Examining the perceptions and self-reported practices of L2 teachers in China regarding oral fluency: A conceptual replication and extension

Abstract

This study, a partial replication of Tavakoli and Hunter, examined the perceptions and self-reported practices of 72 second language (L2) teachers of English or Japanese in China regarding oral fluency. The research employed a mixed-methods approach, collecting both quantitative and qualitative data through questionnaires and interviews. The findings aligned with the original study conducted in the UK, suggesting that L2 teachers in China often viewed fluency as encompassing speaking ability in a broad sense. Therefore, these teachers reported a large majority of free production tasks that enhanced speaking ability, rather than specifically focusing on fluency development. Collectively, these findings highlight a discrepancy between the scholarly definition of fluency and teachers’ understanding of it, and underscore the importance of examining the impact of teacher perceptions on their practices for promoting fluency. Additionally, this study provided novel insights by exploring the influence of teacher variables on teachers’ confidence in their knowledge about fluency. The findings demonstrated that certified teachers, teachers of multiple L2s, more experienced teachers, and teachers in more socioeconomically advantaged regions exhibited greater confidence in their knowledge about fluency. Furthermore, the interaction effects among various teacher variables were found, underlining the complexity of teachers’ roles in perceiving fluency. Considering the findings from both studies, adopting a narrower definition of fluency and incorporating more fluency-focused activities in the classroom could enhance the effectiveness of fluency teaching. Moreover, this study emphasizes the importance of considering teachers’ variability in their understanding of fluency.

Keywords

classroom practices defining fluency L2 fluency teacher perceptions teacher variables

I Introduction

This replication study investigates the beliefs and self-reported practices of second language (L2) teachers in China regarding oral fluency. To replicate Tavakoli and Hunter’s (2018) study, this research first examines teachers’ definitions of oral fluency within the Chinese context. It then investigates teachers’ self-reported classroom practices aimed at promoting fluency in this context. In addition to replicating the original study, this research expands upon it by examining the influence of teacher variables on teachers’ confidence in their knowledge about fluency.

Speaking a language fluently is often the ultimate goal for language learners (Kormos & Dénes, 2004, p. 145). Consequently, a substantial body of research has been conducted to examine the definitions of oral fluency (Freed, 2000; Lennon, 1990; Tavakoli, 2016), its aspects (Koponen & Riggenbach, 2000; Préfontaine et al., 2016; Segalowitz, 2010), and the practices that promote it (Boers, 2014; Rossiter et al., 2010; Tavakoli et al., 2016; Tran & Saito, 2021). From a research perspective, fluency has been defined as the ‘flow, continuity, automaticity, or smoothness of speech’ (Koponen & Riggenbach, 2000, p. 6). Thus, in a narrow sense, fluency mainly refers to the ease and automaticity of speech. It is considered one isolatable component of speaking ability (Lennon, 1990) and, therefore, differs from other key characteristics of speaking ability, such as complexity and accuracy (Ellis, 2009; Foster, 2020; Housen et al., 2012; Skehan, 2009). In this narrow sense, fluent speech is ‘unimpeded by silent pauses and hesitations, filled pauses, self-corrections, repetitions, false starts, and the like’ (Lennon, 1990, p. 390). However, in a broad sense, fluency represents speakers’ speaking ability or general L2 proficiency (Lennon, 1990). Thus, fluency has often been defined in two senses, based on its functions: a narrow sense and a broad sense.

To distinguish the aspects of fluency, Segalowitz (2010) proposed a triadic model from a cognitive perspective. The first aspect is utterance fluency, which refers to ‘the fluidity of the observable speech as characterized by measurable temporal features, such as syllable rate, duration and rate of hesitations, filled and silent pauses, and . . . breakdown fluency and repair fluency’ (Segalowitz, 2016, p. 81). The second aspect is cognitive fluency, which refers to ‘the fluid operation (speed, efficiency) of the cognitive processes responsible for performing L2 speech acts’ (Segalowitz, 2016, p. 82). Unlike these two aspects that primarily reflect the objective characteristics of fluency, the third aspect is perceived fluency, which pertains to ‘subjective judgments of L2 speakers’ oral fluency’ (Segalowitz, 2016, p. 86). Some studies indicate that utterance fluency interacts with both cognitive and perceived fluency, suggesting an interplay between the different aspects of fluency in reflecting its complex and multifaceted nature (e.g. de Jong & Bosker, 2013; Préfontaine, 2013; Suzuki & Kormos, 2023).

Furthermore, many studies have found that the development of fluency can be facilitated through the use of extensive speaking practices and explicit instructions targeting specific aspects of fluency (e.g. Gatbonton & Segalowitz, 1988). The main activities promoting fluency include consciousness-raising tasks (Dudley, 2007; Riggenbach, 1999), pre-task planning (Tavakoli & Skehan, 2005; Yuan & Ellis, 2003), task repetition (Ahmadian & Tavakoli, 2011; Lynch & Maclean, 2000), the 4/3/2 technique (Boers, 2014; Thai & Boers, 2016; Tran & Saito, 2021), and instructions on formulaic chunks or lexical bundles (Wood, 2010; Wray, 2008).

These studies demonstrate that, despite fluency being a complex and multifaceted construct in second language acquisition (SLA), its definitions and aspects have been established from a research perspective. Based on research findings, many classroom practices for promoting fluency have been proposed, which have also been verified to be effective in various educational contexts. Nevertheless, despite the crucial role that teachers play as facilitators and providers of knowledge regarding fluency, only a limited number of studies have investigated whether teachers have embraced the scholarly definition of fluency and implemented the recommended fluency-focused teaching practices. The most relevant study in this regard was conducted by Tavakoli and Hunter (2018), which investigated how L2 teachers perceive and define fluency and the classroom practices they report to promote fluency in the UK. The results showed that the majority of teachers conceptualized fluency in a broad sense, considering it as a cover term for speaking ability and general L2 proficiency. Furthermore, most reported classroom practices for promoting fluency were free production activities, with only a small proportion focusing on fluency. These findings suggested an interaction between teachers’ definitions and their reported classroom practices regarding fluency, and a discrepancy between ‘what fluency research recommends and what teachers do in class’ (Tavakoli & Hunter, 2018, p. 330).

To extend the applicability of these findings, this study conceptually replicated Tavakoli and Hunter (2018) within the Chinese context. We adopted Porte and McManus’s (2019) label of ‘a conceptual replication’ because many aspects of the original study were matched as closely as possible, with some major variables being different (see Section III). The most notable modification in the design pertained to the change in research context. The decision to choose China as the replication context was motivated by the fact that previous studies investigating teachers’ beliefs and practices regarding fluency were predominantly conducted in highly mobile and developed countries (Dore, 2016; Morrison, 2018; Tavakoli, 2023; see Section II.1), such as the UK and Italy. In these countries, learners and teachers benefit from exposure to and contact with English and other foreign languages due to factors like international tourism, transnational business, multilingual culture, and multiethnic educational settings. In contrast, China relies primarily on different Chinese varieties (e.g. Mandarin, Cantonese) for education, business, and daily communication, resulting in limited exposure to foreign languages for L2 learners (e.g. Hu, 2003). Furthermore, classroom-based formal instruction serves as the primary method of L2 learning in China (Wu, 2001), underscoring the significant role of teachers and their classroom practices in facilitating learners’ language development.

Given these distinctions, we anticipate potential discrepancies in classroom teaching practices and outcomes arising from teachers’ varying understanding of fluency, as well as the impact of teacher variables on their understanding of fluency within the Chinese context. Consequently, this study selected a representative sample of the L2 teacher community in China, addressing a new research question exploring how teacher variables, such as teaching experience, training experience, and work regions, influence teachers’ understanding of fluency. To further enrich the research design, semi-structured interviews with teachers were incorporated as an additional data source, providing triangulation of the findings.

II Literature review

1 Teachers’ perceptions of oral fluency

Although SLA researchers have established definitions and applications of fluency, there remains a lack of clarity among language teachers. Particularly, teachers often interpret fluency in a broad sense, encompassing speaking ability or general L2 proficiency, which aligns with its colloquial usage, rather than its more specific and precise academic sense. However, research on teachers’ perceptions and definitions of fluency is still in its early stages. To our knowledge, only four studies have been conducted in this area.

