November 2017 – Volume 21, Number 3
Junya Fukuta
Faculty of Education, Shizuoka University, Japan
<fukuta
shizuoka.ac.jp>
Abstract
This study explored the longitudinal changes of Japanese EFL learners’ psychological attributes of unwillingness to speak English in a task-based classroom. Thirty-three university students participated in this study. They engaged in a communicative task once a week for two semesters. They completed a questionnaire examining the psychological attributes of their task performance. The results suggest that overall task-based communicative lessons were effective in reducing unwillingness-related psychological attributes. The following qualitative analysis revealed that the learners became aware of the gap between their perceived and actual L2 skills, which reduced some learners’ confidence. Additionally, the results indicated that personality, such as introversion, low-cooperativeness, or lack of sociability affected the extent to which unwillingness-related attributes were impacted. The paper concludes with a discussion of the positive effects of task-based lessons on Japanese EFL learners’ unwillingness to speak English, and the possible latent variables which weaken the effectiveness of these lessons.
Background
Willingness to communicate (WTC) is one of the individual affective factors explored in applied linguistics. WTC has been defined as a student’s intention to interact with others in the target language, given the chance to do so (Oxford, 1997, p. 449). Second language (L2) WTC research has empirically shown that WTC increases the frequency of use of foreign languages both inside and outside of the classroom, and a high frequency of use is associated with more L2 learning (e.g., Yashima, Zenuk-Nishide, & Shimizu, 2004). Moreover, WTC is also considered as a nonlinguistic outcome of the L2 learning process (MacIntyre, 2007).
L2 WTC has been regarded as having dual constructs. One is trait-like and the other is situational or dynamic WTC (e.g., Cao, 2011; McCroskey & Richmond, 1991). The trait “WTC” refers to tendencies to approach communication as stable across different situations, and situational WTC predicts the decision to initiate communication in a specific situation. The trait WTC (hereafter, simply “WTC”) is stable but may change as a result of learning experience (Cao, 2011).
How, then, does one’s learning experience change WTC? Two main influential psychological attributes are proposed; low perceived competence, which refers to low sense of own competence, and anxiety, which is the feeling of being worried about, in this case, communication. These attributes also relate to unwillingness to communicate, which is defined as a chronic tendency to avoid and/or devalue oral communication and to view the communication situation as relatively unrewarding (Burgoon, 1976, p. 60). Unwillingness to communicate can be considered to be induced by low WTC, and willingness to communicate and unwillingness to communicate are not simply the two sides of the same coin. This is because zero-unwillingness does not necessarily refer to high willingness to communicate. This unwillingness induces classroom silence, which is an obstacle to acquiring the target language in a classroom, and results from avoidance of communication (Harumi, 2011). For successful communicative lessons, therefore, reducing unwillingness to communicate (or unwillingness to speak English) is considered important.
An increase in unwillingness to communicate takes the following course (Figure 1).
Figure 1. The changes of the three sub-factors.
As Figure 1 shows, lower perceived communicative competence and high communicative anxiety, which are related to L2 confidence, negatively affect WTC (e.g., Cao, 2011) and are therefore considered to increase unwillingness. As mentioned previously, high unwillingness incorporates the tendency of more avoidance of starting and continuing communication. Therefore, reducing low perceived competence and anxiety is important for decreasing unwillingness to speak in an L2, which results in learner’s behavior, say, starting communication. This process changes dynamically, associated with other psychological factors such as motivation, international posture, attitudes, and personality (e.g., MacIntyre & Charos, 1996; Yashima et al., 2004).
However, very little research has been conducted longitudinally to explore how unwillingness-related attributes change in association with each other. Also, it is unclear how these attributes are affected by other factors, in the communication-based classroom. Longitudinally, it is possible that some learners are encouraged to speak in communicative lessons, but others are rather discouraged. This issue can be revealed by qualitative data analysis, but previous studies did not emphasize this point. The current study examines these two aspects by means of qualitative and quantitative data analysis.
The study also has pedagogical importance. Teachers would like to know whether task-based language teaching (TBLT) is applicable to a language classroom with low proficient, unmotivated learners. Teachers would also like to know whether and why such learners are encouraged or discouraged to speak in the L2 in communicative TBLT. Since speaking has been considered to be the most anxiety-provoking modality (Horwitz, Horwitz & Cope, 1986; MacIntyre & Gardner, 1991), low proficient learners could potentially decrease their willingness through communicative lessons.
To address the issues discussed above, the current study sets the following research questions:
RQ1: Does the repetition of communicative tasks enable English learners of low proficiency level to decrease their unwillingness to communicate?
