking accuracy, while native-speaking teachers can help learners achieve higher speaking fluency and lexical complexity.However, the interpretation of these findings should not simply suggest that non-native-speaking teachers require additional training.Instead, the results suggest that both native and non-native-speaking teachers may require specific types of training to improve their instruction.For example, non-native-speaking teachers may require additional training in providing better input to promote lexical complexity and fluency, while nativespeaking teachers may require skills training in English grammar or training that highlights the differences between the learners' L1 and L2 to promote accuracy.Therefore, it is important to consider the training needs of both teacher types to provide more effective language instruction.The findings of this study contribute to the ongoing discussion on the effectiveness of native and non-native-speaking teachers in English language instruction and highlight the need for targeted training programs to improve language instruction for both teacher types [8,25].
The students taught by the native teacher would attempt to express their ideas freely and without stress.This is as a Education Research International result of motivating the students to speak by raising interesting topics and ideas in the class without demanding a focus on grammar.That is, the teacher would not stop students for error correction while providing them with the required lexicon.It is worthwhile to mention that the native teacher, in this study, was not proficient in explaining of some grammatical points due to lack of receiving training and experience in teaching; for instance, he felt quite incompetent in explaining the conditional forms or clarifying present perfect continuous and future perfect tenses.The native teacher, therefore, seems to have used a less prescribed instruction, by mostly exposing the learners to his input, which is in accordance with Modiano [51] mentioning that "the very goals of the instruction are based on learners' preferences and needs as opposed to the mastering of a form and structure defined by those who believe that a prescriptive standard, by default, is a superior form of language" (p.172).
The students taught by the non-native teacher in this study would, on the other hand, focus their attention on producing grammatically correct spoken output, which prevented them from speaking freely and spontaneously about the ideas which came to mind.This is due to the fact that the non-native teacher would interrupt and provide learners with immediate feedback on articles, pluralization, tenses, verb forms, etc.The finding that the learners taught by the non-native teacher proved more accurate in speaking than those taught by the native teacher supports the findings by Medgyes [18], Arva and Medgyes [12], and Moussu [52] arguing that grammar was the main focus of non-native teachers, although in this study, the teaching context was not devoted to pure grammar teaching.
The non-native teacher, in this study, is highly likely to have grounded his teaching on the way he experienced learning English himself in the Iranian EFL context (for instance, through resorting to metalinguistic feedback, explicit correction, repetition, etc.).According to Yasuda [53], in EFL pedagogical milieus, focus on grammar is preferred to pragmatic aspects, which can be due to a lack of sufficient fluency and low self-confidence on the part of the non-native teacher, and hence fearing their status to be threatened [10,11,14].Thus, the non-native teacher in this study seems to have accentuated the grammatical aspects, favoring accuracy to fluency.
The results of this study support the beliefs of Cook [13] and Arva and Medgyes [12], arguing that native teachers have the privilege of knowing the target language in using its spoken form as well as the related cultural issues; this study, further, advocates Medgyes [18] and Seidlhofer [7] pointing out that non-native teachers have the advantage of knowing the structures of both L1 and L2 better.Hence, each type of teacher is likely to result in different consequences in learners' language learning.However, the findings in this study do not seem to be in line with the studies conducted by Maftoon et al. [54] and Fard et al. [55] since they reported that native teachers in Iran outperformed nonnative teachers in teaching all aspects of oral skills.The results, on the other hand, support the assumption that native and non-native teachers contribute to different aspects of student learning, each by bringing certain qualities to the practice [30].
Although the findings of this study are insightful, they should be treated with caution as the investigation only included one teacher for each teacher type, which detracts from its reliability.Research comprising a larger number of native and non-native teachers would provide more insight into any context.Additionally, the learners who participated in this study were of only one level, and the findings should not be generalized to other levels.Furthermore, this study used a small-scale dataset, and further research with a larger sample in other contexts and with different age groups is needed to investigate the issue.Moreover, this study considered the complexity of speech production only in terms of lexicon, and future research could extend this to structural and discoursal features.Finally, to replicate this study, it is suggested to use different instruments such as communicative information exchange, free discussion, opinion exchange, and picture description tasks.

Conclusion and Implications
This study aimed to examine the effect of teacher type (native and non-native) on EFL students' oral abilities.The results showed that each teacher type can bring specific abilities in teaching students.While the native teacher could motivate the students with his fluent speaking, the non-native teacher could help the learners become aware of their grammatical errors.
The results of this research have implications for policymakers, administrators, and those involved in the employment and training of language teachers in Iran.First, hiring a native speaker of English, irrespective of their teaching knowledge and experience, is not a viable option.A language teacher should not only be a good conversationalist but also have academic knowledge about language teaching, different methods and postmethods practiced in the past and present, and their pros and cons to adopt a principled eclectic method.Second, this study highlights the importance of organizing language teaching practices.Any deficits in the education of English teachers should be urgently addressed since it will impact learners and language schools, favoring highly qualified teachers over those who are just native speakers.Such a policy will have a significant impact on language pedagogy, as well as other social issues such as sociology and the economy.

TABLE 1 :
Independent sample t-tests for accuracy, fluency, and lexical complexity on the pretest.

TABLE 2 :
Independent sample t-tests for accuracy, fluency, and lexical complexity on the posttest.

TABLE 4 :
Results of paired-sample t-test for accuracy, fluency, and lexical complexity of the students taught by the non-native teacher.

TABLE 3 :
Results of paired-sample t-test for accuracy, fluency, and lexical complexity of the students taught by the native teacher.