Title Page

Title of manuscript: The Roles of Multimodal Pedagogic Effects and Classroom Environment in Willingness to Communicate in English

Acknowledgement: This work was supported by the Ministry of Education of China (MOE) Project of Humanities and Social Sciences [13YJA740041].

Author: Jian-E Peng

Affiliation: College of Liberal Arts, Shantou University, China

Correspondence address:

Department of Foreign Languages and Literature College of Liberal Arts, Shantou University

243 Da Xue Road, Shantou City Guangdong Province, China Post code: 515063

Office Number: +86 754 8650 4008

Mobile number: +86 1380 2712 719

Email address: pengjiane@stu.edu.cn

The Roles of Multimodal Pedagogic Effects and Classroom Environment in Willingness to Communicate in English

ABSTRACT

Multimodality has become a prominent concept in communication and language education research, and pedagogic discourse in second language (L2) classrooms is fundamentally multimodal. While research on willingness to communicate (WTC) has been thriving, little is known about how L2 WTC is related to multimodal classroom pedagogies. This article presents findings from the first large-scale survey study of its kind on EFL students’ perceptions of multimodal pedagogies, and the interrelationships between multimodal pedagogic effects, classroom environment, and WTC in English. Data were collected from 2058 Chinese EFL university students and analyzed using frequency analysis and structural equation modeling. The results showed that the use of audio/video and teachers’ voices/facial expressions were perceived by the participants as the most satisfactory, whereas the visual design of PowerPoint slides was the least satisfactory. Effective use of audio/video significantly predicted classroom environment and WTC, while teachers’ voices/facial expressions contributed to classroom environment, and teachers’ gestures and spatial positions predicted WTC. In addition, classroom environment was the strongest predictor of WTC. The findings have immediate implications for L2 teaching and will enable L2 teachers to exploit the potential of multimodal pedagogies to promote students’ WTC and to benefit their learning.

Keywords: Willingness to communicate; multimodal pedagogic effects; classroom environment; multimodality; structural equation modeling

1. Introduction

The prevalence of English as a world language has compelled its learners to enhance their communicative competence, a task particularly pertinent to those who learn English as a foreign language (EFL) and lack sufficient authentic linguistic contact. These students mostly learn and use English in their language classrooms. Hence, to maximize learning in class, language teachers often painstakingly expend efforts to promote EFL learners’ willingness to communicate (WTC), but find that many of their students are reluctant to speak in English.

Research into WTC in a second language (L2) has been thriving and has revealed many factors that hinder or promote L2 learners’ WTC, such as anxiety, motivation, and classroom environment, to mention but a few. However, many studies focusing on L2 classrooms have been preoccupied with linking contextual factors such as topic, teacher, and learning task with students’ WTC, without giving sufficient attention to the multimodal nature of

classroom communication. Human communication is fundamentally multimodal (O’Halloran, 2011), involving not only language but also other modes, such as gesture, gaze, and facial expression. While relevant concepts such as teachers’ nonverbal behaviors (NVBs) have been considered in several studies of WTC or willingness to talk (Hsu, Watson, Lin, & Ho, 2007;

Menzel & Carrell, 1999; Wen & Clément, 2003), little is known about how WTC is associated with other semiotic resources such as images, movies, or television (TV) series, which are often woven into modern language classrooms. These various modalities that are inherent in classroom communication can influence students’ meaning making of events, and the pedagogic effects of these modalities may impact on the classroom environment, which can influence students’ WTC (Author, 2010).

The relationship between WTC and multimodal pedagogies is relatively uncharted, but it may have significant implications for language teachers. This study extends qualitative findings regarding the association of multimodal pedagogies and classroom WTC (Author, 2017) and sets out to test the roles of multimodal pedagogic effects and classroom environment in WTC in the Chinese EFL context.

2. Multimodality and multimodal classroom pedagogies

Multimodality has become a prominent concept in communication and language education research (Canagarajah, 2018; Kress, 2010; Rowsell & Collier, 2017), which reflects that human communication is an amalgamation of various modes, including not only language but also other modes such as gesture and facial expression. The increasing attention given to multimodality is manifested in the 2018 special issues of Language Learning and System, devoted respectively to exploring multimodality from cognitive perspectives (e.g.

Basirat, Brunellière, & Hartsuiker, 2018; Biau, Fromont, & Soto-Faraco, 2018) and with a focus on teaching and learning in higher education (e.g. Bonsignori, 2018; Morell, 2018).

An important line of multimodal studies draws on Halliday’s (1978) systemic functional linguistics, which views language as social semiotic and emphasizes that meaning arises from social interaction. This social semiotic view of language was then applied to other modes of communication, which gave rise to the broad domain of systemic functional multimodal discourse analysis (SF-MDA) (Kress, 2010; O’Halloran, 2011). SF-MDA has been widely deployed in research areas such as visual images (Kress & van Leeuven, 2006) and classroom pedagogy (Hood, 2011; Lim, O’Halloran, & Podlasov, 2012).

In the field of SF-MDA, O’Halloran (2005) differentiated between semiotic resources and modes, with the former referring to instances such as “l(fā)anguage, visual images, and

mathematical symbolism” and the latter meaning “the channel (e.g., auditory, visual or tactile) through which semiotic activity takes place” (p. 20). However, these two notions are not treated as distinct by Kress (2010), who defined mode as “a socially and culturally given semiotic resource for making meaning” (p. 79), and stated that image, writing, layout, gesture, spatial position, and facial expression are all examples of modes. Multimodality is often used as an overarching term that includes both multisemiotic and multimodal resources (Knox, 2009). This inclusive use of multimodality is adopted in the present study to examine EFL classroom pedagogies that necessarily involve multimodal texts (e.g., PowerPoint slides integrating fonts, colors, images, and audio/video) and multimodal pedagogic discourse (e.g., teachers’ gestures, facial expressions, and spatial position).

Multimodal texts in university EFL classrooms are typically written texts, music, audio/video clips, and images (with colors), which are often assembled in PowerPoint (PPT) slides (Levasseur & Sawyer, 2006; Li & Walsh, 2011). These modes enhance visual and aural sensory stimulation, which are often associated with heightened arousal (Levasseur & Sawyer, 2006) and thereby can influence students’ emotions and motivation. Students have reported that computer-generated slides could “enliven the classroom environment” (Davis, 1998, p. 9) and described classes with these slides as “more enjoyable, more interesting, and more exciting” (Levasseur & Sawyer 2006, p. 116).

