The current study was conducted to explore the relationship of English listening performance to two learner variables, English listening anxiety and tolerance of ambiguity for learners at high school levels. In addition, the study examined the two learner factors’ predictability of English listening performance. Data in the study were gathered from 223 high school students from a municipal high school in Taipei, Taiwan, through questionnaires and a GEPT intermediate level listening test. The GEPT
listening test was given to evaluate learners’ English listening performance. To investigate the participants’ anxiety level, Second Language Listening Anxiety Scale (SLLAS) was utilized. Multiple Stimulus Types of Tolerance of Ambiguity Scale (MSTAT-I) was employed to examine the participants’ level of tolerance of ambiguity.
Pearson product-moment correlation coefficient was employed to investigate the correlations between the two learner variables and English listening performance. As their significant correlations had been proven, stepwise multiple regression was
conducted to further explore if the learner variables could serve as significant predictors for English listening performance. This chapter summarizes the findings, along with potential explanations. Finally, pedagogical implications, limitations of the present study, and suggestions for future research are yielded.
English Listening Performance, English Listening Anxiety and Tolerance of Ambiguity
Pearson product-moment correlation was conducted to answer the first research question: What is the relationship of English (FL) listening performance to English (FL) listening anxiety and tolerance of ambiguity among high school learners? As revealed in
44
the previous chapter, a significant, negative and moderate correlation was found between listening anxiety and listening performance (r = -.50, p < .001), which indicates the tendency that learners with a higher level of listening anxiety had lower scores in the English listening comprehension test. Such results echo Brownell’s (1996) model of HURIER listening process and Flowerdew and Miller’s (2005) model of second language listening. As previously discussed, Brownell (1996) argued that anxiety can be a hindrance for the messages entering the listening process and that it lowers a listener’s remembering capability. In the same light, Flowerdew and Miller (2005) claimed that anxiety has an impact on listeners’ motivation and can lower their desire for listening to and understanding the message. Hence, it is reasonable to find that learners with a lower level of listening anxiety had a better listening comprehension performance and vice versa.
The correlations between one’s listening anxiety and listening performance are consistent with the ones uncovered in Yang’s (2012) and Chou’s (2017) studies. The correlations found in both Yang’s (2012) and Chou’s (2017) studies for junior high school students are lower (r = -.32, p < .05; r = -.32, p < .01) than that found in this study (r = -.50, p < .001), yet the correlation attained in the Chou’s study (2017) for senior high school students (r = -.41, p < .01) is close to what was found in the present study. In both Yang’s (2012) and Chou’s (2017) studies with junior high school
students, the instrument administered to evaluate participants’ listening performance was the elementary level GEPT listening test whereas the test used in this study and that for senior high school students in Chou (2017) was the intermediate level of GEPT. The difference in the intensity of correlations between listening anxiety and listening
comprehension performance in the two levels of education (i.e., junior high school and senior high school) may result from the dissimilar level of difficulty of the listening
45
tests.
According to Vogely (1998), attributes of the input such as the nature of the speech and level of difficulty can be the source of anxiety, and based on Yerkes-Dodson Law (Smith, Sarason & Sarason, 1982, cited in MacIntyre, 1995), there is “a curvilinear relationship between anxiety and performance as a function of task difficulty” (p. 92).
As elaborated in Chou (2017), while anxiety may demand increased effort and thus have a facilitating effect on performance when the given task is relatively simple, the effect can turn debilitating as the difficulty of the given task increases, for the extra effort needed may outpace the learner’s ability and fail to offset the increased cognitive demands.
The present finding of a negative correlation between listening anxiety and listening performance is also consistent with the results of Elkhafaifi’s (2005) study on undergraduate and graduate students. However, the correlation found in Elkhafaifi (2005) was relatively stronger (r = -.70, p < .01). Elkhafaifi (2005) conducted the study in an Arabic-as-a-foreign-language context and adopted the listening comprehension grades provided by the participants’ instructors without specifying the types of listening texts and tasks given in class. It is difficult to decide whether the strong correlation attained in Elkhafaifi’s study was derived from the greater difficulty in the listening tasks or not. Note that Elkhafaifi’s (2005) study differs from the current study and Chou (2017) in the target foreign language and learners’ education level. Further research on students at postsecondary level and/or learners of a foreign language other than English is needed to explain what may contribute to differences in the intensity of the
correlation between listening anxiety and listening comprehension performance.
