Factors associated with improvement in oral fluency when using Video-Synchronous Mediated Communication with native speakers
The Challenge
Technology increasingly offers opportunities for language learners to engage with others. Does video-synchronous mediated communication with native speakers offer an effective way to improve oral fluency? What factors are most strongly associated with increased fluency?
Abstract
Computer-mediated, communication-based foreign language learning programs are showing great promise. Among these, video-synchronous mediated communication seems to offer an effective way to provide speaking practice, although research findings have been inconclusive. Furthermore, among studies that document the effectiveness of
video-mediated communication, it is not clear why some learners improve more than others. This study reports data from three separate learner groups who engaged in video-synchronous mediated communication with native speakers. Specifically, the study investigated its effectiveness on learners’ oral fluency, and explored the impact of enjoyment, target language speaking time, or instructional level on improvement. The data suggest that participation did not necessarily guarantee greater improvement when compared with students in a control group, that instructional level is associated with improvement, that allocated time is associated with a decrease in pausing, and that students’ reasons for enjoying the program, rather than their overall enjoyment, are related to improvement.
Keywords: Japanese as a foreign language (JFL), Post-secondary/higher education, Computer-mediated communication
Introduction
With the advancement of modern technology, computer-mediated communication (CMC) has become more common in the foreign language (FL) classroom (see Lin, 2014 for an overview). While Lin (2014) found that both synchronous and asynchronous FL CMC can be successfully utilized to improve a number of different foreign language skills, from writing to oral fluency, his critical review of recent CMC studies revealed that the largest learner gains were found in voice-based CMC studies. Furthermore, Lin (2014) suggested that the greatest gains in oral proficiency would be found in programs in which learners communicated with other learners, not with native speakers, although this result was based on a meta-analysis that did not actually include data from any studies with pairs of learners and native speakers. This notion is partially supported by Yang and Chang (2008) who failed to find improvement in the oral fluency of participants in video-synchronous mediated communication (VSMC) with native speakers. However, studies such as Kato, Spring and Mori (2016), Canto, Jauregi, and Van Den Bergh (2013), and Yang, Gamble and Tang (2012) contradict these studies, all reporting that VSMC with native speakers led to improvement in participants’ oral skills. Thus, the impact of VSMC with native speakers on the development of oral skills is still unclear.
Even among the studies that reported significant improvement in speaking abilities for participants in VSMC programs (e.g. Kato et al., 2016; Yang et al., 2012), it was
observed that some learners improved more than others, although the researchers were unable to pin-point why. Lin’s (2014) analysis found that a number of factors influenced the effectiveness of CMC programs in general, including task type, interlocutor type, and whether the learners were in a foreign language or second language learning environment1. Specifically, Lin (2014) posited that FL learners were aided by CMC more than second language learners, that opinion-exchange task types were effective in improving writing skills but not speaking skills, that peer-to-peer CMC environments supported learners’ spoken and written language development more than peer-to-native speaker environments, and that lower-level learners showed more improvement than higher-level learners.
However, studies that report on VSMC with native speakers (e.g. Kato et al., 2016; Yang et al., 2012) often find significant improvement at a range of instructional levels, making it unclear how this factor affects VSMC specifically.
Furthermore, while Lin (2014) focused on the types of tasks, partnership, and environments in a wide range of CMC studies, the present study focused solely on VSCM with native speakers and the individual learner characteristics of participants in an attempt to determine what other factors specific to VSCM programs might have an impact on learner outcomes. Specifically, the study drew on a large body of data from a VSCM program that paired L2 Japanese learners with native speakers and investigated the
relationship between improvement in participants’ oral fluency and beginning proficiency level, participant enjoyment and target language speaking time.
1 Foreign language environments are those in which the target language is not present or commonly spoken in the surrounding community, whereas second language environments are those in which the target language is used in the locale of the learner.
Review of the Literature
This study examined three factors that could potentially be associated with improvement in oral fluency in a VSCM program: initial proficiency level, participant enjoyment, and the amount of time devoted to speaking in the target language (engaged time).