Dore (2016) conducted a survey among native speaker university teachers of English in Italy and the UK, using online questionnaires to assess their perceptions of fluency. The samples differed in terms of gender, age, teacher qualifications, and years of teaching experience. However, both samples were comparable in terms of the number and level of academic qualifications. The results showed that, contrary to the scholarly definition of fluency, which emphasizes speed, breakdown, and repair, teachers attributed greater importance to coherence, chunking, intonation, and rhythm of speech. A comparison of the two samples suggested that participants in Italy had a broader understanding of fluency. Additionally, Dore (2016, p. 41) investigated the effects of three teacher variables – academic qualifications, teacher qualifications, and years of experience – on teachers’ perceptions of fluency. The correlation tests revealed a positive correlation between teacher qualifications and perceptions of automaticity in fluency, indicating that teachers who received more training placed greater importance on automaticity. Conversely, years of experience were negatively correlated with perceptions of speech rate, indicating that teachers with longer teaching experience were less concerned with rapid L2 speech. However, there was no significant correlation between academic qualifications and teachers’ perceptions of fluency.

Building on the research findings of Dore (2016), Tavakoli and Hunter (2018) developed a questionnaire to investigate the perceptions of L2 teachers of English and Modern Foreign Languages (e.g. Spanish) regarding fluency in the UK. Although over 80% of the teachers reported knowing the meaning of fluency and its definable features to a large or some extent in a rating task, their answers to the open-ended questions revealed inconsistencies. Only a small proportion of terms (13.4%) defined fluency in a narrow sense, focusing on flow, continuity, and automaticity, while the majority of definitions (76.3%) related to the aspects of speaking ability and general L2 proficiency, including pronunciation, intonation, vocabulary, and grammar. These findings indicated a mismatch between teachers’ confidence in their knowledge about fluency and their definitions of fluency. Similar patterns were found in the Chilean context, where Morrison (2018) examined the perceptions of high school L2 teachers of English regarding fluency. Using a modified Spanish version of the questionnaire developed by Tavakoli and Hunter (2018), Morrison surveyed 60 participants and found that only 10.6% of the definitions of fluency provided by Chilean L2 teachers related to fluency in a narrow sense, while over 56% focused on speaking ability and general L2 proficiency.

To enhance L2 teachers’ understanding of fluency, Tavakoli (2023) conducted a one-day professional workshop in the UK. Before and after the workshop, participants completed the questionnaire developed by Tavakoli and Hunter (2018). The results showed significant improvements in the teachers’ understanding of fluency in all questionnaire items, including definitions, aspects, and teaching strategies of fluency. These findings suggested that professional training or workshops on fluency had a positive impact on teachers’ clearer understanding of fluency and their self-reported teaching practices for promoting fluency.

It is worth noting that all these studies were conducted in highly mobile, multilingual, and developed countries. Furthermore, while Dore (2016) examined three teacher variables in the data collected from native speaker university teachers of English in Italy and the UK, there have been no follow-up studies to validate the generalizability of these findings to non-native speaker teachers and other educational contexts. Additionally, although in the original study there were several interesting teacher variables, their effects on teachers’ understanding of fluency were not examined. These factors serve as the primary motivations for conducting this replication study in the Chinese context (for further details, see Section III).

2 Teaching practices for promoting oral fluency

Fluency research has generated prolific findings that directly relate to and benefit learning and teaching practices aimed at promoting fluency in various educational contexts (e.g. de Jong & Perfetti, 2011; Freed et al., 2004; Hunter, 2017; Tavakoli et al., 2020). Among these practices, task planning and task repetition have received considerable attention. It has been found that manipulating preparation time and training learners in planning time use can facilitate the development of fluency (e.g. Tavakoli & Skehan, 2005; Mehnert, 1998). Additionally, repeating the same task multiple times or under different time constraints promotes the automatization of speech production, thereby enhancing fluency in speaking performance (e.g. Ahmadian & Tavakoli, 2011; Gass et al., 1999; Sample & Michel, 2014). One widely used task repetition technique is the 4/3/2 technique and its variations (Boers, 2014; Thai & Boers, 2016; Tran & Saito, 2021). In the 4/3/2 task, students are required to deliver a speech on a chosen topic for 4 minutes initially, then for 3 minutes, and finally for 2 minutes. Teaching formulaic sequences to learners is another important activity. These sequences, as frequent collocations in language use, entail lower processing demands and greater accessibility for learners, thereby being more readily producible in speech and promoting fluency (Chan, 2019; Wood, 2010; Wray, 2008). Classroom activities that raise awareness of the significance and advantages of becoming fluent L2 speakers – such as recording and listening to one’s own speech – are considered effective in promoting fluency (Dudley, 2007; Riggenbach, 1999). Such tasks are argued to enhance learners’ fluency in oral performance by increasing their awareness of fluent speech and motivating them to improve fluency (Seifoori & Vahidi, 2012; Tavakoli et al., 2016).

Though many classroom practices have been suggested and verified by fluency researchers, they are not well received by developers of education materials and language teachers (e.g. Ahmadian et al., 2017). Rossiter et al. (2010) surveyed fluency activities in English as a second language (ESL) learner and teacher resource materials in Canada and found that free production tasks – such as group discussions, role-plays, and debates – were the most common activities in all materials. These activities are not specifically designed to improve learners’ fluency but rather their speaking ability or communicative competence. However, less focus was placed on fluency-focused activities such as rehearsal, planning time, and repetition. These authors argued that free production tasks alone are unlikely to have a significant impact on fluency. These findings have been confirmed by teachers’ self-report of their classroom practices in the UK (Tavakoli & Hunter, 2018). The majority of reported practices (53.6%) were free production tasks, while only a relatively small proportion of the activities (10.4%) were fluency focused. Collectively, these studies suggest that there is a mismatch between recommendations from researchers and practices employed by teachers in promoting fluency. This contradictory relationship will be further investigated in this study with a sample of teachers in China.

III Motives for replication

Apart from the general need for more replication research, as ‘replication is both a much-neglected and much-needed feature of applied linguistics research’ (Porte & McManus, 2019, p. 176), there are several reasons why the design and findings of Tavakoli and Hunter (2018) invite a partial replication.

First, Tavakoli and Hunter (2018) is one of the first studies on L2 teachers’ understanding of fluency and their related practices, and thus it needs to be verified, consolidated, and its generalizability to other contexts needs to be examined to establish a solid foundation and strengthen the field’s evidence base for future studies on this topic.

Second, the authors of the original study are considered key researchers in the field of fluency research, and they have been open and transparent about the data and methods used, providing venues for replication studies.

Third, the questionnaire used in the original study was developed based on a UK sample, thus requiring validation and reliability testing in other educational contexts. This replication study was conducted in China, where there are notable differences in language environment, cultural diffusion, political status quo, and educational systems compared to the UK (e.g. Hu, 2003; Wu, 2001).

Fourth, the original study employed a convenience sampling approach in participant selection, which may limit the representativeness of their sample. In the replication study, a purposeful sampling approach was employed to enhance the representativeness of the sample within the L2 teacher community in China (Dörnyei, 2007).

Fifth, building on L2 teachers’ data, the original study expanded on the dichotomy of fluency proposed by Lennon (1990) to a four-level model. This model encompasses four approaches to defining fluency, namely ‘very narrow’, ‘narrow’, ‘broad’, and ‘very broad’. This significant theoretical contribution to the field of fluency warrants replication to validate the model and improve its generalizability and interpretability among teachers in the Chinese context.

Sixth, although there were important teacher variables in the original study, they were not examined. This replication study extended the original study by investigating the effects of teacher variables on teachers’ understanding of fluency. This research question was addressed by conducting linear regression modeling in R (R Core Team, 2023), which ensured methodological and analytical rigor and enhanced the reproducibility of statistical analysis in this open-access platform (Mizumoto & Plonsky, 2016).

Seventh, the original study acknowledged the limitation of not conducting interviews with the participants to triangulate the data source and deepen the interpretation of findings. In the replication study, follow-up interviews were conducted with a representative focal group, providing further insights into the mechanisms behind teachers’ perceptions and practices of fluency.

Motivated by these reasons, our first and second research questions followed Tavakoli and Hunter’s (2018) research question 1-a and research question 2. For our research question 1, we hypothesize that, similar to the original study, teachers in China would often define fluency in a broad sense, considering it as a cover term encompassing speaking ability and general L2 proficiency. For research question 2, we hypothesize that, due to teachers’ broad perspective on fluency, their reported practices would prioritize free production rather than fluency-focused activities.

Research question 1: How do L2 teachers in China define fluency and fluent L2 speakers?

Research question 2: What classroom practices do L2 teachers in China use to promote fluency?