However, the design of PPT slides necessitates an optimal combination of fonts, colors, layout, images, and clip art. Abundant neuroimaging evidence has shown that written text and

spoken words involve overlapping neural structures, and thus large passages of text in PPT slides could impair the audience’s comprehension and retention of orally presented information (see Horvath, 2014). Besides, while images and spoken words trigger different neural regions and these two modalities do not normally conflict (Horvath, 2014), irrelevant visual slides may divert students’ attention, since visual information is more readily processed than verbal messages (Levasseur & Sawyer, 2006; Mackiewicz, 2008). Poorly designed slides with small font, lengthy texts, superfluous images, and poor layout or color combinations have been reported to generate side effects such as distraction, boredom, and impaired learning among students (Levasseur & Sawyer, 2006; Schmaltz & Enstr?m, 2014).

While the use of PPT slides was not found to result in improved learning in general (see Levasseur & Sawyer, 2006), its positive effects on foreign language learning specifically have been more frequently reported. Researchers have found that teacher-facilitated PPT presentations together with pedagogical tasks contributed to the learning of English conditional constructions (Jacobsen, 2018), and visual images embedded in slides had a superior effect on the learning of Chinese abstract words among beginner learners of Chinese as a foreign language (Shen, 2010). Furthermore, corpus-based analyses have reported high

coverage of frequent word families in internet television or American and British movies (Lin, 2014; Webb & Rodgers, 2009). Exposure to English TV broadcasts or TV series with

subtitles has been found to facilitate the accruing of grammatical rules and vocabulary (Bylund & Athanasopoulos, 2015; Frumuselu, De Maeyer, Donche, & Plana, 2015) and trigger L2 WTC (MacIntyre, Burns, & Jessome, 2011). Bonsignori (2018) presented a detailed analysis of multi-semiotic modes in clips from films and TV series and the benefits

of using these clips in teaching English for specific purposes (ESP). González-Lloret and Ortega (2014) strongly proposed that visual and aural input from the internet should be incorporated into learning tasks in L2 classrooms.

Multimodal pedagogic discourse involves the modes of gesture, gaze, voice, facial expression, and spatial position (Hood, 2011; Lim et al., 2012), which are compatible with L2 teachers’ NVBs examined previously (Kamiya, 2018; Majlesi, 2018; Sime, 2006). Sime (2006) reported that EFL students attributed three types of functions to teachers’ pedagogic gestures: cognitive (e.g., enhancing comprehension and orienting attention), emotional (e.g., conveying enthusiasm and encouragement), and organizational (e.g., allocating speech turns). Recent evidence from event-related potentials have also shown that audiovisual speech cues in terms of speakers’ articulatory gestures influence the early stages of lexical processing (Basirat et al., 2018), and speakers’ beat gestures influence the syntactic parsing of ambiguous sentences in Spanish (Biau et al., 2018).

Among NVBs, teachers’ nonverbal immediacy behaviors are particularly relevant to students’ WTC. Nonverbal immediacy behaviors refer to “nonspoken actions which are approach behaviors, signals of availability for communication, typically multichanneled, and communications of interpersonal closeness and warmth” (Menzel & Carrell, 1999, p. 32).

Instances of these behaviors are “positive head nods, smiles, eye contact, vocal expressiveness and close physical distance” (Wen & Clément, 2003, p. 28). In Sime’s (2006) study, students construed positive emotions from teachers’ “relaxed body posture, abundance of gestures and eye contact and positive facial expressions” (p. 221). Other studies have also

shown that teachers’ nonverbal immediacy behaviors were positively correlated with students’

motivation and positive affect (Christensen & Menzel, 1998; Christophel, 1990; Comstock, Rowell, & Bowers, 1995) and willingness to talk in English class (Gol, Zand-Moghadam, & Karrabi, 2014; Hsu et al., 2007). Lim et al. (2012) demonstrated that teachers could construct different relationships with students by deploying physical space, such as being authoritative (i.e., standing in front of the teacher’s desk or in the front of the classroom) or interactional (i.e., standing among students). Two scales measuring nonverbal immediacy behaviors are often used. Richmond, McCroskey, and John (2003) developed the 26-item Nonverbal Immediacy Scale (NIS) that describes people’s behaviors in conversations (e.g., “I gesture when I talk to people”), and high reliability estimates (Cronbach’s α above .90) were reported. Zhang and Oetzel (2006) constructed a scale of 15 items eliciting students’ perceptions of Chinese teachers’ immediacy behaviors (e.g., “[The teacher] is committed to teaching”). High reliability estimates for this scale (Cronbach’s α above .80) were also reported.

3. L2 WTC and classroom environment

The concept of WTC originated in first language (L1) communication research, which refers to individuals’ stable predisposition towards talking (McCroskey & Baer, 1985). This concept was soon introduced to the L2 domain given the importance of “talking in order to learn” (MacIntyre & Legatto, 2011, p. 152). The development of L2 communicative competence requires learners to be willing and inclined to talk at the outset. L2 WTC refers to learners’ intention to enter into L2 discourse in a specific situation (MacIntyre, D?rnyei, Clément, & Noels, 1998). MacIntyre et al. (1998) proposed a seminal model that integrates various enduring factors (e.g., personality) and transient factors (e.g., state communicative

self-confidence), which jointly influence L2 WTC that in turn predicts L2 use. Underpinned by this theoretical model, L2 WTC research has been gaining momentum in many EFL contexts, such as Japan (Fushino, 2010; Yashima, 2002), China (Author, 2010; Yu, 2015), Iran (Ghonsooly, Khajavy, & Asadpour, 2012; Khajavy, Ghonsooly, Fatemi, & Choi, 2016), and Poland (Mystkowska-Wiertelak & Pawlak, 2017).

Early research in L2 WTC has mainly used quantitative methods to test the roles of a number of individual factors in predicting L2 WTC. This strand of research has contributed fruitful insights into how L2 WTC is influenced by learner internal factors such as personality, motivation and attitudes, self-confidence, learner beliefs, and enjoyment (Fushino, 2010; Ghonsooly et al., 2012; Khajavy, MacIntyre, & Barabadi, 2017; Yashima, 2002).

Due to its pedagogic consequences, L2 WTC in educational contexts has received particular research attention. Researchers have employed qualitative methods such as classroom observations, interviews, and journals (e.g., Cao, 2014; Kang, 2005) to identify the influence of contextual factors on students’ WTC. Many factors such as topic, task, interlocutor, teacher, and group size within the classroom context were found to influence L2 WTC (Cao, 2014; Cao & Philp, 2006; Kang, 2005; Yu, 2015). Kang (2005) also identified the joint impact of L2 learners’ psychological conditions (i.e., excitement, responsibility, and security) and situational variables (i.e., topic, interlocutors, and conversational context) on WTC.