RQ2: If not, what disturbs the positive change?
Methodology
Participants
Thirty-three first year students (male: n = 7, female: n = 26) majoring in arts or design in a Japanese university participated in this study. They attended two 90-minute English oral communication classes taught each week. None of them had any experience of living in English-speaking countries for more than a month. Their English proficiency was very limited; most participants had difficulty in comprehending short oral sentences such as Where are you from? No students rated themselves with a score higher than 2 (beginner) in a 7-point self-rating scale of the questionnaire (including speaking, reading, writing, listening, grammar, and vocabulary) ranging from 0 (introductory) to 6 (advanced).
Materials
This paper defines task in terms of the following criteria summarized by Ellis and Shintani (2014, pp. 135-136).
- The primary focus should be on meaning;
- There should be some kind of gap;
- Learners should largely rely on their own resources;
- There is a clearly defined outcome other than the use of language.
The main tasks used included a picture-description task (which asked the learner to orally describe the information presented in a picture to a partner), two spot-the-difference tasks (in which paired participants had similar pictures with some differences and communicated with each other to find the differences), 13 problem-solving tasks (which had some information gaps to fill and each had a single outcome), and four decision-making tasks (which required discussions to reach an agreed solution). These tasks were extracted from three textbooks (i.e., Rocks, 1994; Takashima, 2005; Sato, 2010) and revised by the instructor in accordance with the learners’ proficiency level and cultural background.
Procedure
The task-treatment phase follows the framework of Willis (1996). In this framework, a class consists of three parts: (1) the pre-task, (2) the main task or task cycle, and (3) the post-task or language focus. The pre-task activity aims to introduce the topic of the main task and activate the learners’ background knowledge. In the main task phase, learners perform each task. Finally, if needed, learners engage in form-focused activity in the post-task phase, for example, reading a model-dialog of the main task or watching videos with a focus on expressions or grammar. Each class consisted of the single task-treatment phase, and the phase was repeated 20 times over two semesters. Table 1 summarizes the overall design of the study.
Table 1.
Lesson Procedures
| Semester | Week | Content |
| First Semester | 1 | Orientation, filling in a consent form |
| Assessing psychological attributes (1) | ||
| 2-12 | Task-based lessons | |
| 13 | Final exam | |
| Second Semester | 1 | Orientation Assessing psychological attributes (2) |
| 2-12 | Task-based lessons | |
| 13 | Assessing psychological attributes (3) | |
| Two weeks later | Questionnaire for qualitative analysis |
Questionnaires
Attribution questionnaire. This study administered a questionnaire devised by Isoda (2009), which was constructed and validated for Japanese EFL learners to assess unwillingness to speak English composed of three psychological attributes (item descriptions are written in Japanese). Here, unwillingness to speak English is considered as a formative construct composed of three psychological attributes, namely, low perceived competence, anxiety, and avoidance. Therefore, the questionnaire had three items for three each psychological attribute (e.g., Low Perceived Competence: I think my English is not understood by others; Anxiety: I feel nervous when I speak in English; Avoidance: I do not want to speak in English if it is possible). The participants were asked to answer each question in using a 7-point scale (0: strongly disagree, 7: strongly agree). A high value indicated a negative attitude to speaking English. The questionnaire was conducted three times during the 15-week semester; in the early part of the first class in the first semester, the beginning of the second semester, and the last part of the final class in the second semester. Reliability was calculated for each factor and each survey time. The Cronbachs alpha was shown to be sufficiently high (.77 – .89) for later analysis.
Reflection questionnaire. This questionnaire consisted of two open-ended questions asking about the learners’ feelings during the lessons. It was conducted two weeks after the third questionnaire survey. Before the participants completed the questionnaire, the researcher showed them a table and a figure containing the actual data representing their changes for each attribute and then asked them to write freely why they thought they had changed in the way they did. The participants completed the questionnaire in their native language, Japanese, and these descriptions were translated by the researcher.
Data analysis
Descriptive statistics were calculated for each psychological attribute and data point. The mean score of each attribute was subjected to a within-participant one-way Analysis of Variance (ANOVA) and a subsequent post-hoc analysis whose alpha was adjusted in Holm’s method.
The reflection questionnaire was also qualitatively analyzed in order to better interpret the quantitative data and focus on the minority case which is possibly buried in statistical analysis. All the names of the participants shown in the following sections are replaced with pseudonyms.
Results
The results of the descriptive statistics are summarized in Table 2. The overall score, which refers to the score of each attribute, showed that each value decreased throughout the treatment.
Table 2.