Research question 3: What are the effects of teacher variables on L2 teachers’ confidence in their knowledge about fluency in China?

The third research question was novel in the replication study. We examined six teacher variables, including Teaching Experience, Teacher Certification, L2 Numbers, Institution Type, Region, and Course Type. Based on the findings in Dore (2016) and Tavakoli (2023) regarding the effects of Teaching Experience and Teacher Certification, we hypothesize that more experienced teachers and certified teachers would have greater confidence in their knowledge about fluency in China compared to less experienced teachers and uncertified teachers. The other four teacher variables have not been examined in studies on teachers’ perceptions of fluency, but they have been found to have an impact on L2 emotions and outcomes. Based on previous research (e.g. Borg, 2013; Dewaele et al., 2018; Khau & Huynh, 2022), we hypothesize that teachers with teaching experience of multiple L2s, university teachers, teachers working in more socioeconomically advantaged regions, teachers of speaking-focused courses would be more confident in their knowledge about fluency compared to teachers with teaching experience of only one L2, high school and training school teachers, teachers working in less socioeconomically advantaged regions, and teachers of general language courses.

IV Methodology

1 Participants

To sample a representative population of the L2 teacher community in China, we adopted a purposeful sampling approach (Dörnyei, 2007), considering various teacher characteristics during the participant selection process. Initially, we sent invitation letters to 80 L2 teachers of either English or Japanese through our social networks to participate in this study. However, 8 of them did not reply, resulting in a total of 72 participants and a response rate of 90%. All participants were native speakers of Chinese but non-native speakers of the foreign language(s) they were teaching.

In Tavakoli and Hunter (2018), they employed a convenience sampling approach, and their participants represented a range of different first languages and nationalities. Hence, the sampling approaches, as well as the language and national backgrounds of participants, differed between both studies. These intentional modifications aimed to recruit a representative sample for analysing the effects of teacher background on teachers’ understanding of fluency in China.

The detailed characteristics of the participants in both studies were summarized in Table 1. In this study, the majority of participants (94%) were ESL teachers, while the remaining 6% were teachers of Japanese as a second language (JSL). This distribution was not a sampling bias but rather a reflection of the composition of L2 teachers in China, where ESL teachers constituted the dominant group. Among the 72 teachers, 14% reported having taught a third language for a period of time (range = 1.5 ~ 8 years, M = 2.3, SD = 2.39), including Korean, French, and German. These teachers were categorized as teachers of multiple L2s. Furthermore, the participants worked in three types of institutions, with 43% being university teachers, 31% being high school teachers, and 26% being private language training school teachers.

Table 1.

Participant characteristics in both studies.

Categories	Current study (N = 72)			Original study (N = 84)
	Sub-categories	n	Percentage	Sub-categories	Percentage
L2 variety	English	68	94	English	49
L2 variety	Japanese	4	6	MFL	51
Number of L2s	Multiple L2s	10	14	Multiple L2s	7
Number of L2s	One L2	62	86	One L2	93
Institution type	University	31	43	University	45
	High school	22	31	Primary and secondary	32
	Private training	19	26	Private and state-funded	23
Teaching experience	More experienced	44	61	More or equal to 5 years	70
Teaching experience	Less experienced	28	39	Less than 5 years	30
Teacher qualification	Certified	51	71	Advanced	82
Teacher qualification	Uncertified	21	29	Preliminary	18
Regional socioeconomic status	More advantaged	23	68	n/a	n/a
Regional socioeconomic status	Less advantaged	49	32	n/a	n/a
Course type	Speaking-focused	10	14	n/a	n/a
Course type	General language	62	86	n/a	n/a

Notes. The original study only provided percentages for sub-categories. L2 = second language. MFL = modern foreign languages (e.g. Italian, German). n/a = not applicable.

Additionally, participants’ teaching experience showed variability, but its distribution suggested a two-way division: more experienced and less experienced. Specifically, 61% of participants had an average teaching length of 11 years (range: 5–32 years, SD = 6.56), categorizing them as more experienced teachers. The remaining 39% of participants were less experienced teachers, with an average teaching length of only 2.7 years (range: 0.5–4 years, SD = 1.14). We also collected information on teacher certification, which revealed that 71% of participants held nationally and internationally recognized certificates, such as TESOL (Teaching English to Speakers of Other Languages) Certificate, TEFL (Teaching English as a Foreign Language) Certificate, CELTA (Certificate in English Language Teaching to Adults), Cambridge TKT (Teaching Knowledge Test), and Higher Education Qualification Certificate.

According to the most recent census published by China’s National Bureau of Statistics (2021), regions in Northwest China were less socioeconomically advantaged compared to regions in East China and the four municipality cities (Beijing, Shanghai, Tianjin, and Chongqing), based on regional gross domestic product (GDP) and the average educational level of the population. Following these criteria, 68% of teachers in this sample worked in the less socioeconomically advantaged regions, including the Xinjiang Uygur Autonomous Region, Qinghai Province, and Ningxia Province, while 32% of them worked in more socioeconomically advantaged regions, including Beijing, Shanghai, Guangdong Province, and Zhejiang Province. Participants’ sociodemographic information further revealed that 14% of them taught speaking-focused courses, while the majority (86%) taught general language courses.

As is evident, the participant characteristics in both studies were similar in terms of the distributions of institution type and teaching experience, but differed slightly in terms of L2 variety, number of L2s, and teacher qualifications. Furthermore, the current study collected information on regional socioeconomic status and course type.

2 Instruments

We adopted the questionnaire developed by Tavakoli and Hunter (2018) and subsequently tested by Tavakoli (2023) to capture L2 teachers’ definitions, perceptions, and self-reported teaching practices regarding fluency. The original questionnaire was in English, which presented challenges for JSL teachers in fully understanding its content. To address this, we translated all the items into Chinese, creating a bilingual questionnaire that included both English items and their Chinese translation equivalents, ensuring better participant comprehension.

Since the original English version of the questionnaire was developed in a British context, translation and cultural differences might hamper the measurement of the English–Chinese bilingual questionnaire used in China, given its distinct cultural and educational context. To enhance the validity and reliability of the bilingual questionnaire, we undertook three steps. First, we consulted the online Cambridge Dictionary (English–Chinese Simplified) to obtain Chinese translation equivalents for fluency and fluent, which resulted in 流利 (‘the ability to speak or write a language easily, well, and quickly’) and 流利的 (‘when a person is fluent, they can speak a language easily, well, and quickly’), respectively. These findings indicate that the colloquial use of fluency and fluent to denote proficiency is consistent in both English and Chinese. We further validated the Chinese translation equivalence of these terms by reviewing fluency papers written in Chinese and published in Chinese journals, where 流利 (‘fluency’) and 流利的 (‘fluent’) were consistently employed. Second, we maintained the style, order, and number of items in the bilingual questionnaire to match the English version. Third, before its formal use, we piloted the bilingual questionnaire with 5 university and 5 high school L2 teachers in China (who did not participate in the main study) and engaged in discussions with them to gather feedback.

Specifically, the questionnaire consisted of 6 sections, comprising short answer questions, Likert-scale items, and open-ended descriptions (see Appendix A in supplemental material). Section 1 requested teachers to list and comment on the characteristics of fluent L2 speech. Section 2 asked for their definitions of fluent L2 speakers. In Section 3, there were 11 4-point Likert-scale items focusing on their opinions about the features of L2 oral fluency, their experience in teaching and researching it, and their confidence in promoting fluency through teaching practices. To provide participants with space to illustrate their teaching practices, Section 4 asked for three examples of classroom activities they used to enhance fluency. Section 5 inquired about the importance of promoting fluency. Finally, participants were asked to provide their sociodemographic and contact information.

Following the suggestion from Tavakoli and Hunter (2018), we incorporated semi-structured interviews guided by 8 questions into the research design to triangulate the data source (see Appendix B in supplemental material).

3 Data collection

The questionnaire was created using the online platform Microsoft Forms, and its link was shared with targeted L2 teachers in China. As described in Section IV.1, invitation letters were sent to teachers before the formal study. Out of the invited teachers, 72 agreed to take part, and subsequently, all of them completed the questionnaire, resulting in a questionnaire response rate of 100%. Participants’ consent was obtained at the beginning of the survey. All responses were written in Chinese and were downloaded as an Excel sheet for further analysis.