Classroom environment is an important contextual antecedent of WTC. The influence of classroom environment or climate on WTC has been statistically tested using structural equation modeling (SEM) (Author, 2010; Khajavy et al., 2016). Author (2010) elaborated on

the role of classroom environment in WTC and operationalized classroom environment in three aspects: teacher support, student cohesiveness, and task orientation. Classroom environment was found to exert both direct and indirect effects on WTC. Khajavy et al.’s (2016) study of 243 English-major university students in Iran confirmed that classroom environment was the strongest predictor of L2 WTC (see also Khajavy et al., 2017). Joe et al. (2017) also reported that classroom social climate indirectly predicted L2 WTC among 381 Korean secondary school students. However, it should be noted that the majority of studies on WTC and/or classroom environment have been premised on the mode of verbal communication, with only a few exceptions explicitly or indirectly addressing L2 WTC in relation to multimodal pedagogies. This is further discussed below.

4. L2 WTC in relation to multimodal pedagogies

As noted earlier, SF-MDA envisages that meaning making in conversation is accomplished not just by verbal means but also by using other semiotic resources. In this light, classroom WTC can be viewed as developing from the meaning that students dynamically attribute to multimodal classroom communication. Author (2017) adopted an SF-MDA approach and identified nuanced differences in the semantic features of the teacher’s verbiage, gestures, and gaze between two teaching scenarios rated as high and low WTC, respectively. This study, which was the original effort in this line of inquiry, has revealed associations between multimodal pedagogies and classroom WTC.

Several other studies on WTC, although not framed within a multimodal perspective, have also touched upon multimodal pedagogies. For instance, Hsu et al. (2007) found that

teacher immediacy behaviors of touching, relaxed body position, gaze, and gestures significantly predicted 235 Taiwanese students’ willingness to speak in English. In Buckingham and Alpaslan’s (2017) experimental research, significant improvement in speaking proficiency and WTC in English was observed in two classes of Turkish students who accomplished out-of-class speaking activities through recording their answers to questions embedded in PPT slides from teachers, which contained “video and voice recordings of the teacher, images and text” (p. 31). Munezane’s (2015) study of 373 Japanese university EFL students used a movie clip from Harry Potter to instruct students on the importance of visualization. She found that students who received instruction in visualization and goal-setting exhibited a significant increase in their WTC. Despite these sporadic studies, however, the relationship between multimodal pedagogies and L2 WTC is still a nascent yet significant research area that warrants more compelling empirical evidence.

The study reported in this paper is the first large-scale survey of its kind that investigates the roles of multimodal pedagogic effects and classroom environment in WTC. This survey approach is argued to be a timely contribution to the state of the art of research in multimodality and classroom WTC. While many studies have offered fine-grained analyses on how multimodality is lived out in classroom teaching (e.g., Author, 2017; Morell, 2018) or executed in cognitive mechanisms of L2 learning (e.g., Basirat et al., 2018; Biau et al., 2018), those studies have adopted micro perspectives or “the proverbial magnifying glass” (Mystkowska-Wiertelak & Pawlak, 2017, p. 76) to reveal uniqueness instead of more general trends. Invaluable as the micro or situated perspectives are, methodologies adopting a macro perspective, such as surveys involving large samples and inferential statistics, have their

merits in modeling patterned and generalizable relationships between WTC and its influential factors in question (Mystkowska-Wiertelak & Pawlak, 2017). More importantly, while the multimodal nature of L2 teaching has been widely recognized, it is the students’ interpretation of the pedagogies that ultimately matters. Sime (2008) argued for credit to be given to addressees’ interpretations in gestural studies instead of meanings being imposed on speakers’ gestures. Holsanova (2012) also advocated a “reception perspective” (p. 252) to reveal how recipients process multimodal information. Their ideas are taken up in the present survey study that focuses on students’ perceptions of multimodal pedagogic effects.

In addition, since teaching and learning in L2 classrooms, as in any educational context, necessarily “happens through speech, writing, gesture, image and space” (Archer, 2014, p.

189), this study explores both multimodal pedagogical texts (i.e., audio/video and the visual design of PPT slides) and pedagogic discourse (i.e., teachers’ NVBs), which are typically found in EFL classrooms. It should be noted that it was the visual design of PPT slides rather than their content that was examined in this study. This was because the content of teaching slides could greatly vary across English courses with different syllabus goals (e.g., targeting different macro skills such as speaking, reading, or writing), the consideration of which could complicate the intended exploration of the visual modality.

5. The present study

The present study investigates university students’ perceptions of multimodal pedagogies in EFL classrooms, and the relationships between multimodal pedagogic effects, classroom environment, and WTC. It addresses the following research questions:

1.

How do the participants perceive the multimodal pedagogies in their English language classrooms?

2. What are the roles of multimodal pedagogic effects and classroom environment in WTC in English?

The first research question was answered with descriptive and frequency statistics, and SEM was employed to address the second question. The measurement instruments were validated through exploratory factor analyses (EFAs) in a pilot study and confirmatory factor analyses (CFAs) in the main study. Based on these analytic results, multimodal pedagogic effects were identified to encompass five dimensions: audio/video, visual design of PPT slides, voice/facial expressions, gestures, and spatial position (presented below). The first two dimensions were mainly related to multimodal pedagogical texts, and the latter three to multimodal pedagogic discourse.

Since the roles of different types of multimodality in classroom environment and WTC are largely unknown, the five dimensions of multimodal pedagogic effects were treated as five exogenous variables, while classroom environment and WTC in English were the two endogenous variables with their sub-dimensions serving as indicator variables. Figure 1 shows the hypothesized model. Model specification was based on relevant theories and previous research (Hair, Black, Babin, & Anderson, 2010). First, multimodal pedagogic effects are hypothesized to predict classroom environment, since teachers’ nonverbal immediacy behaviors (i.e., part of their multimodal pedagogies) were proposed to engender

“interpersonal closeness and warmth” (Menzel & Carrell, 1999, p. 32), which could contribute to a supportive classroom environment. The use of audio, video, and multimodal slides was also found to stimulate student interest and engagement (Levasseur & Sawyer, 2006) and enliven the classroom environment (Davis, 1998). Second, multimodal pedagogic effects are expected to predict WTC based on extant findings that WTC was responsive to teachers’ nonverbal immediacy behaviors (Hsu et al., 2007; Wen & Clément, 2003) and the watching of movies or TV programs (MacIntyre et al., 2011). In addition, a path from classroom environment to WTC in English was added based on existing findings (Author, 2010; Khajavy et al., 2016). Since the present study represents the first attempt to test the relationships between these variables, the current analysis was exploratory in nature, and post hoc analysis (Byrne, 2016) was necessary to scrutinize the results.