Initial proficiency level
It has been suggested that a learner’s initial proficiency level can influence the degree to which VSMC supports participants’ developing speaking abilities. Li, for example, posited that lower-proficiency VSMC participants would exhibit greater gains in speaking (2014), perhaps due to the fact that they have more room to improve and that it takes increasingly greater time to move into each successive proficiency band (i.e. a ceiling effect; Rifkin, 2005). A contrasting hypothesis posits that, since beginning learners will have acquired fewer lexical resources (i.e. vocabulary, grammatical structures, etc.), this could limit their ability to engage and thus improve their speaking skills during VSMC sessions. In addition, since oral fluency is generally measured using the amount (number of words or syllables spoken) or speed (speech rate or articulation rate) of speech, learners who bring more linguistic resources may be able to focus more on improving fluency, meaning beginning learners might exhibit fewer measureable gains in oral fluency.
Participant enjoyment
With respect to FL learning, a number of studies have investigated the connection between enjoyment and improvement in FL ability (e.g. Brantmeier, 2003; 2005; Schultz, 2012). Schultz (2012) and Brantmeier (2003), for example, both looked at the relationship between
enjoyment and a wide range of FL abilities, but neither study found significant correlations. Other studies, however, are more aligned with general research on enjoyment, motivation and learning. For example, Brantmeier (2005) found that enjoyment was correlated with self-assessed overall language ability and written recall scores. More recently, Dewaele and Alfawzan (2018) found that FL enjoyment “was associated with higher self-assessed foreign language ability and scores on a reading comprehension test” (p.39). Similarly, a longitudinal cross-sectional study by Saito, Dewaele, Abe and In’nami (2018) found that “more frequent L2 use with positive emotions directly impacts acquisition” (p.709), including speaking skills. Although enjoyment is likely to lead to increased interest in maintaining the conversation partnership beyond a specific course or program (Kato et al., 2016) and although several studies focusing on VSCM found students’ levels of enjoyment to be important enough to report (e.g. Yang et. al, 2012; Yanguas, 2012), none attempted to determine if enjoyment was correlated with improvement in oral abilities.
Engaged time
A large body of work on student engagement across a range of subject areas and instructional levels has confirmed that there is a correlation between the amount of time spent engaged in educational activities and students’ learning outcomes (e.g. Krause & Coates, 2008; Nauffal, 2011; etc.). As in other disciplines, Ellis (1985) suggested that the most influential factors for improving second language acquisition were the quality and quantity of input. In the context of VSMC, quantity refers to the amount of time participants spend using VSMC (allocated time) and the percentage of that time in which communication was conducted in the target language (engaged time). Although quantity
does not necessarily equate with quality (Taillefer & Munoz-Luna, 2014; Tian & Wang, 2010), VSMC with native speakers can provide learners with opportunities for meaningful interactions that support progress toward higher levels of proficiency (see Kato et al., 2016; Scarcella & Oxford, 1992).
Measuring fluency
There are generally three factors by which oral proficiency are measured: accuracy, complexity and fluency (Lambert & Kormos, 2014). In this study, fluency was used to evaluate the effectiveness of a VSMC program for several reasons. First, fluency has been argued to likely be the most important component of oral proficiency (Bernstein, Cheng & Suzuki, 2010; Kormos, 2006). Second, reliable tools for objective measurement of fluency are readily available and can be utilized across languages (Boeersma & Weenink, 2011). Finally, when measuring oral accuracy and complexity, a number of standards have been created for English (Thai & Boors, 2016), but the same cannot be said for Japanese, especially for Japanese FL learners (Hirotani, Matsumoto & Fukuda, 2012). This study focused on the five metrics of fluency that were suggested by Boeersma and Weenink (2011) and Hirotani, Matsumoto and Fukuda (2012). These factors have been successfully used to judge the fluency of FL learners of Japanese in previous studies (e.g. Hirotani et al., 2012; Matsushita, 2011) and include: number of syllables spoken, number of pauses, speech rate (number of syllables spoken per second), articulation rate (number of syllables per phonation time), and average syllable duration (ASD – speaking time per syllable).
Framed by previous studies that have returned conflicting findings about the effectiveness of VSMC with native speakers, and by a body of work that suggests that learner characteristics might help to explain the lack of conclusive findings, this study analyzed three years of survey data, post-program reflections, and fluency data to answer the following research questions:
(1) To what extent does participation in a VSMC program support FL learners’ developing oral fluency, and
(2) To what extent are learner factors (initial instructional level, enjoyment, and engaged time) associated with improvement in oral fluency?
Method
Participants
American students from X University who were studying Japanese as a foreign language (JFL) were partnered with Japanese students from Y University. This cross-sectional study reports data from three separate JFL (American) cohort groups. To clarify, the individual L2 Japanese learners each year were entirely different; data for students who took part in the program during their beginning level class were not collected again in any subsequent semester.