Descriptive Statistics in Each Survey and Factor (N = 33)
| First | Second | Third | ||
| Overall | M | 5.74 | 5.42 | 4.90 |
| SD | 1.25 | 1.20 | 1.22 | |
| Competence | M | 6.20 | 5.53 | 4.78 |
| SD | 0.82 | 1.11 | 1.14 | |
| Anxiety | M | 5.80 | 5.59 | 5.20 |
| SD | 1.27 | 1.16 | 1.23 | |
| Avoidance | M | 5.22 | 5.15 | 4.74 |
| SD | 1.37 | 1.26 | 1.24 |
The results of an ANOVA showed the differences were statistically significant, and the post-hoc analysis also confirmed this (Table 3).
Table 3
The Results of Post-hoc Analysis of Overall results, p is adjusted in Holms way
| Pair | Difference | t | p | r | |
| 1st – 2nd | -0.32 | 2.19 | .036 | * | .36 |
| 2nd – 3rd | -0.50 | 4.04 | < .001 | * | .58 |
| 1st – 3rd | -0.85 | 5.17 | < .001 | * | .67 |
With respect to each of the attributes, a two-way ANOVA (Factor * Time) found a significant interaction, F = 4.17, p = .009, ?p2 = 0.12. Therefore, the simple effects for the interaction were further examined. The results showed that all effects of Time for low perceived competence (F [2, 64] = 27.24, p < .001, ?p2 = 0.46) and anxiety (F [2, 64] = 4.44, p = .016, ?p2 = 0.12) were statistically significant, and that for avoidance was marginally significant, F (2, 64) = 3.18, p = .059, ?p2 = 0.46. Multiple comparisons of all factors are presented in Table 4. With respect to low perceived competence, positive effects were shown in every data point. Anxiety showed similar results, but the easing effects were much more moderate. Although the change throughout the period (difference between the first and third surveys) was statistically significant, the differences between the first and second, and second and third did not reach statistical significance. The avoidance scale presented a similar result in terms of multiple comparisons, but showed a unique pattern in terms of effect size. That is, there were almost no effects in the first semester (r = .06), but the effects suddenly became larger in the second semester (r = .33).
Table 4
The results of Multiple Comparison in Each Factor
| Pair | Difference | t | p | r | ||
| Competence | 1st – 2nd | -0.68 | 3.82 | < .001 | * | .55 |
| 2nd – 3rd | -0.75 | 4.06 | < .001 | * | .58 | |
| 1st – 3rd | -1.43 | 6.70 | < .001 | * | .76 | |
| Anxiety | 1st – 2nd | -0.21 | 1.05 | .302 | .18 | |
| 2nd – 3rd | -0.43 | 1.98 | .112 | .33 | ||
| 1st – 3rd | -0.63 | 2.68 | .035 | * | .42 | |
| Avoidance | 1st – 2nd | -0.32 | 0.32 | .750 | .06 | |
| 2nd – 3rd | -0.49 | 2.03 | .102 | .33 | ||
| 1st – 3rd | -0.41 | 2.69 | .034 | * | .42 |
To look closely at the individual data, learners who Reduced, Increased, and Not Changed values for each data point were also examined (Table 5). This revealed that, in terms of overall score, 73% of the participants had decreased scores for the unwillingness-related attribute between 1st and 2nd time point, and for 6% the score did not change. However, despite the significant improvement of the mean score, 21 % of the participants showed greater unwillingness. Each attribute, except for low perceived competence, also showed similar results. More than half of the participants successfully reduced their attribute scores throughout the treatment period, notably in terms of anxiety and avoidance, where approximately a quarter of the them increased their scores. A relatively strong positive result was shown in the low perceived competence scale. The positive effects of communicative task on low perceived competence seemed to accumulate through treatments, and finally nearly 80% of the participants showed positive effects while only two participants showed increased negatively lack of perceived competence. Anxiety showed much more moderate reduction effects than low perceived competence, in both the first and second periods. As for avoidance, 13 of the participants showed a positive reducing effect, but 12 of the participants showed a negative effect in the first-second period. However, the reducing effect was more strongly shown in the second-third period. A quarter of the participants showed greater anxiety, but more than half of the participants reduced it.
Table 5.