[Figure 1 near here]

5.1. Participants

Stratified sampling was used to recruit participants. Initially the survey was conducted with 2350 participants recruited from 14 universities distributed across six regions of China: northern, northeastern, eastern, south-central, southeastern, and northwestern. In principle, two universities were chosen from each region, one being a key national university and the other an ordinary university. However, since the eastern and south-central regions have a larger number of universities, three universities were chosen from each of these areas. The prospective participants were non-English-major freshmen and sophomores. Juniors and

seniors were not recruited because non-English majors of these two grades are generally not required to take college English as a mandatory course. To avoid sampling errors, English majors were not included due to the heterogeneity in the courses taken and the expected learning outcomes for the two groups of students.

During the initial data exploration, listwise deletion was used to remove multivariate outliers identified by Mahalanobis distance (p < .001) and cases with missing values and with apparent traces of cavalier responses (e.g., the written responses formed certain patterns or figures). This resulted in 2058 valid cases, from participants whose ages ranged between 17 and 22 years (M=19.01, SD=0.93). Table 1 details the demographic information about the participants.

[Table 1 near here]

5.2. Instruments

The questionnaire comprised two parts. The first part requested the participants’ demographic information (i.e., age, gender, grade, name of institution, and disciplinary major) but not their names. The second part contained scale items capturing multimodal pedagogic effects, classroom environment, and WTC. The item pools for the three scales were first identified from the literature and the face validity of the scales was enhanced by inviting a panel comprising two experts in SF-MDA and three academics specializing in L2 teaching to review the items (see Neuman, 2014). Only the items agreed by the panel to have measured their designated variables were retained. Translation and back-translation procedures were

undertaken to produce bilingual versions of the items, and the accuracy of the translation was confirmed by a renowned professor in translation. The final Chinese scales were used in the survey. The items for multimodal pedagogic effects are shown in Table 4, and the items for classroom environment and WTC in English are shown in Appendix A.

The scales were first piloted with 428 university students who shared the same background as the participants in the main study. EFAs were run with the pilot data to identify the factor structures of the three variables. Assumptions about the factorability of the data were met since Bartlett’s test of sphericity was significant at the .05 level, and the

Kaiser-Meyer-Olkin value was larger than .06 (Field, 2009). Principal axis factoring was used as the factor extraction method and an oblimin rotation was used for factor rotation. Items with low or split loadings or low communalities were removed progressively for data reduction. The factor structures were used for specifying the measurement models in the main study using CFAs (described below).

Multimodal Pedagogic Effects: Since there was no established scale measuring multimodal pedagogic effects, the pilot study commenced with 28 items in an item pool generated from extant discussions on the design of PPT slides and their effects on affective learning (Davis, 1998; Levasseur & Sawyer, 2006) and Richmond et al.’s (2003) NIS. It should be reiterated that since this study focused on the quality use of multimodal elements, the polarity of items was scrutinized and adapted so that they indicated or implied possible effects of the multimodal pedagogies. The items for voice/facial expressions and spatial position described immediacy behaviors such as “smiling when talking to students” and “going to the middle of the class when teaching”, which were deemed to imply positive

effects because these behaviors made the relations between the teacher and students closer. However, the original items related to gestures in Richmond et al.’s (2003) NIS (e.g., “I gesture when I talk to people”) conveyed neutrality, so revisions were made to indicate positive effects. The items were rated on a six-point Likert scale (from 1=strongly disagree to 6=strongly agree).

The results of EFAs yielded a five-factor solution for multimodal pedagogic effects. A close examination of the item content suggested that the five factors reflect five dimensions of multimodal pedagogic effects: audio/video, visual design of PPT slides, voice/facial expressions, gestures, and spatial position. Twenty items were retained after the CFA in the main study.

Classroom Environment: Classroom environment was measured originally with 17 items rated on a six-point Likert scale (from 1=strongly disagree to 6=strongly agree) based on Author’s (2010) study and Zhang and Oetzel’s (2006) items capturing teachers’ instructional immediacy (e.g., “well-prepared in teaching”). Revisions were made (e.g., removing the item “smiles while talking”) to avoid content overlap with the items related to multimodal pedagogic effects. The EFA results suggested a three-factor solution, which corresponded to the three factors identified in Author’s (2010) study: teacher support, student cohesiveness, and task orientation. Fourteen items were used after the CFA in the main study.

WTC in English: Nineteen items from Menzel and Carrell’s (1999) scale of willingness to talk were used as the item pool to measure WTC in English. This instrument was chosen for two reasons: a) existing L2 WTC scales often use items in McCroskey and Baer’s (1985) L1 WTC instrument, which measures many situations that are unlikely to happen in an EFL

classroom; and b) Menzel and Carrell’s (1999) scale portrays typical classroom situations (e.g., “When I know the correct answer”), which have been found suitable for investigating classroom WTC (Yu, 2015) and teachers’ nonverbal immediacy behaviors (Hsu et al., 2007). The items were also rated on a six-point scale (from 1=definitely unwilling to 6=definitely willing).

The results of the EFAs yielded a 3-factor structure for this scale. The content of the items appeared to measure students’ WTC in three situations: when they were well prepared (e.g., “know the correct answer”), sensed a responsibility to speak up (“sitting in the front of the class”), and felt stimulated to speak up (“the topic is interesting”). Therefore, the three dimensions were named communication preparedness, communication responsibility, and contextual stimulation respectively.

5.3. Data collection

Hard copies of the questionnaire printed in a booklet format (see D?rnyei & Taguchi, 2010) were administered to the participants in their regular classes. The students received explanations of the research purposes and were assured of the anonymity and optionality of their participation, which would not affect their course achievement in any way. Those who gave consent to participate were then given the questionnaire. The instructions written on the questionnaire were verbally repeated to remind the participants to respond to the items based on their experiences in one recent and/or their most impressive English class. It took the participants about 15 minutes to complete the questionnaire.

5.4. Data analysis

The data was found to be normally distributed since the values of skewness and kurtosis for the scale items were within the range of –1 to +1 (Schumacker & Lomax, 1996). The Statistical Package for the Social Sciences (SPSS) 22 software was used to input data and process the descriptive and frequency statistics that provided answers to the first research question. To answer the second question, CFAs were conducted to test the factor structures of multimodal pedagogic effects, classroom environment, and WTC as identified in the pilot study, and SEM was used to test their interrelationships using the Analysis of Moment Structures (AMOS) 21 software.

The internal consistency reliability of the instruments was first assessed using Cronbach’s alpha (α); the achieved value, higher than .70, was considered acceptable (Hair et al., 2010). Composite reliability (i.e., construct reliability), which accounts for effects of latent variables and measurement errors, was also computed. Its value above .70 was also considered acceptable (Hair et al., 2010).