The follow-up interviews, conducted in Chinese, took place approximately 3 weeks after we collected all the questionnaires. Before the interviews, we conducted a preliminary analysis of participants’ demographic data and their survey answers, focusing the interview questions on these survey responses. Specifically, we conducted one-on-one semi-structured interviews with 11 participants. This focal group represented a substantial portion of the broader participant sample, as their selection was based on the consideration of various teacher factors. The interviewees consisted of 3 university teachers, 2 high school teachers, 3 private training school teachers, and 3 teachers with extensive experience in teaching speaking. These teachers also differed in terms of their work regions and teaching experience. The duration of these interviews ranged from 25 to 42 minutes (M = 33, SD = 5.48).

The interview data was qualitatively analysed using content analysis and served as a supplementary data source to interpret the relevant questionnaire results. Creswell and Plano Clark (2011, p. 81) argued that the qualitative analysis of semi-structured questions is an add-on to the quantitative items, as it allows for the identification of ‘emergent themes and interesting quotes’ that complement and validate the findings of the quantitative analysis. As the interviews were conducted in Chinese, we translated all the transcriptions into English and referred only to the English translation in the following sections.

V Analysis and results

Research question 1: How do L2 teachers in China define fluency and fluent L2 speakers?

To address research question 1, we initially followed Tavakoli and Hunter (2018, p. 337) in conducting a lexical frequency analysis of the responses in Sections 1 and 2 of the questionnaire. Additionally, we conducted a content analysis of the answers to Questions 1–5 in the follow-up interviews to supplement the relevant questionnaire results.

We extracted all the relevant questionnaire responses and compiled a corpus of 3,113 words in which teachers defined fluency and fluent L2 speakers. Subsequently, we conducted a lexical frequency analysis that identified 326 items. Each item represented a lexical chunk or short phrase that expressed a single, coherent meaning. Following the original study, we employed an inter-rater coding approach to categorize all these items and calculated inter-rater reliability as a percentage agreement. The original study identified four themes: fluency in a narrow sense, speaking ability, general L2 proficiency, and other vague or uninformative terms unrelated to fluency. To replicate this categorization, the first author developed a codebook with the names of these four themes, their content, and two examples of each theme (e.g. theme 1: fluency in a narrow sense; its content: speed, breakdown, and repair; its examples: appropriate speed, no salient pauses and hesitations). Both authors discussed the content of the codebook and made modifications until full agreement was reached. We took three steps to code the data. First, based on the codebook, the first author coded all the items, while the second author coded approximately half of the items (n = 150). Second, we calculated the inter-rater agreement rate (89%, n = 133/150) and then resolved all disagreements through discussion. Third, the second author coded an additional 90 items, resulting in an increased agreement rate of 96% (n = 86/90) (these disagreements were resolved by discussion). Thus, the overall inter-rater agreement rate for coding was high at 91% (n = 219/240).

The results of the lexical frequency analysis and item categorization in both studies are summarized in Table 2. In Tavakoli and Hunter (2018), the average number of lexical items per participant was 5.4 (452/84), which was higher than that in the replication study (4.5, 326/72). Regarding the ranking of the four themes, both studies were consistent, with speaking ability ranking first, overall L2 proficiency second, fluency third, and other vague or irrelevant answers fourth. However, the distributions of each theme differed. In the replication study conducted in China, 21.8% of the items defined fluency in a narrow sense, focusing on aspects of utterance fluency such as speed, breakdown, and repair. Examples included definitions like ‘appropriate speed, duration, and place of pauses’ and ‘no salient pauses and hesitations, such as fillers “um” and “uh”.’ In the interviews, teachers mentioned that fluency meant ‘not getting stuck but pausing reasonably’, that ‘speaking fast doesn’t necessarily mean being fluent’, and that ‘speed is not crucial, but it can’t be too fast or too slow’ because even ‘native speakers also pause according to expressive meaning and use discourse markers, fillers, and self-corrections to facilitate conversations’. In contrast, only 13.4% of the items in the original study related to definitions of fluency in a narrow sense.

Table 2.

Teachers’ definitions of fluency in both studies.

Categories	Current study (n = 326)			Original study (n = 452)
	Frequency	Percentage	Examples	Frequency	Percentage
Fluency (narrow sense)	71	21.8	合适的语速、时长和停顿的位置 appropriate speed, duration, and place of pauses; 没有明显的停顿和卡壳，比如’嗯’、’啊’ no salient pauses and hesitations, such as ‘um’, ‘ uh’	61	13.4
L2 speaking ability	121	37.1	准确的语音语调 accuracy of pronunciation and intonation; 流利表达自己的观点 fluently expressing personal opinions; 类似母语者的说法方式 native-like ways of speaking; 语调和语境适配 intonation fits with the expressive context; 超音段的准确性 suprasegmental accuracy	198	43.8
General L2 proficiency	107	32.8	用实词 use of content words; 只有少许语法错误 making very few grammatical mistakes; 熟悉多种句法结构 familiar with a range of syntactic structures; 词汇语法的准确度 lexico-grammatical accuracy	147	32.5
Other vague or irrelevant answers	27	8.3	母语迁移 transfer effects of mother tongue; 难以量化 difficult to quantify; 教师看不到它的重要性 teachers do not see its importance	46	10.2

Notes. For their examples, see Tavakoli and Hunter (2018, p. 338).

In the second category, speaking ability (37.1% of the items), the examples indicated that the aspects of fluency were similar to those of speaking performance. The most frequent items in this category included ‘accuracy of pronunciation and intonation’, specifically, ‘intonation and suprasegmental accuracy’, ‘native-like ways of speaking’, ‘fluently expressing personal opinions’, and ‘intonation that fits with the expressive context, for example, raising the sound volume and changing the tone when emphasizing points’. In the interviews, teachers often mentioned L2 speakers’ comprehensible accent, accurate pronunciation, and intonation. Although this category comprised the largest proportion of the data in both studies, the percentage of this category was higher in the original study, at 43.8%.

The third category encompassed broader views of fluency. 32.8% of the items viewed fluency as overall L2 proficiency. The frequently used terms included ‘no salient grammatical errors’, ‘use of content words’, ‘proper understanding of interlocutors’ content’, ‘replacement of complex words with simpler ones’, ‘lexico-grammatical accuracy’, and ‘adequate mastery of L2 culture and idioms’. In the interviews, teachers also mentioned that ‘fluent L2 speakers have very few grammatical errors in their speech, and watching them speak brings a sense of visual enjoyment. Additionally, the vocabulary, grammar, and syntax in their speech are clear, logical, and well-organized.’ The percentage of this category here was similar to that in the original study, at 32.5%.

In the replication study, 8.3% of the items were irrelevant to the definitions of fluency or vague in their referential meanings, such as ‘difficult to quantify’, ‘native speakers’, ‘fluent’, and ‘no occasion to speak English’. In contrast, a slightly higher percentage of vague or irrelevant terms, 10.2%, was reported in the original study.

Research question 2: What classroom practices do L2 teachers in China use to promote fluency?

The data used to address research question 2 was from Section 4 of the questionnaire, where each participant was required to report three activities they used to promote fluency in class. The expected number of examples was 216, but 12 slots were left blank, resulting in a final number of reported examples as 204. The blank slots were coded as ‘None’. The self-reported activities underwent a three-step annotation process. First, the first author developed a codebook based on categories from Rossiter et al. (2010) and Tavakoli and Hunter (2018). However, unlike the studies conducted in Canada and the UK, the L2 teachers in China did not report any consciousness-raising activities (e.g. ‘asking students to listen to their recorded performance’ (Tavakoli & Hunter, 2018, p. 341)). In contrast to the original study, they reported some activities that were not directly related to promoting fluency in a pedagogic context, such as ‘browsing English websites’. These activities were classified as a new category called ‘Others’. The categorization criteria were modified accordingly, resulting in five categories in the codebook. Second, both authors discussed the codebook, including category names and examples, until reaching full agreement (e.g. name: fluency-focused activities; examples: tracking or shadowing, pre-task planning time, 4/3/2 technique). Third, both authors independently coded the entire dataset. Cohen’s Kappa was calculated to assess inter-rater consistency, yielding a high value of κ = 0.925, p < 0.001, indicating strong inter-rater reliability. Any disagreements were resolved through discussion.