In this study, Pearson product-moment correlation was administered to have a closer inspection of the inter-correlations among the subscales of SLLAS. As can be
46
seen in Table 11, the subscales of SLLAS were modestly correlated with one another (with r ranging from .30 to .40), which concurred with Lang’s (1971) proposition that while dominated by one’s brain mechanisms, the three systems of emotion, namely, verbal-cognitive, somatic, and behavioral system, are independent. In other words, each system can potentially share the same importance when assessing one’s anxiety
(Lawyer & Smitherman, 2004). This result further lent support to Cheng’s (2014, 2017) proposal for developing a three-dimensional measurement of language skill-specific anxiety.
In contrast to listening anxiety, tolerance of ambiguity had a positive yet weak correlation with listening performance (r = .26, p < .001) while the correlation between tolerance of ambiguity and listening anxiety was moderate yet negative (r = -.32, p
< .001). The former correlation is consistent with the ones found in studies conducted by Naiman et al. (1978) and Chapelle and Roberts (1986), both reporting that one’s level of TA positively correlated with the listening performance in language tests. The result also confirmed the contention that good language learners tend to embrace ambiguous situations and are willing to proceed the communication at the cost of
making mistakes (Naiman et al., 1978; Rubin, 1975). With respect to the intensity of the correlation between tolerance of ambiguity and listening performance, the feature of tolerance of ambiguity can be a factor that helps explain the result of such a weak correlation. According to McLain (1993), items included in the MSTAT-I scale are unidimensional, reflecting general tolerance of ambiguity. However, as proposed by Durrheim and Foster (1997), tolerance of ambiguity should be taken as a situation-specific variable, for “within a single individual, high levels of tolerance ambiguity within one content might associate with low levels in another domain, and might be unrelated to AT in a third domain” (p. 748). Ely (1989) also claimed that tolerance of
47
ambiguity could be situation-specific and vary according to the situations encountered.
He thus developed a tolerance of ambiguity scale for second language learning, though without providing any validation evidence for the scale he developed. Due to its
questionable validity and the lack of items concerning English listening in the scale, this study adopted MSTAT-I, instead of Ely’s instrument. It is possible that a stronger correlation may be found if a valid and reliable measurement of TA for L2 listening is used. In other words, despite the low correlation shown in the results, it is premature to conclude that tolerance of ambiguity has little to do with one’s listening performance.
On the other hand, the moderate yet negative correlation found between tolerance of ambiguity and listening anxiety was in line with the contentions made by Bochner (1965) and Smock (1955), both of whom considered anxiety to be a behavioral characteristic of and a negative correlate with tolerance of ambiguity. In other words, less ambiguity-tolerant people tend to feel more anxious when an environment is full of uncertainty. That is, in the present context of English learning, in which unfamiliarity concerning phonemes, word meaning, and grammar abounds, learners with a lower level of TA would feel more anxious when listening to English compared with their L1.
Predictors of English Listening Performance
Stepwise multiple regression was administered to help explore and answer the second research question: Which of the two variables, listening anxiety and tolerance of ambiguity, can serve as a significant and better predictor of English (FL) listening performance for learners at high school levels? The results showed that only listening anxiety served as a significant predictor of one’s English listening performance, explaining 25% of the variance in one’s listening performance. In addition, a further inspection on the subscales of SLLAS specified the behavioral dimension of listening anxiety to be the major predictor among the three, accounting for 33% of the variance in
48
one’s listening performance, with cognitive listening anxiety being the other significant predictor, which explained only additional 3% of the variance in English listening comprehension performance. Based on Lang’s (1971) study, the different predictability of the subscales could also result from the independence among the three response systems of anxiety.
On the other hand, tolerance of ambiguity was excluded from having significant contributions to one’s English listening performance. The potential factor contributing to this result can be the instrument, McLain’s (1993) MSTAT-I, administered in the research. As previously discussed, the 22 items in the scale measure one’s general tolerance of ambiguity and are not directly associated with one’s listening to
second/foreign language. In other words, MSTAT-I scores may not adequately reflect one’s tolerance of ambiguity in situations of listening to English (if L2 listening TA is indeed distinguishable from general TA), lowering MSTAT-I scores’ predictability for one’s listening comprehension performance. Further research is still needed in the future to explore and attest the plausibility of the reason.
Implications of the Study
Based on the finding of the present study, some pedagogical implications for the teaching of English listening comprehension to EFL learners can be provided.
To begin with, since listening anxiety has been proven to significantly correlate to and serve as a predictor to one’s English listening attainment, it is critical that
instructors observe students’ anxiety level and create a non-threatening classroom. As suggested in the results, the behavioral dimension of learners’ listening anxiety best predicted listening performance. Therefore, teachers should become alert when learners perform actions indicative of anxiety, such as avoidance, showing hypervigilance, or having deficits in attention. To reduce learners’ level of anxiety, it is prerequisite that
49
instructors provide appropriate scaffolding and comprehensible input, and be careful about the difficulty level of materials and task types to be included in listening tests.