Construct validity was assessed by considering a range of fit indexes, which were also used for evaluating the structural model. Since Chi-Square (χ2) or the normed χ2 can substantially inflate with a large sample size, the model fit was assessed by referring to other goodness-of-fit indexes (see also M?ntym?ki & Salo, 2011). Hair et al. (2010) pointed out that “in circumstances with larger samples (greater than 750)” (p. 579) the normed χ2 is not applicable. Given that this study involved an exceptionally large sample, instead of χ2 or normed χ2, other indexes were considered. Generally, GFI (Goodness-of-Fit Index), CFI (Comparative Fit Index), and NFI (Normed Fit Index) above 0.90, RMSEA (Root Mean

Square Error of Approximation) below .08 with 90% confidence interval (CI) between 0.05 and 0.08, and SRMR (Standardized Root Mean Square Residual) below 0.08 indicate that the model is acceptable (Byrne, 2016). The fit indexes for the measurement models are shown in Appendix B. Table 2 shows the dimensions, sample items, and the Cronbach’s α and composite reliability values of each dimension for the three variables in question.

[Table 2 near here]

6. Results

6.1. Perceptions of multimodal pedagogic effects

Descriptive statistics for the dimensions of the three variables and their correlation coefficients are shown in Table 3. The data show that teachers’ use of voice/facial expressions received the highest score (M=4.88), followed by audio/video (M=4.86), gestures (M=4.58), and spatial position (M=4.38). The mean score of visual design was the lowest (M=4.31), suggesting that the participants were least satisfied with the visual design of the PPT slides in their English class.

[Table 3 near here]

Table 4 shows the frequency of responses to the items related to multimodal pedagogic effects. For purposes of clarification, the sum of responses (Σ) for disagreement (scores of 1 to 3) and agreement (scores of 4 to 6) are also shown, respectively.

[Table 4 near here]

Table 4 shows that generally the participants held positive attitudes towards their teachers’ multimodal pedagogies. In particular, more than 90% of them agreed that the audio/video used in class facilitated their understanding of the lesson content (Item M1) and provided useful information (Item M3). The majority of them also embraced the statements regarding the effects of teachers’ voice/facial expressions, agreeing that their teachers smiled (Item M13, 97.1%) or maintained eye contact with them (Item M12, 94.1%).

Comparatively, the items on the visual design of PPT slides and the teacher’s spatial position received lower agreement. Less than 80% of the participants believed that the template, colors, and fonts of the PPT slides (Items M6, M7, and M9) enabled them to stay focused. Similarly, less than 80% of them reported that their teachers stood close to them during conversations (Items M17, M18, and M20).

6.2. Roles of multimodal pedagogic effects and classroom environment in WTC

The initial SEM results indicated that the model provided a reasonable fit to the data, but the regression weights of five hypothesized paths were non-significant: from visual design to classroom environment and WTC, respectively, from voice/facial expressions to WTC, from gestures to classroom environment, and from spatial position to classroom environment. For the sake of model parsimony, and, more importantly, due to the exploratory nature of this study, the non-significant paths were removed progressively. The final model is shown in Figure 2. The fit indexes displayed a reasonable model fit: GFI = 0.92; TLI = 0.90; NFI =0.91;

CFI = 0.92; RMSEA = 0.06, 90% CI [0.06, 0.07]; SRMR = 0.07. The model explained 70%

of the variance in classroom environment and 41% of the variance in WTC in English.

[Figure 2 near here]

As seen in Figure 2, four dimensions of multimodal pedagogic effects were significant predictors of either or both of the two endogenous variables. Voice/facial expressions and audio/video significantly predicted classroom environment, the standardized coefficient of the former (.67) being much higher than that of the latter (.23). Three of the four types of modalities significantly predicted WTC in English, but their standardized coefficients were relatively low, ranging between .10 and .12. Classroom environment was the strongest predictor of WTC (.41).

7. Discussion

This study investigates Chinese EFL university students’ perceptions of multimodal pedagogical texts and multimodal pedagogic discourse in their EFL classrooms. The results indicated that in terms of multimodal pedagogical texts, the effective use of audio/video described in the survey items received higher endorsement than the visual design of PPT slides. These results may be interpreted with reference to the EFL classroom context.

Different from teaching other subjects, teaching a foreign language is supposed to provide rich linguistic input in class, which can be realized by taking audio/video materials from sources such as English news broadcasts, films, and TV series. These multimodal resources

complement traditional verbal teaching and can facilitate learning (Frumuselu et al., 2015; Webb & Rodgers, 2009), which may account for the positive appraisal attached to these resources. In contrast, the finding that the visual design of PPT slides was less satisfactory seemed consistent with previous results (Levasseur & Sawyer, 2006). It is likely that the participants had experienced slides with dense text, small fonts, or disorganized images which, according to neuroscientific evidence, may impair cognitive processing (Horvath, 2014), have a negative effect on interest, and cause distraction or boredom (Schmaltz & Enstr?m, 2014).

The relatively low score of spatial position suggested that the participants tended not to perceive interpersonal closeness from the pedagogic mode of space. This might be attributed to the large classes in China, where college classes of over 100 students are common due to the rapid expansion in college enrollment (Wang & Zhang, 2011). It is conceivable that large classes may constrain the teacher from interacting with students. That said, since teachers are generally accepted as authoritative figures in Chinese culture (Wen & Clément, 2003), the psychological distance from their teachers that students perceive may mask the salience of physical distance during classroom communication.

The hypothesized effects of audio/video resources on classroom environment and WTC were confirmed. While no prior studies have statistically tested the relationships between these variables, the identified effects are understandable. As suggested before, audio and video resources such as English broadcasts or movies can illustrate authentic, contextualized use of the target language, which simultaneously enhances students’ sensory stimulation.

Compared to teachers’ verbal teaching and written textbooks, these audiovisual resources

enable students to ‘see’ the language used in actual contexts (Bonsignori, 2018) and, if effectively used, they can sustain students’ attention and interest, enliven the classroom atmosphere (Davis, 1998; Levasseur & Sawyer, 2006), and thereby promote their willingness to speak in English.

As for multimodal pedagogic discourse, teachers’ voice and facial expressions were found to exert a significantly direct effect on classroom environment, but not on WTC. The items measuring these modalities were related to teacher immediacy behaviors such as teachers’ vocal variety, eye contact, and smiles (Christophel, 1990). This result indicated that teachers can employ multiple modes of voice and facial expressions to communicate “interpersonal closeness and warmth” (Menzel & Carrell, 1999, p. 32), which convey teacher support and contribute to the students’ ability to make sense of the classroom environment.