The frequency distribution of activities in both studies is summarized in Table 3. In the current study, 15.3% of the reported activities fell under the fluency-focused category, comprising 33 examples. This percentage was much higher than that in the original study (2.8%, n = 7/252). The fluency-focused category encompassed three types of activities: planning, rehearsal and repetition tasks (n = 19), tracking and shadowing tasks (n = 8), and the 4/3/2 technique (n = 6). The most prevalent activities within this category were planning, rehearsal, and repetition tasks, with 2 examples related to time planning (e.g. ‘managing pre-task planning time by repeating the same task’) and 17 examples involving rehearsal and repetition tasks (e.g. ‘taking turns to build the plot for the same story’). Tracking tasks, where learners repeat an aural passage while listening to it, and shadowing tasks, where they repeat it slightly after the speaker, were described in 8 examples. Additionally, the 4/3/2 technique, recognized as a highly effective fluency-promoting task, was mentioned in 6 examples. Conversely, activities related to fluency strategy training constituted only 2.8% of the examples (n = 6/216), which was lower than that in the original study (6.0%, n = 15/252). These 6 activities included 4 examples of using formulaic language, 1 example involving the use of fillers and repair, and 1 example explaining the definitions/aspects of fluency.

Table 3.

Teachers’ self-reported activities for promoting fluency in both studies.

Categories	Current study (n = 216)			Original study (n = 252)
	Frequency	Percentage	Examples	Frequency	Percentage
Fluency-focused	33	15.3	影子练习shadowing; 四/三/二 4/3/2; 重复相同的任务来把握任务前的计划时间managing pre-task planning time by repeating the same task	7	2.8
Fluency strategy training	6	2.8	解释流利度的定义和内涵explaining the definitions/aspects of fluency; 练习填补和修补practicing fillers and repairs; 练习程式语practicing formulaic language	15	6.0
Communicative free production	132	61.1	角色扮演role-play; 视频配音video dubbing; 辩论debate; 对话conversation; 图片描述picture description	135	53.6
General L2 proficiency	23	10.6	即时短文写作instant essay writing; 短文阅读essay reading;翻译translation	34	13.5
Others	10	4.6	浏览英语网站browsing English websites;出国旅行travelling abroad	n/a	n/a
None	12	5.6	No examples provided	57	22.6

Notes. An additional 1.6% of the activities reported in Tavakoli and Hunter (2018, p. 341) were categorized as consciousness-raising. n/a = not applicable.

The majority of examples belonged to communicative free production tasks, which appeared 132 times and accounted for 61.1% of the total number of examples. This percentage was higher than that in the original study (53.6%, n = 135/252). Within this category, there were 100 dialogic production activities, such as role-plays, debates, conversations, group discussions, and map tasks, which frequently appeared. Another 32 examples pertained to monologic production activities, such as picture descriptions and presentations.

Furthermore, 23 examples were not aimed at promoting fluency or speaking ability. These activities, such as instant essay writing, essay reading, listening to foreign language songs, and translation, were coded as ‘General L2 proficiency’ activities. This category accounted for 10.6% of the examples, which was lower than that in the original study (13.5%). Regarding activities that were not directly related to promoting fluency in a pedagogic context, the participants reported 10 examples, such as browsing English websites, co-teaching with a native speaker, and traveling abroad.

In summary, the results of both studies revealed variations in the specific categories of self-reported classroom activities and their respective percentages. In the current study, no consciousness-raising activities were reported by teachers in China, but other activities unrelated to direct fluency promotion were mentioned. In terms of percentage differences, compared to the original study, higher percentages of fluency-focused and free production tasks were reported in the current study. However, lower percentages were observed for the other two categories, namely fluency strategy training and general L2 proficiency tasks.

Research question 3: What are the effects of teacher variables on L2 teachers’ confidence in their knowledge about fluency in China?

To address research question 3, which was not explored in Tavakoli and Hunter (2018), we examined six teacher variables described in Section IV.1. These variables included the number of L2s, institution type, regional socioeconomic status, teaching experience, teacher certification, and course type. Additionally, we investigated the 11 items in Section 3 of the questionnaire, with each item responded to on a 4-point Likert scale ranging from 4 (to a large extent) to 3 (to some extent) to 2 (to a limited extent) to 1 (hardly at all). These items measured three specific constructs: L2 teachers’ understanding of fluency, their familiarity with fluency research, and their confidence in promoting fluency in their classrooms (Tavakoli & Hunter, 2018, p. 336). Henceforth, we used the term ‘L2 teachers’ confidence in their knowledge about fluency’ to represent these constructs. To calculate the scores of teachers’ confidence in their knowledge about fluency, we added up the score of each item for each participant and then divided it by the total number of items (n = 11), resulting in an average score for each participant. The range of the scores was from 1.6 to 4.0, with a median of 3.0 (M = 2.94, SD = 0.46). A higher score corresponded to a more scholarly definition of fluency and, consequently, a greater confidence in knowledge about fluency in a narrow sense. Scale analysis of the items revealed high internal reliability (Cronbach alpha = 0.887).

For the statistical analysis, we employed linear regression modeling to simultaneously investigate the main and interaction effects of teacher variables on L2 teachers’ confidence in their knowledge about fluency. We fitted a maximal model using the lm function in the lme4 package (Bates et al., 2015) in R (R Core Team, 2023) with predictors L2Numbers, Institution, Region, Teaching Experience, Teacher Certification, and Course Type. Subsequently, we followed the backward elimination procedure to prune the models, testing and removing interaction and fixed effects if they were not significant (alpha level set at 0.05) (Zuur et al., 2010). Model fit was assessed using the anova function in the car package (Fox & Weisberg, 2019), and when model fit was not significantly different, the simpler model was selected.

The final model, which predicted teachers’ confidence scores in their fluency knowledge, revealed significant main effects for 4 predictors and 3 two-way interaction effects. Notably, Institution did not exhibit a main effect. The variable Course Type lacked both a main effect and an interaction effect with other predictors, leading to its exclusion during the model selection process.

Table 4 provided a summary of the coefficient estimates (β), standard errors, and levels of significance for each predictor. These coefficient estimates provided insights into the relative magnitudes of the effects for predictors. Among the main effects, Region displayed the largest effect size (β = −0.54), followed by TeachingExperience (β = −0.50), L2Numbers (β = −0.48), and Certification (β = −0.34). Regarding the interaction effects, the interaction between Institution and L2Numbers exhibited the most substantial effect size (β = 1.02 for one level and 0.49 for the other level). This was followed by the interaction between Region and Certification (β = 0.39) and the interaction between Region and TeachingExperience (β = 0.34).

Table 4.

Model summary for the linear regression on the scores of teachers’ confidence in their knowledge about fluency by the effects of Region (reference level = MoreAdvantaged), TeachingExperience (MoreExperienced), Certification (Certified), Institution (HighSchool), and L2Numbers (OneL2).

	Estimate	Std. error	t value	p
(Intercept)	3.60	0.13	26.74	< 0.001
Region (LessAdvantaged)	–0.54	0.14	–3.74	< 0.001
TeachingExperience (LessExperienced)	–0.50	0.12	–4.11	< 0.001
Certification (Uncertified)	–0.34	0.12	–2.72	< 0.01
Institution (University)	–0.15	0.08	–2.01	0.056
Institution (PrivateSchool)	–0.08	0.10	–0.80	0.422
L2Numbers (MultipleL2s)	–0.48	0.18	–2.61	< 0.01
Region (LessAdvantaged): TeachingExperience (LessExperienced)	0.34	0.14	2.42	< 0.05
Region (LessAdvantaged): Certification (Uncertified)	0.39	0.15	2.60	< 0.05
Institution (University): L2Numbers (MultipleL2s)	1.02	0.27	3.74	< 0.001
Institution (PrivateSchool): L2Numbers (MultipleL2s)	0.49	0.22	2.21	< 0.05

Notes. Positive estimates of coefficients are associated with higher scores. Bold p-values indicate statistically significant effects.

To aid further interpretation of the directions of these effects, we visualized them in Figures 1 to 4. In each plot, the x-axis represented the specific categories of a teacher variable, while the y-axis represented the predicted scores of teachers’ fluency knowledge.

Figure 1.

Plots of main effects in the regression model.

Figure 2.

Effect plot of the interaction between Institution and L2Numbers.

Figure 3.

Effect plot of the interaction between Region and Certification.

Figure 4.

Effect plot of the interaction between Region and TeachingExperience.

Figure 1 illustrated the results as follows: In the upper left panel, the main effect of teaching experience revealed that teachers with more experience had higher predicted scores than those with less experience. This suggested that experienced teachers exhibited greater confidence in their knowledge about fluency. In the upper right panel, the main effect of certification indicated that certified teachers had higher confidence in their fluency knowledge compared to uncertified teachers. Regarding the effect of regional socioeconomic status, the lower left panel demonstrated that teachers working in more socioeconomically advantaged areas reported greater confidence in their fluency knowledge compared to those in less advantaged areas. As observed in the lower right panel, teachers who taught multiple L2s (e.g. English and Korean) exhibited greater confidence in their knowledge about fluency compared to those who taught only one L2.