A Breakdown List of Reduced, Increased, and Not Changed, Comparing Each Survey.
| First – Second | Second – Third | First – Third | |||||
| n | % | n | % | n | % | ||
| Overall | Reduced | 20 | 61% | 24 | 73% | 24 | 73% |
| Increased | 8 | 24% | 6 | 18% | 7 | 21% | |
| Not Changed | 5 | 15% | 3 | 9% | 2 | 6% | |
| Competence | Reduced | 21 | 64% | 25 | 76% | 26 | 79% |
| Increased | 3 | 9% | 5 | 15% | 2 | 6% | |
| Not Changed | 9 | 27% | 3 | 9% | 5 | 15% | |
| Anxiety | Reduced | 15 | 45% | 17 | 52% | 20 | 61% |
| Increased | 8 | 24% | 6 | 18% | 8 | 24% | |
| Not Changed | 10 | 30% | 10 | 30% | 5 | 15% | |
| Avoidance | Reduced | 13 | 39% | 18 | 55% | 17 | 52% |
| Increased | 12 | 36% | 8 | 24% | 8 | 24% | |
| Not Changed | 8 | 24% | 7 | 21% | 8 | 24% | |
These results suggest that the communicative TBLT were holistically effective for changing unwillingness to speak English. Although to confirm the causal effect we need further research with an appropriate research design to obtain highly refined evidence (Imbens & Rubin, 2015), the results, at least as overall trends, indicate that learners did not increase their unwillingness to speak English, contrary to the prediction from Horwitz, Horwitz and Cope (1986).
To answer Research Question 2, the reflection questionnaire completed by the participants whose negative attributes increased was analyzed. The examples below are of participants of whose scores lack of competence, anxiety, or avoidance increased.
Example 1: Participants whose low perceived competence increased
Nana: I thought I could use English better. When I tried to use English, however, I sometimes used Japanese and I noticed I could not use English as well as I thought before.
Shinya: I have not felt such strong unwillingness to speak English, although I cannot speak English so well…
Both participants in Example 1 originally seemed to have had confidence in speaking English. The experience of speaking English, however, seemingly changed their mind. This suggests that realistic experience of task engagement possibly made learners realize their actual communicative and linguistic skills.
The excerpts in Example 2 are extracted from the descriptions by participants whose anxiety scores increased. Their descriptions lead us to believe that they might actively communicate with other students if they had more confidence.
Example 2: Participants whose anxiety increased
Kota: The class was fun. I like to speak, but I do not have confidence at all. I wanted to speak Japanese.
Namiko:I could not communicate with other people well, so I feel that I could not use English so well.
Example 3 is from participants whose avoidance score increased. All these participants hesitated to talk with other people. Neither Natsumi nor Rina seemed to have a positive attitude toward communicative TBLT. Notably, the influence of personality as a mediating factor between TBLT and unwillingness to speak English is indicated from their descriptions.
Example 3: Participants whose avoidance increased
Natsumi: I’m shy. Therefore, I could talk with my friends who I knew well, but I could not talk with people I did not know so well.
Rina: By my nature, I do not like talking with others, either inside or outside the class. I was very uncomfortable when I had to communicate with other people. I would like to study English by myself.
To sum up, the results of the qualitative analysis revealed that some of the learners became aware of the gap between their perceived and actual L2 skills and this experience possibly reduced some learners’ confidence. In addition, communication-based activities may increase the unwillingness to speak of learners who have certain personality traits such as, strong introversion.
Discussion
The first research question asked whether repeating communicative tasks enables learners to reduce their unwillingness to communicate. The results showed that the communicative lessons positively affected the learners’ unwillingness, whereas a few learners showed no or negative effects. The positive effect was particularly evident on low perceived competence, whereas less effectiveness for easing anxiety and avoidance for some learners was implied.
The following hypotheses emerge from the results. First, communicative tasks are holistically effective, but it takes a long time for unwillingness to speak English to be reduced. The results indicated that just one semester (tasks repeated 10 times) may not be enough to reduce anxiety and its consequence, avoidance. Second, the results showed that the tendency of avoidance may be reduced after both of the L2 confidence-related attributes (i.e., low perceived competence and anxiety) reach ideal levels. That is, reducing only low perceived competence is not enough to reduce avoidance.
The second research question asked what contributed to an increase in unwillingness to communicate. The quantitative analysis showed that realistic experiences of task engagement made learners realize their actual communicative and linguistic skills. The results also indicated that the opportunity to communicate in English tended to increase the unwillingness to speak of learners with an introverted nature.
This study has some implications. First, although learners experienced positive gains overwhelmingly, communicative TBLT may not be suitable for L2 learning of introverted learners because avoidance of communication induced by increased unwillingness causes an obstacle to L2 learning. A possible solution is to utilize input-based tasks (e.g., Shintani, 2012). Language teachers would be required to choose an appropriate task type depending on the learners in their own classrooms. Future studies to explore the interaction between the degree of introversion and task type (output- or input-based) are also desired.