The non-significant direct relationship between voice/facial expressions and WTC was unexpected, but it corresponded with previous findings. While positive correlations between nonverbal immediacy behaviors and students’ state motivation have been reported (Christensen & Menzel, 1998; Christophel, 1990), Menzel and Carrell (1999) found that students’ willingness to talk did not vary across different levels of teacher nonverbal immediacy. Comstock et al. (1995) even posited that moderate levels of teacher nonverbal immediacy are more effective in facilitating student learning than excessively high or low immediacy, because high levels of teacher nonverbal immediacy may evoke high arousal, which in turn causes avoidance moves. In light of this argument, the teacher’s voice/facial expressions, such as eye contact and smiles, may send encouraging messages, but overuse of them may impose stress on a student—for instance, if he or she is struggling with a difficult

topic under discussion. On the other hand, teacher eye contact may equally increase another student’s confidence to speak up, particularly if that student is awaiting an encouraging message from the teacher. Put another way, the effects of teachers’ voice/facial expressions on students’ WTC may not be direct, but instead may occur through individual students’ momentary affects, such as self-perceived competence or anxiety (MacIntyre et al., 1998).

The results also show that teachers’ gestures that were perceived to attract attention or stimulate interest positively predicted WTC, which aligns with Hsu et al.’s (2007) finding. These collective findings suggest that teachers’ deliberate or involuntary immediacy behaviors may function as tokens of encouragement that generate students’ motivation and confidence (Sime, 2008) and thus their willingness to talk in English. Hood’s (2011) analysis of pedagogic gestures suggested that a palms-up gesture used by a teacher opens up space for students’ voices. In addition, teachers’ spatial position close to students was found to exert a positive effect on WTC, which supports the important role of space in classroom pedagogy (Lim et al., 2012). In the current case, the teacher’s close physical proximity may oblige students to respond to his/her prompts, which places students in situations where they are willing to communicate because of a sense of responsibility to do so. In other words, teachers can employ ‘spatial pedagogy’ to construct an interpersonal relationship with their students (Lim et al., 2012). The non-significant effects of gestures and spatial position on classroom environment in the present study may be because classroom environment was conceptualized to encompass teacher support, task orientation, and student cohesiveness, and the latter two dimensions are unlikely to be influenced by teachers’ fleeting nonverbal behaviors.

The present result also confirms the significant effect of classroom environment on

WTC reported in previous research (Author, 2010; Khajavy et al., 2016; Khajavy et al., 2017), although WTC here was measured using items from Menzel and Carrell’s (1999) study. More importantly, classroom environment was the strongest predictor of WTC (Author, 2010; Khajavy et al., 2016). This result corroborated the established role of classroom environment in WTC. At this juncture, it should be noted that classroom environment is co-constructed by the teacher and students during their verbal and nonverbal interactions. Therefore, although multimodal resources were the foci of this study, the essential role of verbal communication between all parties in building a classroom environment conducive to WTC should not be overlooked.

The current findings may have immediate pedagogic implications. The positive roles of audio/video in both classroom environment and WTC in English imply that language teachers can exploit these modalities to diversify the input in their language class. For instance, the teacher may utilize English songs, news broadcasts, and clips from movies and TV series in the teaching of vocabulary, syntax, or even communication skills. These audio and visual modalities provide contextualized language input tied to real-life situations, which, as endorsed by the participants in this study, can create a rich linguistic context. Various tasks can be designed and performed based on the audio/video materials, such as written discourse completion, summaries of news broadcasts, or roleplaying communication encounters before or after the audio/video materials are played to students (see also González-Lloret & Ortega, 2014).

Given the positive roles of gestures and spatial position in WTC in English, teachers can optimize multimodal pedagogic discourse by deploying head nods, eye contact, and palms-up

gestures, and manipulating their spatial position to bond with students so as to stimulate their willingness to speak in English. The present findings can enforce teachers’ understanding of the effects of these modalities on students’ communication psychology, which is important since these modalities operate in momentary and fleeting manners and may often be neglected by teachers. Language teachers of university students in particular need to be ‘literate in gestures’ at least in order not to put a dampener on students’ communication intention, since adult learners are more able to interpret nonverbal messages than younger students (Kamiya, 2018). In EFL contexts where the teacher is taken as the authority and students’ unwillingness to communicate is significant, langauge teachers, as suggested by the current findings, can exploit gestures in tandem with verbal message to promote students’ WTC (Wen & Clément, 2003).

In addition, although the modalities of voice/facial expressions were not found to directly predict WTC, they can be used to convey teacher support and thereby build a wholesome classroom environment. For instance, teachers need to avoid using a monotone or dull voice (Richmond et al., 2003), which may express an attitude of indifference. In contrast, they can use a variety of tone to express their care about and attention to students. It is worthwhile for teacher education programs to include course modules on multimodal pedagogies, to promote the skills of pre-service and in-service teachers in orchestrating multiple modalities to enhance pedagogic effectiveness.

Despite the novel findings of this study, caution must be taken in interpreting the results.

First, although the five dimensions were measured using separate items, there was no intention to claim that these modalities were independent from each other. On the contrary,

these modalities are often interrelated—for instance, gestures and voice/facial expressions mostly happen simultaneously. The interrelationships between these dimensions are reflected in the covariances specified between them (see Figure 2). However, a detailed analysis of intersemiotic complementarity cannot be obtained from such a large-scale quantitative design, instead mandating other approaches such as observation and retrospection. In addition, the five dimensions of multimodal pedagogic effects identified in this study are by no means exhaustive or conclusive. Further research is needed to explore other types of multimodal pedagogies, especially the use of media platforms (e.g., Padlet or Moodle) or social media (e.g., Facebook or WeChat). These new media forms, which have mushroomed and been ushered into cutting edge classrooms, provide rich semiotic resources and may play a role in promoting classroom communication. Finally, the measurement of multimodal pedagogic effects was based on items taken from relevant research, which should only serve as a starting point for operationalizing this construct. There is a clear need for further refinement based on more empirical evidence.

8. Conclusion

Using a large-scale stratified survey, this study has investigated Chinese university students’ perceptions of multimodal pedagogical texts and multimodal pedagogic discourse in their EFL classrooms, and the roles of multimodal pedagogic effects and classroom environment in WTC. It has found that audio/video and teachers’ voice/facial expressions were perceived as the most satisfactory, whereas the visual design of PPT slides was considered the least satisfactory. SEM analysis showed that while the effective use of

audio/video significantly predicted classroom environment and WTC, teachers’ voice/facial expressions also directly contributed to classroom environment, and teachers’ gestures and spatial positions directly predicted WTC. In addition, classroom environment was the strongest predictor of WTC.

Multimodal pedagogies have been attracting increasing research attention, along with the recent recognition of diverse possibilities for meaning making in communication through multimodal resources. Such multimodal perspectives on communication and L2 teaching can shed new light on how L2 learners attribute meanings to multimodal affordances in class.

Research in this area can inspire L2 teachers to exploit the potential of multimodal pedagogies to promote students’ WTC, classroom participation, and learning attainment. This study is a preliminary attempt to go beyond anecdotal reports and statistically test the relationships between students’ meaning making of multimodal pedagogies and their perceptions of classroom environment and L2 WTC. The significance of this study also lies in its initiative to broaden L2 WTC research by keeping abreast with the recent research agenda in applied linguistics, which features multimodality and translingual practice (Canagarajah, 2018).