Furthermore, while the present study failed to verify tolerance of ambiguity as a significant predictor of learners’ listening performance, there was still a significant, modest correlation between tolerance of ambiguity and listening performance. There was also a significant, negative correlation between tolerance of ambiguity and listening anxiety (especially in the cognitive and avoidance behavioral dimensions of anxiety).
Since learners with a low level of tolerance of ambiguity can demonstrate avoidance behaviors and have higher anxiety (Furnham, 1994), which has been proven to have detrimental effects on EFL learning. The results suggest that, on the one hand,
instructors should lower the degree of ambiguity that can occur in EFL class by clearly informing students of the teaching procedures and familiarizing them with the teaching activities. On the other hand, with an attempt to enhance learners’ level of tolerance of ambiguity, teachers should manage to design versatile tasks that allow room for ambiguity so as to encourage students to take risks, make guesses, and become better language learners that are more tolerant of ambiguity (Naiman et al., 1978). For instance, instead of designing activities that require students to fill in blanks based on the provided listening material, teacher can incorporate activities such as note-taking and summarizing, in which learners do not necessarily need to understand every single word, but can still be productive based on the part they hear and understand.
Meanwhile, as opposed to pausing the listening material and translating the listening material word by word or passage by passage, instructors themselves should be tolerant and provide learners time and opportunity to deal with the ambiguity and continue on listening. Moreover, as previously discussed, instructors should stay vigilant that one’s tolerance of ambiguity can be situation-specific. That is, even when learners show traces
50
of high tolerance of ambiguity in daily, general situations, they do not necessarily have high tolerance of ambiguity in a foreign language learning context.
Limitations and Suggestions for Future Study
The present study has its limitations. To begin with, in the current study, the instrument adopted to understand participants’ ability of comprehension was a published listening test on the intermediate level of General English Proficiency Test (GEPT), after which participants filled out the listening anxiety questionnaire. Even though GEPT is a standardized test with validity and reliability, the intermediate level of listening test does not include all the text types and task types, such as monologues and summarizing, which learners are likely to encounter in real life. As text types such as monologue and task types like summarizing can be more difficult and anxiety-provoking (Anderson & Lynch, 1988; Shohamy & Inbar,1991), the GEPT listening test may not fully reflect the participants’ listening ability or provoke the same level of listening anxiety participants would have in other communicative situations. Therefore, it is recommended that future studies incorporate other text types and task types when probing into participants’ listening ability and anxiety.
Second, as the GEPT listening test was only administered once in the present study, the one-shot sampling may fail to fully illustrate the participants’ English listening ability.
In addition, since learner factors such as anxiety and tolerance of ambiguity may have dynamic interplay with various variables inherent in the social context of the learning situations (Dörnyei, 2014), the one-shot sampling from the self-report listening anxiety and tolerance of ambiguity questionnaires in this study may not reflect learners’ levels of anxiety or tolerance of ambiguity throughout the listening process. Hence, it is recommended to adopt a longitudinal design, and include interviews in future studies to help collect more in-depth data.
51
Third, the instrument utilized to investigate the participants’ tolerance of ambiguity was the MATAT-I developed by McLain (1993). It should be noted that what MSTAT-I evaluates is one’s general tolerance of ambiguity rather than one’s tolerance of ambiguity in listening to second/foreign language. Therefore, the participants’ tolerance of ambiguity for comprehending foreign language messages may not be authentically presented. Regardless of the fact that Ely (1989) developed a second/foreign language tolerance of ambiguity scale, only one item from the scale is related to listening, and its validity is still to be established. Accordingly, in order to depict one’s actual tolerance of ambiguity in second/foreign language listening, it is recommended that future studies first develop a valid, reliable scale of second/foreign language listening tolerance of ambiguity.
Finally, the participants in the current study were restricted to 223 eleventh grade students from one municipal high school in Taipei. As a result, it remains debatable whether the results can be applied to students at different grades at high school level or in other places. It is suggested that careful replication with populations in different regions and at different grade levels be conducted in the future.
Conclusions
The major findings corresponding to the two research questions addressed in the present study are summarized as follows: (1) Listening anxiety was found to significantly and moderately correlated to English listening performance in a negative direction.
Among the three dimension of anxiety, behavioral dimension was proven to have the strongest correlation with English listening performance. In terms of the correlation between tolerance of ambiguity and English listening performance, only a significant yet modest correlation was discovered. (2) For the two learner variables explored in the study, only listening anxiety served as a significant predictor for English listening performance, with the behavioral dimension demonstrating the highest predictability; tolerance of
52
ambiguity had no significant predictability for English listening performance.