Following IRB approval and once appropriate consent forms were signed, a total of 79 JFL students participated in some aspect of the study (26 in the first year, 18 in the second, and 35 in the third year). Due to missing post-program surveys and/or speaking posttests, data from 56 JFL learner participants were analyzed. These 56 participants were
between the ages of 19 and 26 and were all American native speakers of English with the exception of two students who had learned Chinese as a first language at the same time as English. They were further divided into two groups: Those who had completed 3 semesters of Japanese were considered to be the “beginning group” while participants who had studied for 5 semesters were designated the “intermediate class” These learners were required to participate in the Skype partner program as part of their out-of-class homework in a Japanese language class. Students understood that participating in the speaking tests, survey and reflection paper would not impact their grade, and they were asked to complete these on a volunteer basis (with no financial compensation) according to university policy during all three years.
Data were also collected each year from a control group. The 29 total participants in the control group (year 1 = 8; year 2 = 9 year 3 = 12) ranged in age from 19 to 25, were American native speakers of English, were taking the same types of Japanese language classes to those in the VSMC group, and had also studied Japanese either for 3 or 5 semesters. Although they also participated in pre- and posttests, they did not participate in the Skype VSMC program. Participation in the pre- and posttests was not required for their course work, and control group participants received financial compensation (a $16 gift card) for completing the tests as the VSMC group according to university policy during all three years.
The Japanese conversation partners in Year 1 were first-year students from the same English as a foreign language (EFL) class. Additional students at other grade levels as well
as individuals who were not currently taking any English classes were allowed to participate in the two subsequent years.
Scheduling and expectations
The VSMC Skype program in which students participated over the three-year period under consideration (2015 to 2017) has been described in Kato et al. (2016). Apart from minor changes each year, described below, the overall scheduling pattern, general requirements, pre- and post- speaking tests, questionnaires and post-program reflection prompts remained the same.
For each of the three years, the Skype partner program took place during the American spring semester (from January to May) for 15 weeks. Students were requested to speak with their Skype partner at least twice per week for at least 30 minutes per session, spending at least 15 minutes speaking English and at least 15 minutes speaking Japanese, resulting in a minimum of 30 15-minute speaking sessions in each language. Students were allowed to schedule sessions as best fit their other commitments and were encouraged to extend the sessions based on their availability and motivation. In 2015, the participants were assigned topics to discuss during one of the two weekly sessions, and during the other session, they were allowed to choose the topic. Because some students complained about running out of conversation topics (Kato et al., 2016), in 2016 students were assigned a conversation topic for one weekly session and were asked to talk about a joint project during the other weekly session. However, since the guidelines were admittedly weak, students generally spoke about topics of personal interest and ignored the project. In 2017, a more specific project -- the construction of a bilingual webpage which students presented
as a final project – was assigned as the discussion topic for one session each week. Students reported having actually collaborated with their partners, thus making the project-based learning element of the 2017 program more successful. Although the topics and discussion tasks evolved over time, the basic premise of using VSMC to communicate with a native speaker of the target language for at least 15 minutes in the target language twice per week remained the same.
Fluency measures: Description, coding and analysis
The oral pre- and post-tests were scheduled at the beginning and end of the semester and were used to evaluate participants’ fluency were drawn randomly from the IELTS oral test2 (See Appendix 1) and were the same each year. Participants were given three questions, one at a time. Participants were allowed one minute to prepare their response once each question was posed. They were allowed to respond orally for up to two minutes to each prompt but were allowed to stop their recording sooner if they had nothing more to say. Non-answers were not allowed.
Responses were recorded and analyzed using Praat (Boersma & Weenink, 2011), a software that automatically calculates the number of syllables spoken, number of pauses, speech rate (number of syllables spoken divided by overall time – given in seconds), articulation rate (the number of syllables spoken divided by phonation time – given in seconds) and the average syllable duration (speaking time divided by number of syllables). Differences between pre- and posttests were checked with dependent t-tests, and significant interaction between participation in the Skype partner program and improvement was
2 The IELTS is an international standardized test of English language proficiency for nonnative English speakers that tests all four skills.
checked by conducting two-way one-repeated measures ANOVA tests using the pre- and posttest data from the program participant group and control group. Cohen’s d was calculated as a measure of effect size. A multiple analysis of variance (MANOVA) was conducted to determine if any of the independent factors (class level, enjoyment, or engaged time) had a significant effect on any of the dependent variables. Pillali’s trace was reported for effect size, and Tukey post-hoc tests were used to determine which factors affected which dependent variables when statistically significant results were found.