References

Archer, A. (2014). Power, social justice and multimodal pedagogies. In C. Jewitt (Ed.), The routledge handbook of multimodal analysis (2nd ed.). London: Routledge.

Author, 2010.

Author, 2017.

Basirat, A., Brunellière, A., & Hartsuiker, R. (2018). The role of audiovisual speech in the early stages of lexical processing as revealed by the ERP word repetition effect.

Language Learning, 68(S1), 80–101.

Biau, E., Fromont, L. A., & Soto-Faraco, S. (2018). Beat gestures and syntactic parsing: An ERP study. Language Learning, 68(S1), 102–126.

Bonsignori, V. (2018). Using films and TV series for ESP teaching: A multimodal perspective.

System, 77, 58–69.

Buckingham, L., & Alpaslan, R. S. (2017). Promoting speaking proficiency and willingness to communicate in Turkish young learners of English through asynchronous

computer-mediated practice. System, 65, 25–37.

Bylund, E., & Athanasopoulos, P. (2015). Televised Whorf: Cognitive restructuring in advanced foreign language learners as a function of audiovisual media exposure. The Modern Language Journal, 99(S1), 123–137.

Byrne, B. M. (2016). Structural equation modeling with AMOS: Basic concepts, applications, and programming (3rd ed.). New York: Routledge.

Canagarajah, S. (2018). Translingual practice as spatial repertoires: Expanding the paradigm beyond structuralist orientations. Applied Linguistics, 39(1), 31–54.

Cao, Y. (2014). A sociocognitive perspective on second language classroom willingness to communicate. TESOL Quarterly, 48(4), 789–814.

Cao, Y., & Philp, J. (2006). Interactional context and willingness to communicate: A comparison of behavior in whole class, group and dyadic interaction. System, 34(4), 480– 493.

Christensen, L. J., & Menzel, K. E. (1998). The linear relationship between student reports of teacher immediacy behaviors and perceptions of state motivation, and of cognitive, affective, and behavioral learning. Communication Education, 47(1), 82–90.

Christophel, D. M. (1990). The relationships among teacher immediacy behaviors, student motivation, and learning. Communication Education, 39(4), 323–340.

Comstock, J., Rowell, E., & Bowers, J. W. (1995). Food for thought: Teacher nonverbal immediacy, student learning, and curvilinearity. Communication Education, 44(3), 251– 266.

Davis, M. S. (1998). The electronic biology classroom: Implementation and student opinion.

Kapiolani Community College. Paper contributed to the Teaching in the Community Colleges Online Conference. (ERIC Document Reproduction Service No. ED449976).

D?rnyei, Z., & Taguchi, T. (2010). Questionnaires in second language research: Construction, administration, and processing (2nd ed.). Mahwah, NJ: Lawrence Erlbaum Associates.

Frumuselu, A. D., De Maeyer, S., Donche, V., & Colon Plana, M. d. M. G. (2015). Television series inside the EFL classroom: Bridging the gap between teaching and learning informal language through subtitles. Linguistics and Education, 32, 107–117.

Fushino, K. (2010). Causal relationships between communication confidence, beliefs about group work, and willingness to communicate in foreign language group work. TESOL Quarterly, 44(4), 700–724.

Ghonsooly, B., Khajavy, G. H., & Asadpour, S. F. (2012). Willingness to communicate in English among Iranian non-English major university students. Journal of Language and

Social Psychology, 31(2), 197–211.

Gol, M., Zand-Moghadam, A., & Karrabi, M. (2014). The construct of willingness to communicate and its relationship with EFL learners' perceived verbal and nonverbal teacher immediacy. Issues in Language Teaching, 3(1), 135–160.

González-Lloret, M., & Ortega, L. (Eds.). (2014). Technology-mediated TBLT: Researching technology and tasks. Amsterdam: John Benjamins.

Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2010). Multivariate data analysis

(7th ed.). Upper Saddle River, NJ: Prentice Hall.

Halliday, M. A. K. (1978). Language as social semiotic: The social interpretation of language and meaning. London: Edward Arnold.

Holsanova, J. (2012). New methods for studying visual communication and multimodal integration. Visual Communication, 11(3), 251–257.

Hood, S. (2011). Body language in face-to-face teaching: A focus on textual and interpersonal meaning. In S. Dreyfus, S. Hood & M. Stenglin (Eds.), Semiotic margins: Meaning in multimodalities (pp. 31–52). London: Continuum.

Horvath, J. C. (2014). The neuroscience of PowerPointTM. Mind, Brain, and Education, 8(3), 137–143.

Hsu, L.-I., Watson, T., Lin, C.-H., & Ho, T.-C. (2007). Explorations in teachers’ nonverbal immediacy behaviors and students’ willingness to talk in English. English Teaching & Learning, 31(1), 1–27.

Jacobsen, N. D. (2018). The best of both worlds: Combining cognitive linguistics and pedagogic tasks to teach English conditionals. Applied Linguistics, 39(5), 668–693.

Joe, H.-K., Hiver, P. H., & Al-Hoorie, A. H. (2017). Classroom social climate,

self-determined motivation, willingness to communicate, and achievement: A study of structural relationships in instructed second language settings. Learning and Individual Differences, 53, 133–144.

Kamiya, N. (2018). The effect of learner age on the interpretation of the nonverbal behaviors of teachers and other students in identifying questions in the L2 classroom. Language Teaching Research, 22(1), 47–64.

Kang, S.-J. (2005). Dynamic emergence of situational willingness to communicate in a second language. System, 33(2), 277–292.

Khajavy, G. H., Ghonsooly, B., Fatemi, A. H., & Choi, C. W. (2016). Willingness to communicate in English: A microsystem model in the Iranian EFL classroom context. TESOL Quarterly, 50(1), 154–180.

Khajavy, G. H., MacIntyre, P. D., & Barabadi, E. (2017). Role of the emotions and classroom environment in willingness to communicate: Applying doubly latern multilevel analysis in second language acquisition research. Studies in Second Language Acquisition, 40(3), 605–624.

Knox, J. S. (2009). Multimodal discourse on online newspaper home pages: A

social-semiotic perspective (Unpublished doctoral thesis). The University of Sydney, Sydney, Australia.

Kress, G. (2010). Multimodality: A social semiotic approach to contemporary communication.

London: Routledge.

Kress, G., & van Leeuven, T. (2006). Reading images: The grammar of visual design (2nd

ed.). London: Routledge.

Levasseur, D. G., & Sawyer, J. K. (2006). Pedagogy meets PowerPoint: A research review of the effects of computer-generated slides in the classroom. The Review of Communication, 6(1-2), 101–123.