The significant interaction between Institution and Number of L2s was visualized in Figure 2. The figure showed that teachers who taught only one L2 demonstrated reasonable and similar levels of confidence in their fluency knowledge, regardless of the institution type (scores clustered around 3.0). However, for L2 teachers who taught multiple L2s, the institution type had an effect. Teachers of multiple L2s at universities had the highest scores of confidence in their knowledge about fluency, while those in high schools had the lowest scores, and those in private language training schools fell in between.

Figure 3 illustrated the significant interaction between Region and Certification. Certified teachers working in more socioeconomically advantaged regions displayed the highest scores of confidence in their knowledge about fluency. Conversely, certified teachers working in less socioeconomically advantaged regions exhibited the lowest levels of confidence in their fluency knowledge. Regarding uncertified teachers, they showed similar levels of confidence in their knowledge about fluency regardless of whether they worked in more or less socioeconomically advantaged areas.

Additionally, an interaction between Region and Teaching Experience was observed, as depicted in Figure 4. This interaction revealed that more experienced teachers in more socioeconomically advantaged regions displayed the highest levels of confidence in their knowledge about fluency. Conversely, less experienced teachers in more socioeconomically advantaged regions exhibited the lowest levels of confidence in their fluency knowledge. Within the less socioeconomically advantaged regions, more experienced teachers showed a higher degree of confidence in their knowledge about fluency compared to their less experienced counterparts.

VI Discussion and implications

In this study, we replicated the first two research questions from Tavakoli and Hunter (2018), so we discussed our results in relation to those of the original study. For research question 1, in line with the original study, our findings indicated that the majority of L2 teachers in China defined fluency broadly, considering it as a comprehensive term encompassing speaking ability and general L2 proficiency. Additionally, we replicated the four-level model of defining fluency identified in the original study, which includes a very narrow perspective (speed, repair, and breakdown), a narrow perspective (ease, continuity, and flow), a broad perspective (pronunciation, speech accuracy and complexity), and a very broad perspective (general L2 proficiency). These results collectively highlight a lack of clarity within the language teacher community regarding fluency, despite its established definitions by SLA researchers. This finding was unexpected considering the longstanding recognition of fluency as a significant component of communicative competence (e.g. Bachman, 1990; Skehan, 2003) and the extensive research conducted on its definitions and aspects (e.g. Lennon, 1990; Segalowitz, 2010; Tavakoli & Skehan, 2005). One possible explanation for this lack of clarity is the complex and multifaceted nature of fluency (Freed, 2000; Segalowitz, 2010, 2016; Tavakoli, 2016), which makes it challenging to define precisely. During the interviews, when we asked the teachers to elucidate the relationship between fluency, speaking ability, and general L2 proficiency, they unanimously agreed that fluency is distinct from the other two concepts. The most commonly mentioned responses were ‘fluency is just one aspect of speaking ability, separate from complexity and accuracy’ and ‘fluency serves as an indicator of general L2 proficiency, but it does not equate to it’. However, in the questionnaire responses, the majority of definitions (69.9%) referred to fluency in a broad sense, focusing on accuracy, complexity, and lexis. Conversely, a relatively small proportion of items (21.8%) conceptualized fluency in a narrow sense, emphasizing speed, breakdown, and repair of speech. These patterns indicate that although teachers believe they possess a good understanding of fluency, they face challenges in providing a precise definition for the term due to its complex and multifaceted nature. The observation highlights the differentiation between three aspects of ‘construct’ in teachers’ cognition: ‘theoretical construct’ (theories of a construct), ‘perceived construct’ (individual understanding, beliefs, and attitudes towards a construct), and ‘stated construct’ (explicit description of a construct) (Macqueen, 2022, pp. 243–244).

There were notable variations in the distribution of different types of definitions between the two studies. In the UK, only 13.4% of items defined fluency in a narrow sense, which was lower than the proportion reported by teachers in China. Furthermore, the percentage of definitions that conceptualized fluency as speaking ability was smaller in the Chinese context compared to the UK context (37.1% vs. 43.8%). However, the proportions for the third category, general L2 proficiency, were similar in both studies. In Morrison’s (2018) study of high school teachers of English as a foreign language (EFL) in Chile, the rankings of definition types were also different. Only 10.6% of the items defined fluency in a narrow sense, which was lower than the figures reported in the UK and China. The percentages of items conceptualizing fluency as speaking ability and general L2 proficiency were similar, at 28.1% and 28.3%, respectively, which were also lower than those reported in the UK and China. These findings demonstrate the variations in the extent of understanding fluency among L2 teachers in different educational contexts. However, there is a consensus that fluency is not a dichotomous construct with only narrow and broad definitions (Lennon, 1990). Instead, fluency is defined as a continuum of language performance, encompassing multiple language features such as speed, breakdown, pronunciation, accuracy, and complexity. These findings align with prior research, which emphasizes the intricate nature of fluency (e.g. Chambers, 1997; Freed, 2000; Segalowitz, 2010; Tavakoli, 2016). They provide additional evidence from the teachers’ perspective, further supporting the notion that understanding fluency poses challenges for teachers (Tavakoli & Hunter, 2018).

In relation to research question 2, the results revealed that for L2 teachers in China, the majority of their reported activities aimed at promoting fluency were free production tasks, with a relatively small proportion of fluency-focused activities reported. These are in line with the original study conducted with a sample from the UK. Collectively, these findings suggest that teachers’ confusion regarding the distinctions between fluency, speaking ability, and general L2 proficiency influences their reported teaching practices related to promoting fluency. It has been argued that free production tasks are primarily designed to enhance learners’ speaking performance and overall L2 proficiency, so these tasks alone are unlikely to significantly foster fluency (e.g. Rossiter et al., 2010). However, the results of both studies indicate that these research-based recommendations have not been widely embraced by language teachers in the UK and China. This may be attributed to the lack of communication between the SLA researcher community and the language teacher community, a gap that has been frequently reported in previous studies (Foster & Hunter, 2016; Nassaji, 2012; Tavakoli, 2015). Our interview data yielded additional supporting evidence, with 9 out of the 11 teachers expressing their unfamiliarity with fluency research. They attributed this lack of familiarity to the limited time and restricted access available to engage with research materials. Another factor hindering the implementation of fluency-promoting practices recommended by researchers is teachers’ negative attitudes towards certain practices. For example, a study conducted by Ahmadian et al. (2017, p. 473) on L2 teachers’ and learners’ perceptions of task repetition revealed that ‘virtually all teachers’ considered this activity to be ‘boring and disinteresting’, whereas most learners recognized it as a valuable classroom task (e.g. Lambert et al., 2016). This disparity in attitudes further underscores the challenges in adopting and integrating research-based practices into classroom teaching.

Additionally, differences were identified in the categories of teachers’ reported practices and their rankings between the two studies. A notable distinction emerged in the absence of awareness-raising activities reported by teachers in China, while in the original study, awareness-raising tasks accounted for 1.6% of the examples provided by teachers in the UK, albeit being the smallest category. This discrepancy may be attributed to variations in the tasks emphasized in textbooks, teaching materials, and training resources used in each context. However, further research on these materials is needed to validate this assumption. A shared finding in both studies is that awareness-raising activities are rarely reported. This indicates that such activities are relatively infrequently integrated into the materials, aligning with similar findings from Canada, where awareness-raising activities were found to have the smallest representation in both learner and teacher materials (Rossiter et al., 2010). Regarding the rankings of categories, in the original study, only 10.4% of the reported activities aligned with those suggested as fluency-enhancing by research (e.g. Boers, 2014; de Jong & Perfetti, 2011; Tavakoli et al., 2016). In contrast, the replication study exhibited a larger proportion of these activities, at 18.1%. Furthermore, while the proportion of reported activities that aimed to enhance speaking ability and overall L2 proficiency was similar between the two studies, the replication study slightly surpassed the original study (71.7% vs. 67.1%). As shown in Table 2, teachers in China exhibited a higher proportion of narrowly-defined fluency definitions compared to teachers in the UK, but both studies demonstrated similar proportions in broadly-defined fluency definitions. These findings provide further evidence supporting the notion that teachers’ understanding of fluency impacts their teaching approaches to promoting it. They also underscore the intricate nature of fluency, which presents challenges for teachers ‘at both conceptual and practical levels’ (Tavakoli & Hunter, 2018, p. 331).