Survey: Description, coding and analysis
After completion of the program each year, students completed a survey. In all three years, students completed three Likert-type questions that investigated their enjoyment of the program (1 = not at all – 5 = very much). Students who responded that participation was very enjoyable (5) were categorized as “high” while those who rated their level of enjoyment 1, 2, 3 or 4 were coded as “low.”
Student also reported the amount of time they spent in the target language (TL) (1 = less than 40% of the time, 2 = about 50% of the time, 3 = more than 60% of the time), and about how long each session lasted on average (1 = less than 30 minutes, 2 = 30-45 minutes, 3 = 45-60 minutes, 4 = more than an hour). Participants who reported spending the recommended amount of time talking (30 minutes) on average and using their target language about 50% of the time were considered coded as having “medium” engaged time. Participants who reported spending fewer than 30 minutes on average talking during each session or used less than 50% of the time using their target language were categorized as having “low” engaged time. Participants who reported spending more than the
recommended time on average and/or spending more than 50% of the time using their target language were coded as having “high” engaged time.
Reflections: Description, coding and analysis
In their reflection papers, which were written in English, students were asked to describe their experiences in the Skype partner program in response to a global prompt (My
experience in the Skype partner program). Before analyzing the reflection papers, the most and least improved participants were identified: the most improved participants were those who improved on all six metrics of fluency and were within the top quartile in at least four metrics (n=10); the least improved participants were those in the bottom quartile on four or more metrics (n=10). The opinions of these participants were extrapolated from their reflection papers, based on a conceptually clustered matrix analysis (Miles and Huberman, 1994). This approach clusters comments in order to qualitatively analyze students’ opinions and perspectives by grouping similar opinions and comments together to reveal similarities, patterns and trends among participants who evidence high improvement in fluency and those who did not.
Results
Fluency
The pre- and posttest scores for the JFL participant and control groups are given in Table 1. Data are reported by level (beginner, intermediate), level of enjoyment (low, high), and engaged time (low, medium, high) as described in the previous section for the pre- and posttest scores on all five metrics of fluency in Table 2.
Table 1
Averages (SD) scores for program participant group and control groups’ pre- and posttests
Group (N) Test # of syllables # of pauses Speech Rate Articulation Rate ASD Participant Group (56) Pretest 412.8 (233.2) 60.2 (43.6) 1.92 (0.93) 2.73 (1.07) 0.66 (1.24) Posttest 549.6 (266.1) 80.8 (52.5) 2.11 (0.89) 2.92 (0.92) 0.4 (0.22) Control Group (29) Pretest 541.9 (290.8) 71.3 (47.2) 2.13 (0.73) 3.06 (0.73) 0.35 (0.14) Posttest 626.5 (357.6) 91.4 (53.7) 2.08 (0.94) 3.04 (0.99) 0.44 (0.43) Table 2
Average improvement (SD) for program participants divided by class, enjoyment and immersion Type of Division Class or Level (N) # of syllables # of pauses Speech Rate Articulation Rate ASD Divided by Class Beginner (32) 32.9 (182) 21.8 (57) 0.03 (1.04) -0.03 (1.23) 0 (0.24) Intermediate 275.4 19 0.42 0.47 -0.61
(24) (235.7) (45.8) (0.69) (1.05) (1.83) Divided by Level of Enjoyment Low (31) 161.9 (204.9) 22.5 (55) 0.28 (1.02) 0.27 (1.43) -0.45 (1.64) High (25) 105.64 (274.5) 18.2 (49.1) 0.09 (0.78) 0.08 (0.75) -0.03 (0.17) Divided by Engaged Time Low (9) 35.6 (256.8) 2.4 (43.5) -0.06 (1.13) -0.14 (1.49) -0.51 (1.95) Medium (28) 179.3 (207.5) 41.4 (48) 0.05 (0.7) 0.23 (1.13) -0.32 (1.38) High (19) 122.2 (266.2) -1.5 (51.4) 0.54 (1.04) 0.27 (1.11) -0.05 (0.2)
The pre- and posttest scores for all program participants (experimental and control) showed significant differences in number of syllables spoken (t(55)=4.3, p<0.001, d=0.81) and number of pauses (t(55)=2.96, p<0.001, d=0.56), but not for speech rate (t(55)=1.61,
p=0.11), articulation rate (t(55)=1.18, p=0.24) or ASD (t(55)=1.58, p=0.12). These results
indicate that, although participants spoke more on the post-test, they also tended to pause more, which makes it unclear as to whether they became able to speak more quickly. When these data were compared to the control group, significant interactions were not found for any of the five measures: number of syllables spoken (F[1, 169] = 0.79, p=0.38), number of pauses (F[1, 169] = 0.01, p=1), speech rate (F[1, 169] = 1.16, p=0.28), articulation rate (F[1, 169] = 0.6, p=0.44), or ASD (F[1, 169] = 1.61, p=0.15).