Li, L., & Walsh, S. (2011). Technology uptake in Chinese EFL classes. Language Teaching Research, 15(1), 99–125.

Lim, F. V., O’Halloran, K. L., & Podlasov, A. (2012). Spatial pedagogy: Mapping meanings in the use of classroom space. Cambridge Journal of Education, 42(2), 235–251.

Lin, P. M. S. (2014). Investigating the validity of internet television as a resource for acquiring L2 formulaic sequences. System, 42, 164–176.

M?ntym?ki, M., & Salo, J. (2011). Teenagers in social virtual worlds: Continuous use and purchasing behavior in Habbo Hotel. Computers in Human Behavior, 27(6), 2088–2097.

MacIntyre, P. D., Burns, C., & Jessome, A. (2011). Ambivalence about communicating in a second language: A qualitative study of French immersion students’ willingness to communicate. The Modern Language Journal, 95(1), 81–96.

MacIntyre, P. D., D?rnyei, Z., Clément, R., & Noels, K. A. (1998). Conceptualizing willingness to communicate in a L2: A situational model of L2 confidence and affiliation. The Modern Language Journal, 82(4), 545–562.

MacIntyre, P. D., & Legatto, J. J. (2011). A dynamic system approach to willingness to communicate: Developing an idiodynamic method to capture rapidly changing affect. Applied Linguistics, 32(2), 149–171.

Mackiewicz, J. (2008). Comparing PowerPoint experts’ and university students’ opinions

about PowerPoint presentations. Journal of Technical Writing and Communication, 38(2), 149–165.

McCroskey, J. C., & Baer, J. E. (1985). Willingness to communicate: The construct and its measurement. Paper presented at the annual convention of the Speech Communication Association, Denver, CO.

Menzel, K. E., & Carrell, L. J. (1999). The impact of gender and immediacy on willingness to talk and perceived learning. Communication Education, 48(1), 31–40.

Morell, T. (2018). Multimodal competence and effective interactive lecturing. System, 77, 70–79.

Munezane, Y. (2015). Enhancing willingness to communicate: Relative effects of visualization and goal setting. The Modern Languae Journal, 99(1), 175–191.

Mystkowska-Wiertelak, A., & Pawlak, M. (2017). Willingness to communicate in instructed second language acquisition: Combining a macro- and micro-perspective. Bristol: Multilingual Matters.

Neuman, W. L. (2014). Social research methods: Qualitative and quantitative approaches

(7th ed.). Boston: Pearson.

O’Halloran, K. L. (2005). Mathematical discourse: Language, symbolism and visual images.

London: Continuum.

O’Halloran, K. L. (2011). Multimodal discourse analysis. In K. Hyland & B. Paltridge (Eds.),

Companion to discourse (pp. 120–137). London: Continuum.

Richmond, V. P., McCroskey, J. C., & Johnson, A. D. (2003). Development of the nonverbal immediacy scale (NIS): Measures of self- and other-perceived nonverbal immediacy.

Communication Quarterly, 51(4), 504–517.

Rowsell, J., & Collier, D. R. (2017). Researching multimodality in language and education. In K. A. King, Y.-J. Lai & S. May (Eds.), Research methods in language and education (3rd ed., pp. 311–324). Cham: Springer International Publishing.

Schmaltz, R. M., & Enstr?m, R. (2014). Death to weak PowerPoint: Strategies to create effective visual presentations. Frontiers in Psychology, 5. Retrieved from https://www.frontiersin.org/articles/10.3389/fpsyg.2014.01138/full

Schumacker, R. E., & Lomax, R. G. (1996). A beginner’s guide to structural equation modeling. Mahwah, NJ: Lawrence Erlbaum Associates.

Shen, H. H. (2010). Imagery and verbal coding approaches in Chinese vocabulary instruction.

Language Teaching Research, 14(4), 485–499.

Sime, D. (2006). What do learners make of teachers’ gestures in the language classroom?

International Review of Applied Linguistics in Language Teaching, 44(2), 211–230.

Sime, D. (2008). ‘Because of her gesture, it's very easy to understand’- learners’ perceptions of teachers’ gestures in the foreign language class. In S. G. McCafferty & G. Stam (Eds.), Gesture: Second language acquistion and classroom research (pp. 259–279). New York, NY: Routledge.

Wang, Q., & Zhang, N. (2011). Teaching large classes in China - English as a foreign language. Retrieved from http://www.ntu.edu.vn/Portals/96/Phuong%20phap%20GD/ Teaching%20large%20classes%20(7).pdf

Webb, S., & Rodgers, M. P. H. (2009). The lexical coverage of movies. Applied Linguistics, 30(3), 407–427.

Wen, W. P., & Clément, R. (2003). A Chinese conceptualisation of willingness to communicate in ESL. Language, Culture and Curriculum, 16(1), 18–38.

Yashima, T. (2002). Willingness to communicate in a second language: The Japanese EFL context. The Modern Language Journal, 86(1), 54–66.

Yu, M. (2015). An examination of the dynamic feature of WTC through dyadic group interaction. System, 55, 11–20.

Zhang, Q., & Oetzel, J. G. (2006). Constructing and validating a teacher immediacy scale: A Chinese perspective. Communication Education, 55(2), 218–241.

Appendix A. Questionnaire items Classroom environment

C21. The teacher is committed to teaching. C22. The teacher is well-prepared in teaching. C23. The teacher is passionate about teaching.

C24. The teacher provides a timely response to students’ concerns. C25. My group mates like each other.

C26. I am friendly to members of the class. C27. I work well with other class members. C28. My group mates fit together.

C29. Tasks designed in this class are attractive.

C30. I know what I am trying to accomplish in this classes. C31. Class assignments are clear so everyone knows what to do.

C32. Students know exactly what has to be done in our class.

WTC in English

W33. When I am prepared for class. W34. When I know the correct answer.

W35. When I can really clarify the issue under discussion. W36. When my views differ from my classmates’ views. W37. When no one else is talking.

W38. When I am sitting in the front of the class.

W39. When my views differ from the professor’s views. W40. When the class is engaged in an open discussion. W41 When the topic is interesting.

W42. When my participation is being graded. W43. When the class is engaged in a heated debate.

Appendix B. Fit indexes of the three measurement models

Demographic information about the participants.

Dimensions of the variables and reliability coefficients.

Descriptive statistics and correlation matrix.

**p < .01

? ACCEPTED MANUSCRIPT

42

Table 4

Frequency analysis results of multimodal pedagogic effects (%).

Dimension Item Disagree Agree

? ACCEPTED MANUSCRIPT

38

Fig. 1. Hypothesized model.

Fig. 2. Model of WTC in multimodal EFL classrooms.