In research question 3, a novel inquiry in the replication study, we examined how teacher variables influence teachers’ confidence in their knowledge about fluency in China. The results indicated a mean score of 2.94 on a 4-point scale, suggesting that teachers in China possessed a reasonable level of confidence in their fluency knowledge. This finding corresponded with the original study. However, it was observed that teachers’ confidence levels in their fluency knowledge were influenced by both main and interaction effects among the teacher variables. Specifically, the main effect of number of L2s taught, along with its interaction with institution type, demonstrated that university teachers with experience in teaching multiple L2s exhibited the highest level of confidence in their fluency knowledge compared to high school and training school teachers. These effects can be attributed to the education hierarchy present in China. University teachers with experience in teaching multiple L2s tend to have more extensive formal education and research experience in SLA. This may partially explain why they possess the greatest confidence in their knowledge about fluency. Another contributing factor could be that L2 learners at the university level generally demonstrate better speaking performance compared to those in private training and high schools. As a result, university teachers need to be well-prepared in both theoretical knowledge and practical language teaching techniques (You & Dörnyei, 2016), motivating them to actively enhance their understanding of the distinctions among fluency, speaking performance, and L2 proficiency.

Furthermore, the main effect of regional socioeconomic status and its respective interactions with certification and teaching experience revealed that certified and experienced teachers in more socioeconomically advantaged areas exhibited higher levels of confidence in their fluency knowledge. There are several explanations for these findings. First, more socioeconomically developed regions in China, such as Beijing, Shanghai, and Zhejiang Province, often provide teachers with increased access to professional training opportunities for teaching (Cao, 2010; Hu, 2003). As a result, these teachers tend to have a better understanding of fluency compared to teachers in less developed regions (see the effect of teacher training on fluency in Tavakoli, 2023). Another factor is that more developed regions have a longer history and higher demand for foreign language learning and teaching. Consequently, there is a greater presence of experienced teachers in these areas who have acquired a better understanding of fluency through research and teaching practices (Dore, 2016). These teachers benefit from the accumulated knowledge and expertise prevalent in these regions. These findings highlight the multifaceted nature of teachers’ confidence in their knowledge about fluency (e.g. Khau & Huynh, 2022).

Both studies yield several implications that can be drawn from their findings. First, it is recommended to define fluency in a narrow sense, emphasizing aspects such as speech speed, repair, and breakdown (e.g. Freed, 2000; Lennon, 1990; Tavakoli & Hunter, 2018). This focused definition can guide the selection of fluency-focused classroom teaching activities, such as the 4/3/2 technique and task repetition, which have shown to be effective in improving learners’ fluency (e.g. de Jong & Perfetti, 2011; Foster, 2020; Tran & Saito, 2021). By prioritizing these activities, teachers can provide more targeted and efficient training to enhance learners’ fluency. Second, it is crucial to enhance the accessibility of pedagogic recommendations derived from fluency studies for the teacher community. Organizing workshops and seminars on fluency, involving both SLA researchers and language teachers, can facilitate the establishment of common ground for fluency teaching (e.g. Tavakoli, 2023). These collaborative platforms create opportunities for knowledge exchange and can influence teachers’ classroom practices to be more fluency focused. By improving the dissemination and application of fluency-related pedagogical insights, teachers can better integrate them into their teaching approaches. Third, contextual factors such as teaching experience and teacher qualifications should be considered when examining teachers’ understanding of fluency and their readiness to address it in their language teaching practices. Understanding how these factors influence teachers’ knowledge about fluency can provide valuable insights into designing professional development programs and support systems tailored to the specific needs of teachers in different contexts (e.g. Khau & Huynh, 2022; Tavakoli, 2023).

VII Limitations

There were three major limitations in this replication study. First, similar to Tavakoli and Hunter (2018), we did not collect any classroom observation data in this study. The primary objective was to replicate the original study, focusing on investigating the relationship between teachers’ understanding of fluency and their self-reported practices for promoting it. We expanded upon the original study by examining the impact of teachers’ variables on their confidence in fluency knowledge and incorporating interviews as an additional data source. To gain insights into teachers’ actual classroom practices for promoting fluency and explore the relationship between their understanding of fluency, self-reported practices, and observed practices, employing a different research design may be necessary. Such an approach would warrant its own independent study. Second, the study did not conduct a survey of the textbooks and learning materials used in various institutions and regions in China, which limited the interpretability of the findings. Third, although the sample size in this study was similar to that of the original study, it was still relatively small-scale, thus limiting the generalizability of the findings.

VIII Conclusions

This study on teachers’ perceptions of fluency and their self-reported practices for promoting it in China substantially replicated the original study conducted in the UK. The results were consistent with Tavakoli and Hunter (2018), indicating that the majority of teachers in China held a broad perspective of fluency, perceiving it as a cover term encompassing speaking ability. Consequently, teachers’ understanding of fluency influenced their instructional practices, with free production tasks being more commonly reported than fluency-focused activities suggested by research. Furthermore, this study extended the original study by examining the impact of teacher variables on their confidence in their knowledge about fluency in China. The analysis revealed that certified teachers, teachers with teaching experience in multiple L2s, more experienced teachers, and teachers in more socioeconomically advantaged regions exhibited higher levels of confidence in their fluency knowledge. Additionally, interaction effects among different teacher variables were observed. The implications drawn from both studies underscore the importance of narrowing the definition of fluency, enhancing the accessibility of pedagogic recommendations, and considering contextual factors that influence teachers’ confidence in their fluency knowledge. By addressing these implications, language educators can foster a more effective approach to teaching fluency in the language classroom.

Supplemental Material

sj-docx-1-ltr-10.1177_13621688231186857 – Supplemental material for Examining the perceptions and self-reported practices of L2 teachers in China regarding oral fluency: A conceptual replication and extension

Supplemental material, sj-docx-1-ltr-10.1177_13621688231186857 for Examining the perceptions and self-reported practices of L2 teachers in China regarding oral fluency: A conceptual replication and extension by Qiao Gan and Lin Ma in Language Teaching Research

Footnotes

Acknowledgements

We are very grateful to Irini Mavrou and Susy Macqueen for their support in the design and writing of this study, and to all the teachers who participated in this project. We thank the Language Teaching Research editors and two anonymous reviewers for their constructive feedback on an earlier draft of the article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Qiao Gan

Supplemental material

Supplemental material for this article is available online.

References

Ahmadian

M.J.

Mansouri

Ghominejad

(2017). Language learners’ and teachers’ perceptions of task repetition. ELT Journal, 71, 467–477.

Ahmadian

M.J.

Tavakoli

(2011). The effects of simultaneous use of careful online planning and task repetition on accuracy, complexity, and fluency in EFL learners’ oral production. Language Teaching Research, 15, 35–59.

Bachman

L.F.

(1990). Fundamental considerations in language testing. Oxford University Press.

Bates

Mächler

Bolker

Walker

(2015). Fitting linear mixed-effects models using lme4. Journal of Statistical Software, 67, 1–48.

Boers

(2014). A reappraisal of the 4/3/2 activity. RELC Journal, 45, 221–235.

Borg

(2013). Teacher research in language teaching: A critical analysis. Cambridge University Press.

Cao

(2010). Urban-rural income disparity and urbanization: What is the role of spatial distribution of ethnic groups? A case study of Xinjiang Uyghur Autonomous Region in Western China. Regional Studies, 44, 965–982.

Chambers

(1997). What do we mean by fluency? System, 25, 535–544.

Chan

(2019). Effects of pushed production of single words and multiword patterns on L2 oral fluency: Some evidence from temporal measurements. International Review of Applied Linguistics in Language Teaching, 57, 147–179.

10.

China’s National Bureau of Statistics. (2021). Regional socio-economic status. National Bureau of Statistics.

11.

Creswell

J.W.

Plano Clark

V.L.

(2011). Designing and conducting mixed-methods research. Sage.

12.

de Jong

N.H.

Bosker

H.R

. (2013). Choosing a threshold for silent pauses to measure second language fluency. In Eklund

(Ed.), Proceedings of disfluency in spontaneous speech (pp. 17–20). Royal Institute of Technology.

13.

de Jong

N.H.

Perfetti

C.A

. (2011). Fluency training in the ESL classroom: An experimental study of fluency development and proceduralization. Language Learning, 61, 533–568.

14.

Dewaele

J.-M.