An examination of the impact of the independent variables (class level, enjoyment, engaged time) on fluency revealed that class level (F[6, 46] = 4.1, p=0.002, V=0.35) and engaged time (F[6, 46] = 1.96, p=0.04, V=0.4) were found to significantly affect some measures of fluency, although enjoyment did not (F[6, 46] = 0.82, p=0.56, V=0.1). Post-hoc tests revealed that class level impacted the number of syllables spoken (p<0.001), articulation rate (p=0.05), and ASD (p=0.03). Engaged time only impacted the number of pauses between participants who allocated the required amount of time and those who devoted additional time to speaking with their partner (p=0.009).
Class level
Because a number of dependent variables were impacted by class level, and since the control group included students who had enrolled in the same elementary and intermediate level classes, comparisons could be made of intermediate level class VSMC participants (n=24) and intermediate class control group participants (n=15) by extracting just these participants’ data from the main data pool. First, the intermediate class VSMC participants’ pre- and posttest scores were checked for improvement among the impacted factors. Dependent t-tests showed that the intermediate class VSMC participants improved on the number of syllables spoken (t(55)=5.72, p<0.001), and articulation rate (t(55)=2.2, p=0.04), but not on ASD (t(55)=1.63, p=0.12). Next, the intermediate class VSMC participants’ scores were compared to those of the intermediate class control group participants, revealing significant interactions between participation in the program and improvement in the number of syllables spoken (F[1, 77] = 7.19, p=0.01) and articulation rate (F[1, 77] =
4.83, p=0.03), but not in ASD (F[1, 77] = 2.76, p=0.11). Thus, it was found that intermediate class VSMC participants improved their fluency more significantly than control group students who did not participate in the VSMC program, but this same cannot be said for beginner class VSMC participants.
Enjoyment
An analysis of the reflection papers written by students with top quartile improvement in both WPM and MLU (n=10) and those with bottom quartile improvement in both WPM and MLU (n=10) revealed similarities as well as important differences. Participants in both groups offered positive and negative comments, but their experiences and reflections varied greatly.
Students in both groups mentioned several points with similar frequencies. Three of the best improvers and four of the worst improvers expressed unhappiness that their Skype partner in Japan was a volunteer and did not take participation in the program seriously. Furthermore, three of both the best and worst improvers commented that the time
difference between Japan and America made it difficult to schedule conversation sessions and that they had to complete the activities either very late at night or very early in the morning. Finally, five of the best improvers and three of the worst mentioned that the program was a “good experience”. Since these points were mentioned rather evenly among both the best and worst improvers, it is unlikely that any of these reported experiences had an important effect on participants’ improvement in speaking ability.
Data also revealed five key differences between the best and worst improvers. The first and most notable was that most of the best improvers (70%) were from the
intermediate class, whereas most of the worst improvers (80%) were from the elementary class. It is also striking that, although both groups answered overwhelmingly in their surveys that they enjoyed the program (80% each), among the worst improvers, most commented in their reflection papers how much they enjoyed the program (80%), with representative comments such as “I really liked talking to my Skype partner!” and “This program was very fun to be a part of.” In contrast, very few of the best improvers (20%) mentioned enjoyment in their reflection papers. A second important contrast focused on reported improvement. The worst improvers generally reported feeling that they had improved (70%) [“I improved both my listening and speaking in this class” and “My speech and knowledge of Japan has [sic] improved.”]. In contrast, very few of the best improvers reported feeling that they had improved. Instead, the most improved participants wrote that they had gained confidence in speaking (70%) [“I gained confidence” and “I did gain much more confidence in my Japanese speaking.] and mentioned that the program offered a good opportunity to practice speaking (60%) [“(it) was a great opportunity to practice my Japanese” and “(it was the) perfect way to practice Japanese speaking.”]