Another notable implication is that teachers should take care of learners who encounter a gap between their perceived L2 communicative skill and the actual skill and knowledge they possess, as well as psychological aptitude-treatment interaction. As for the particularly important perspective, students who are aware of the gap between their perceived and actual skills may show an increased unwillingness to communicate. Thus, it is very important that these types of learners, before engaging in communicative lessons, must have confidence, or even skills, to speak in a second language to some extent. This means that an introverted nature is not only involved in risk, but that their unwillingness will increase through communicative lessons. Language teachers should pay considerable attention to learners’ dynamic internal change in a class and must support these types of learners.
We need to acknowledge some limitations of this study. The study is a small one, with non-randomized samples, without a control group. Therefore, the current study cannot generalize the results and cannot show conclusive causal relations between communicative lessons and learners’ unwillingness. Clearly, replications with appropriate research design along with, for example, “potential outcome framework” (Imbens & Rubin, 2015) to inference causal effects and to obtain better evidence are clearly recommended. Finally, the current study did not take account of learners’ linguistic behavioral changes, namely, language acquisition or interlanguage development. This calls for future research investigating unwillingness-related attributes as one of the moderating factors between instruction and learning. This line of study will produce better ideas for successful L2 class management
About the Author
Fukuta Junya is a project assistant professor at University of Shizuoka, Language and Communication Research Center. His research interest includes the role of attention and awareness in second language acquisition and task-based language teaching.
References
Burgoon, J. K., (1976). The Unwillingness to communicate scale: Development and validation. Communication Monograph 43, 6069. doi:10.1080/03637757609375916
Cao, Y. (2011). Investigating situational willingness to communicate within second language classrooms from an ecological perspective. System, 39, 468-479. doi:10.1016/j.system.2011.10.016
Ellis, R., & Shintani, N. (2014). Exploring language pedagogy through second language acquisition research. New York: Routledge.
Harumi, S. (2011). Classroom salience: Voices from Japanese EFL learners. ELT Journal, 65, 260-269.
Horwitz, E. K., Horwitz, M. B., & Cope, J. (1986). Foreign language classroom anxiety. The Modern Language Journal, 70, 125-132.
Imbens, G. W. & Rubin, D. B. (2015). Causal inference for statistics, social, and biomedical sciences: An introduction. Cambridge University Press.
Isoda, T. (2009). Reducing EFL learners’ unwillingness to speak English. JACET Journal, 48, 53-66.
MacIntyre, P. D. (2007). Willingness to communicate in the second language: Understanding the decision to speak as a volitional process. The Modern Language Journal, 91, 564-576. doi:10.1111/j.1540-4781.2007.00623.x
MacIntyre, P., & Charos, C. (1996). Personality, attitudes, and affect as predictors of second language communication. Journal of Language and Social Psychology, 15, 2-26. doi:10.1177/0261927X960151001
MacIntyre, P. D., & Gardner, R. C. (1991). Methods and results in the study of foreign language anxiety: A review of the literature. Language Learning, 41, 85-117.
McCroskey, J. C., & Richmond, V. P. (1991). Willingness to Communicate: A Cognitive View. In M. BoothButterfield (Ed.), Communication, Cognition, and Anxiety (pp. 19-37), Newbury Park, CA: Sage.
Oxford, R. (1997). Cooperative learning, collaborative learning, and interaction: Three communicative strands in the language classroom. The Modern Language Journal, 81, 443 456. doi:10.1111/j.1540-4781.1997.tb05510.x
Rooks, G. (1994). Let’s start talking: Conversation for high beginning and low intermediate students of English. Florence, KY: Heinle Cengage Learning.
Sato, Y. (2010). Fookasu on Foom de dekiru! Atarashii bunpou aidea waaku: Koukou. Tokyo: Meiji Tosho.
Shintani, N. (2012). Input-based tasks and the acquisition of vocabulary and grammar: A process-product study. Language Teaching Research, 16, 253-279. doi:10.1177/1362168811431378
Takashima, H. (2005). Task activities and tasks for form-focused instruction and assessment. Tokyo: Taishukan.
Willis, J. (1996). A framework for task-based learning. Oxford: Oxford University Press.
Yashima, T. (2002). Willingness to communicate in a second language: The Japanese EFL Context. The Modern Language Journal, 86, 54-66. doi:10.1111/1540-4781.00136
Yashima, T., Zenuk-Nishide, L., & Shimizu, K. (2004). Influence of attitudes and affect on L2 communication: A study of Japanese high school students. Language Learning, 54, 119-152.
| Copyright rests with authors. Please cite TESL-EJ appropriately. Editor’s Note: The HTML version contains no page numbers. Please use the PDF version of this article for citations. |