Witney

Saito

Dewaele

(2018). Foreign language enjoyment and anxiety: The effect of teacher and learner variables. Language Teaching Research, 22, 676–697.

15.

Dore

(2016). Perceptions of fluency. British Council ELT Master’s Dissertation Awards: Winner. British Council.

16.

Dörnyei

(2007). Research methods in applied linguistics. Oxford University Press.

17.

Dudley

(2007). Connecting L2 learners to the larger community. The Canadian Modern Language Review, 63, 539–561.

18.

Ellis

(2009). The differential effects of three types of task planning on the fluency, complexity, and accuracy in L2 oral production. Applied Linguistics, 30, 474–509.

19.

Foster

(2020). Oral fluency in a second language: A research agenda for the next ten years. Language Teaching, 53, 446–461.

20.

Foster

Hunter

A.-M.

(2016). When it’s not what you do but the way that you do it. In Tomlinson

(Ed.), SLA research and materials development for language learning (pp. 280–298). Routledge.

21.

Fox

Weisberg

(2019). An R companion to applied regression. Sage.

22.

Freed

(2000). Is fluency, like beauty, in the eyes (and ears) of the beholder? In Riggenbach

. (Ed.), Perspectives on fluency (pp. 243–265). University of Michigan Press.

23.

Freed

Segalowitz

Dewey

(2004). Context of learning and second language fluency in French: Comparing regular classroom, study abroad, and intensive domestic immersion programs. Studies in Second Language Acquisition, 6, 275–301.

24.

Gass

Mackey

Alvarez-Torres

M.J.

Fernández-García

(1999). The effects of task repetition on linguistic output. Language Learning, 49, 549–581.

25.

Gatbonton

Segalowitz

(1988). Creative automatization: Principles for promoting fluency within a communicative framework. TESOL Quarterly, 22, 473–492.

26.

Housen

Kuiken

Vedder

(Eds.). (2012). Dimensions of L2 performance and proficiency: Complexity, accuracy and fluency in SLA: Volume 32. John Benjamins.

27.

G.W.

(2003) English language teaching in China: Regional differences and contributing factors. Journal of Multilingual and Multicultural Development, 24, 290–318.

28.

Hunter

A.-M.

(2017). Fluency development in the ESL classroom: The impact of immediate task repetition and procedural repetition on learners’ oral fluency. Unpublished PhD thesis, University of Surrey, Guildford, UK.

29.

Khau

A.H.

Huynh

V.T.M.

(2022). An investigation into oral fluency perceived by teachers and students in a Vietnamese context of English education. Language Testing in Asia, 12, 27.

30.

Koponen

Riggenbach

(2000). Overview: Varying perspectives on fluency. In Riggenbach

(Ed.), Perspectives on fluency (pp. 5–24). University of Michigan Press.

31.

Kormos

Dénes

(2004). Exploring measures and perceptions of fluency in the speech of second language learners. System, 32, 145–164.

32.

Lambert

Kormos

Minn

(2016). Task repetition and second language speech processing. Studies in Second Language Acquisition, 39, 167–196.

33.

Lennon

(1990). Investigating fluency in EFL: A quantitative approach. Language Learning, 40, 387–417.

34.

Lynch

Maclean

(2000). Exploring the benefits of task repetition and recycling for classroom language learning. Language Teaching Research, 4, 221–250.

35.

Macqueen

(2022). Construct in assessments of spoken language. In Haug

Mann

Knoch

(Eds.), The handbook of language assessment across modalities (pp. 233–249). Oxford University Press.

36.

Mehnert

(1998). The effects of different lengths of time for planning on second language performance. Studies in Second Language Acquisition, 20, 83–108.

37.

Mizumoto

Plonsky

(2016). R as a Lingua Franca: Advantages of using R for quantitative research in applied linguistics. Applied Linguistics, 37, 284–291.

38.

Morrison

(2018). Fluency in the EFL Chilean classrooms. British Council ELT Master’s Dissertation Awards: Commendation. British Council.

39.

Nassaji

(2012). The relationship between SLA research and language pedagogy: Teachers’ perspectives. Language Teaching Research, 16, 337–365.

40.

Porte

McManus

(2019). Doing replication research in applied linguistics. Routledge.

41.

Préfontaine

(2013). Perceptions of French fluency in second language speech production. The Canadian Modern Language Review, 69, 324–348.

42.

Préfontaine

Kormos

Johnson

D.E.

(2016). How do utterance measures predict raters’ perceptions of fluency in French as a second language? Language Testing, 33, 53–73.

43.

R Core Team. (2023). R: A language and environment for statistical computing. R Foundation for Statistical Computing. Available at: https://www.R-project.org (accessed June 2023).

44.

Riggenbach

(1999). Discourse analysis in the language classroom: Volume 1: The spoken language. University of Michigan Press.

45.

Rossiter

M.J.

Derwing

Manimtim

L.G.

Thomson

R.I.

(2010). Oral fluency: The neglected component in the communicative language classroom. The Canadian Modern Language Review, 66, 583–606.

46.

Sample

Michel

(2014). An exploratory study into trade-off effects of complexity, accuracy and fluency in young learners’ oral task repetition. TESL Canada Journal, 31, 23–46.

47.

Segalowitz

(2010). The cognitive bases of second language fluency. Routledge.

48.

Segalowitz

(2016). Second language fluency and its underlying cognitive and social determinants. International Review of Applied Linguistics in Language Teaching, 54, 79–95.

49.

Seifoori

Vahidi

(2012). The impact of fluency strategy training on Iranian EFL learners’ speech under online planning conditions. Language Awareness, 21, 101–112.

50.

Skehan

(2003). Task-based instruction. Language Teaching, 36, 1–14.

51.

Skehan

(2009). Modelling second language performance: Integrating complexity, accuracy, fluency, and lexis. Applied Linguistics, 30, 510–532.

52.

Suzuki

Kormos

(2023). The multidimensionality of second language oral fluency: Interfacing cognitive fluency and utterance fluency. Studies in Second Language Acquisition, 45, 38–64.

53.

Tavakoli

(2015). Connecting research and practice in TESOL: A community of practice perspective. RELC Journal, 46, 37–52.

54.

Tavakoli

(2016). Fluency in monologic and dialogic task performance: Challenges in defining and measuring L2 fluency. International Review of Applied Linguistics in Language Teaching, 54, 133–150.

55.

Tavakoli

(2023). Making fluency research accessible to second language teachers: The impact of a training intervention. Language Teaching Research, 27, 368–393.

56.

Tavakoli

Campbell

McCormack

(2016). Development of speech fluency over a short period of time: Effects of pedagogic intervention. TESOL Quarterly, 50, 447–471.

57.

Tavakoli

Hunter

A.-M.

(2018). Is fluency being ‘neglected’ in the classroom? Teacher understanding of fluency and related classroom practices. Language Teaching Research, 22, 330–349.

58.

Tavakoli

Nakatsuhara

Hunter

A.-M.

(2020). Aspects of fluency across assessed levels of speaking proficiency. The Modern Language Journal, 104, 169–191.

59.

Tavakoli

Skehan

(2005). Strategic planning, task structure, and performance testing. In Ellis

(Ed.), Planning and task performance in a second language (pp. 239–277). John Benjamins.

60.

Thai

Boers

(2016). Repeating a monologue under increasing time pressure: Effects on fluency, complexity, and accuracy. TESOL Quarterly, 50, 369–393.

61.

Tran

M.N.

Saito

(2021). Effects of the 4/3/2 activity revisited: Extending Boers (2014) and Thai & Boers (2016). Language Teaching Research. Epub ahead of print 27 April 2021. DOI: 10.1177/1362168821994136

62.

Wood

(2010). Formulaic language and second language speech fluency: Background, evidence and classroom applications. Bloomsbury.

63.

Wray

(2008). Formulaic language: Pushing the boundaries. Oxford University Press.

64.

Y.A.

(2001). English language teaching in China: Trends and challenges. TESOL Quarterly, 35, 191–194.

65.

You

C.J.

Dörnyei

(2016). Language learning motivation in China: Results of a large-scale stratified survey. Applied Linguistics, 37, 495–519.

66.

Yuan

Ellis

(2003). The effects of pre-task planning and on-line planning on fluency, complexity and accuracy in L2 monologic oral production. Applied Linguistics, 24, 1–27.

67.

Zuur

A.F.

Ieno

E.N.

Elphick

C.S.

(2010). A protocol for data exploration to avoid common statistical problems. Methods in Ecology and Evolution, 1, 3–14.