Similar comments were only made once each by the participants that improved the least, [“I would definitely recommend [the program] to other students who want to gain confidence in speaking;” “I like being able to practice Japanese from the comfort of my home.” Finally, three participants in the most improved group mentioned feelings of embarrassment about not being able to speak or nervousness about conversing with a native speaker, but reported overcoming these issues: [ “[The program] started off kind of awkward, back when
tons in common.” so [in the end we “exchanged contact info, so hopefully we (will) stay friends for a long time!”] Conversely, only one of the worst improvers reported overcoming such problems [“It was frustrating at times and we often had awkward silences while I was trying to think of more questions to push the conversation along;” even at the end, “It didn’t help that my partner would usually give short answers, then expect me to carry the
conversation alone.” Thus, although both the best and worst improvers seemed to have similar communication issues, those who improved most were generally more able to overcome them.
Discussion
This study investigated the impact of participation in a VSCM program with native speaker partners on participants’ oral fluency and sought to determine if instructional level, enjoyment, and/or engaged time were associated with improved fluency. In sum, the initial findings concerning fluency were inconclusive, showing that although participants spoke more, they did not necessarily speak more quickly. These results are congruent with Li (2014) and Yang and Chang (2008) and contrast those reported in Kato et al. (2016), Yanguas (2012) and Yang et al. (2012). They could be due to a number of reasons. First, the metrics in the aforementioned studies are all different. For example, Kato et al. (2016) used speech rate and mean length of utterance as measures of oral proficiency, whereas Yang and Chang (2008) and Yang et al. (2012) used subjective measures and Yanguas (2012) used oral production assessment tasks. As pointed out by Lambert and Kormos (2014), there are a number of metrics for objectively measuring oral communication (i.e.
fluency, accuracy, and complexity) that can potentially influence subjective oral development scores. Furthermore, as noted by Thai and Boors (2016), it is generally the case that only one of the metrics will improve at a time. Thus, it is possible that the participants in this study improved their accuracy or complexity, but such measures were not used. Finally, the settings of the programs in the previous studies and this one could have also influenced the results of each. For example, Yang and Chang (2008) and Yanguas (2012) reported that students had much lower engaged speaking time because VSMC sessions took place in class for only short periods (10-20 minutes), whereas VSMC sessions took place outside of the classroom in Yang et al. (2012), Kato et al. (2016) and the current study and resulted in engaged speaking time that ranged from 15 minutes to two hours per session. Since the current study did find that engaged time resulted in fewer pauses, engaged time could be one reason why most of these studies found some improvement in oral skills.
Concerning the impact of instructional level of the development of oral fluency, the findings of this study suggest that achieving a certain level of proficiency before conducting native speaker paired VSCM can be helpful. This may be due to the fact that more advanced learners are less likely to encounter linguistic structures and vocabulary that are totally unknown, may have more well-developed conversational repair strategies, or are simply more able to sustain a conversation and find points of commonality. This notion is supported by the fact that the best improvers also often noted in their reflection papers that they felt the program was “good practice” and that they “gained confidence” in their speaking ability, while similar comments were rarely offered by the lowest improvers. This
result may also help to explain why studies such as Yang et al. (2012) found improvement while Yang and Chang (2008) did not; it could be that the participants in the latter study were of higher instructional level. However, it is important to point out that this finding contrasts with Lin (2014), who suggested that participants in CMC programs with lower skill levels would likely show greater improvement. Additional studies will be required to determine if this is true at each subsequent level of proficiency or only lower levels, or if this applies only when using VSCM and paired with a native speaker partner.
Concerning enjoyment, data suggest that enjoyment alone does not seem to be heavily associated with improvement. Although questions about enjoyment are often included in studies of VSMC programs (e.g. Yang et al., 2012; Yanguas, 2012) and while some VSMC studies link enjoyment to potential future learning after the program ends (e.g. Kato et al., 2016; Taillefer & Munoz-Luna, 2014), this study suggests that participants can enjoy the program without improving very much, and vice-versa. The qualitative data instead seem to suggest that the circumstances that reduce enjoyment may be of greater importance: some of the best improvers were able to overcome communication and scheduling difficulties, while the worst improvers were less successful in doing so. This may be due to the fact that overcoming communication difficulties is part of meaningful interaction (Woo & Reeves, 2006). Further investigations that solicit deeper explanations concerning program enjoyment may lead to better approaches to preparing students for VSMC and thus helping them to sustain communication and repair miscommunications, and thus support the types of higher-quality interactions that have been shown to be conducive to second language acquisition (Ellis, 1985; Woo & Reeves, 2006).
Taken together, the findings suggest the following specific pedagogical implications when using VSMC with native-speaker partners:
1. Begin utilizing native-speaker partners in VSMC once students have had at least four semesters of language instruction ;
2. Encourage native speakers to adjust their speaking to match partners’ levels; 3. Maximize allocated time by scheduling sessions in mutually acceptable ways; 4. Maximize engaged time by:
(a) Making purposeful partner matches based on common interests;
(b) Offering direct training (e.g. doing in-class rehearsals prior to engaging in Skype sessions);
(c) Offering or assigning topics, tasks, projects or other activities that keep partners engaged and feeling a need to communicate;
(d) Giving direct instruction in conversational maintenance and repair strategies; 5. Require participation by all partners (i.e. the native speakers and learners);
6. Monitor participant satisfaction (e.g. through a weekly survey, worksheet, or report) and help them overcome problems as they arise;
7. Offer advice, instruction or encouragement regarding teamwork, problem solving and negotiation.
While the study uncovered some important findings related to VSCM programs, there are a number of admitted shortcomings. For example, only oral fluency and not complexity or accuracy of speech was evaluated. It is possible that beginning class participants may have shown improvement in these areas, but this could not be gauged by
in this study. Furthermore, although this study found that learners with five semesters of FL learning improved more than learners with three semesters, it is still unclear how much learners at other instructional or proficiency levels will improve. Finally, the present study looked at only a few individual learner characteristics (i.e. instructional level, enjoyment, and engaged time); there are a number of other factors that could potentially be associated with improvement (e.g. changes in confidence, schedule flexibility, and motivation) that could also be addressed. Future work could perhaps evaluate more additional aspects of oral development, observe groups at many different instructional or proficiency levels, and make use of additional psychometrics to investigate other individual learner characteristics that might also be associated with improvement.
Conclusion
The results of this study suggest that there is indeed a place for native speaker paired VSMC in the FL classroom, but the data suggest that participants in such programs should probably have at least five semesters of instruction to make significant gains in oral fluency. Furthermore, the data suggest that increased engaged time -- talking for at least 30 minutes or longer twice per week and speaking in the target language for more than 50% of each VSMC session with native speakers will result in fewer pauses. Finally, it should be noted that high self-reported levels of enjoyment were not found to be associated with improvement in oral fluency. Rather, it seems that the impediments to enjoying participation are more important: the greatest improvers whose reported issues with the program resulted in lower levels of enjoyment generally reported overcoming those
problems, whereas for the worst improvers, problems persisted and participants did not report overcoming them. Implementing the pedagogical strategies suggested above may contribute to the success of this kind of learning opportunity.
References
Bernstein, J., Cheng, J., & Suzuki, M. (2010). Fluency and structural complexity as predictors of L2 oral proficiency. Proceedings of Interspeech 2010, 1241-1244. Boeersma, P. & Weenink, D. (2018). PRAAT: doing phonetics by computer [Computer
program]. Retrieved 14 November 2018 from http://www.praat.org
Brantmeier, C. (2003). Beyond linguistic knowledge: Individual differences in second language reading. Foreign language annals, 36 (1), 33-43.
Brantmeier, C. (2005). Nonlinguistic variables in advanced second language reading: Learners’ self-assessment and enjoyment. Foreign language annals, 38 (4), 494-504. Canto, S., Jauregi, K., & Van den Bergh, H. (2013). Integrating cross-cultural interaction
through video-communication and virtual worlds in foreign language teaching programs: is there an added value? ReCall, 25 (1), 105-121.
Dewaele, J.M., & Alfawzan, M. (2018). Does the effect of enjoyment outweigh that of anxiety in foreign language performance? Studies in second language learning and
teaching, 8 (1), 21-45.
Ellis, R. (1985). Understanding second language acquisition. New York: Oxford University Press.
Hirotani, M., Matsumoto, K. & Fukuda, A. (2012). Longitudinal Study on Fluency Among Novice Learners of Japanese. In L. Bradley & S. Thouësny (Eds.), CALL: Using,
Learning, Knowing, EUROCALL Conference Proceedings (pp. 129-133). Gothenburg,
Sweden: Research-publishing.
Kato, F., Spring, R. & Mori, C. (2016). Mutually beneficial foreign language learning: Creating meaningful interactions through video synchronous computer-mediated communication. Foreign Language Annals, 49(2), 355-366.
Kormos, J. (2006). Speech production and second language acquisition. Mahwah, N.J.: Lawrence Erlbaum Associates.
Krause, K. & Coates, H. (2008). Students’ engagement in first-year university. Assessment
and Evaluation in Higher Education, 33(5), 493-505.
Lambert, C., & Kormos, J. (2014). Complexity, accuracy, and fluency in task-based L2 research: Toward more developmentally based measures of second language acquisition.
Applied Linguistics, 35, 607-614.
Lin, H. (2014). Establishing an empirical link between computer- mediated communication (CMC) and SLA: a Meta-analysis of the research. Language Learning & Technology 18 (3), 120-147.
Matsushita, H. (2011). Computerized oral proficiency test for Japanese: Measuring second
language speaking ability with ASR technology (master’s thesis). Brigham Young
University, Idaho, United States.
Miles, M., & Huberman, A. M. (1994). Qualitative data analysis: An expanded sourcebook (2nd ed.). Thousand Oaks, CA: Sage.
Nauffal, D.I. (2011). Assessment of student engagement: An analysis of trends. Tertiary
Education and Management, 18(2), 171-191.
Rifkin, B. (2005). A ceiling effect in traditional classroom foreign language instruction: Data from Russian. The Modern Language Journal, 89 (1), 3-18.
Saito, K., Dewaele, J.M., Abe, M., & In’nami, Y. (2018) Motivation, emotion, learning experience and second language comprehensibility development in classroom settings: A cross-sectional and longitudinal study. Language Learning, 68 (3), 709-743.
Scarcella, R.C., & Oxford, R. L. (1992). The tapestry of language learning: The individual
in the communicative classroom. Boston: Heinle & Heinle.
Schultz, L.M. (2012). Affect with Chinese learners of English: Enjoyment, self-perception, self-assessment, and abilities across level of language learning. Quarterly journal of
Chinese studies, 5 (2), 65-81.
Taillefer, L., & Munoz-Luna, R. (2014). Developing oral skills through Skype: A language project analysis. Procedia–Social and Behavioral Science, 141, 260–264.
Thai, C. & Boers, F. (2016). Repeating a monologue under increasing time pressure: effects on fluency, complexity and accuracy. TESOL Quarterly, 50 (2), 369-393.
Tian, J., & Wang, Y. (2010). Taking language learning outside the classroom: Learners’ perspectives of E tandem learning via Skype. Innovation in Language Learning and
Teaching, 4, 181–197.
Woo, Y., & Reeves, T. C. (2006). Meaningful interaction in web-based learning: A social constructivist interpretation. The Internet and Higher Education, 10, 15–25.
Yang, Y. C., & Chang, L. (2008). No improvement – Reflections and suggestions on the use of Skype to enhance college students’ oral English proficiency. British Journal of
Educational Technology, 39, 721-725.
Yang, Y. C., Gamble, J., & Tang, S. S. (2012). Voice-over instant messaging as a tool for enhancing the oral proficiency and motivation of English-as-a-foreign-language learners.
British Journal of Educational Technology, 43, 448–464.
Yanguas, I. (2012). Task-based oral computer-mediated communication and L2 vocabulary acquisition. CALICO Journal, 29 (3), 507-531.
Appendix 1 Questions asked on pre- and posttests every year Pretest
1. Tell about the city that you currently live in. What city is it and where is it located? What is interesting or special about the city? What do you like or dislike about the city?
2. Tell about someone who you admire. Who is the person? What good things has this person done? Why do you admire this person?
3. Tell about a place you visited when you were a child. What is the place and where is it? When did you go there and who did you go with? Would you like to go there now? Why or why not?
Posttest
1. Tell about the home you live in. What kind of building do you live in? How long have you lived there? What do you like or dislike about your accommodation?
2. Tell about something you own that is very important to you. What is it and where did you get it from? How long have you had it? Why is it important to you?
3. Tell about a time when you were very happy. Where were you, and what were you doing? Who was with you? Why were you happy?