Overall, the study made some contributions to a better understanding of EFL listening anxiety and tolerance of ambiguity among high school students in Taiwan, regardless of the refinements needed in sampling of participants and measuring tolerance of ambiguity. The results of this study lend support to the detrimental effect anxiety can have on EFL learners’ listening comprehension, alerting teachers to the need of caring for students’ listening anxiety level. By pinpointing the urgency of developing a valid measurement for EFL listening tolerance of ambiguity, it is hoped that this study can be conducive to more refined research on EFL listeners’ tolerance of ambiguity.
53
REFERENCES
Aida, Y. (1994). Examination of Horwitz, Horwitz, and Cope's construct of foreign language anxiety: The case of students of Japanese. Modern Language Journal, 78, 155-168.
American Psychological Association. (n.d.). Anxiety. Retrieved April 9, 2019, from https://www.apa.org/topics/anxiety
Anderson, A., & Lynch, T. (1988). Listening. Oxford: Oxford University Press.
Anderson, J. R. (1995). Cognitive psychology and its implications. (5th ed). New York, NY: Worth Publishers.
Aneiro, S. M. (1990). The influence of receiver apprehension in foreign language learners on listening comprehension among Puerto Rican college students.
Unpublished doctoral dissertation. New York University.
Ba, H. B. (2012). A study of the correlation between processing strategies and tolerance of ambiguity in listening comprehension of college English. Journal of Yangtze Normal University, 28, 104-109.
Bachman, L. F., & Palmer, A. S. (1996). Language testing in practice: Designing and developing useful language tests. Oxford: Oxford University Press.
Baran-Łucarz, M. (2009). The level of ambiguity tolerance as a determinant of accuracy in non-instructed FL pronunciation. Report on a pilot study. Anglica
Wratislaviensia, 47, 91-103.
Basoz, T. (2015). Exploring the relationship between tolerance of ambiguity of EFL learners and their vocabulary knowledge. Journal of Language and Linguistic Studies, 11, 53.
Benjamin, A. J., Riggio, R. E., & Mayes, B. T. (1996). Reliability and factor structure of Budner's tolerance for ambiguity scale. Journal of Social Behavior and
54
Personality, 11, 625-632.
Bochner, S. (1965). Defining intolerance of ambiguity. Psychological Record, 15, 393-400.
Brown, H. D. (2000). Principles of language learning and teaching. White Plains, NY:
Longman.
Brownell, J. (1996). Listening attitudes, principles, and skills. New York, NY: Pearson.
Brownell, J. (2008). Building managers' skills to create listening environments. Cornell Hospitality Tools, 11, 6-17.
Budner, S. (1962). Intolerance for ambiguity as a personal variable. Journal of Personality, 30, 29-50.
Chang, A. C. S. (2008). Sources of listening anxiety in learning English as a foreign language. Perceptual and Motor Skills, 106, 21-34.
Chang, A. C. S., & Read, J. (2008). Reducing listening test anxiety through various forms of listening support. TESL-EJ, 12, 1-25.
Chang, W. C. (2006). English language education in Taiwan: A comprehensive
survey. Bimonthly Journal of Educational Resources and Research, 69, 129-144.
Chapelle, C., & Roberts, C. (1986). Ambiguity tolerance and field independence as predictors of proficiency in English as a second language. Language Learning, 36, 27-45.
Cheng, Y. S. (2005). EFL learners’ listening comprehension anxiety. English Teaching
& Learning, 29, 25-44.
Cheng, Y. S. (2014). Toward a structural model of L2 language-skill-specific anxiety.
(Report No. MOST 100-2410-H-003-098-MY2) Taipei, Taiwan: Ministry of Science and Technology.
Cheng, Y. S. (2017). Development and preliminary validation of four brief measures of
55
L2 language-skill-specific anxiety. System, 68, 15-25.
Cheng, Y. S., Horwitz, E. K., & Schallert, D. L. (1999). Language anxiety:
Differentiating writing and speaking components. Language Learning, 49, 417-446.
Chou, H. C. (2017). A study on the relationships of EFL learners’ listening performance to listening anxiety, listening self-efficacy, and metacognitive awareness.
(Unpublished master’s thesis). National Taiwan Normal University, Taipei City.
Cross, J. (2010). Raising L2 listeners’ metacognitive awareness: A sociocultural theory perspective. Language Awareness, 19, 281-297.
Dalman, R. M. (2016). The relationship between listening anxiety, listening comprehension strategies, and listening performance among Iranian EFL university students. International Journal of Modern Language Teaching and Learning, 1, 241-252.
Dörnyei, Z. (2014). The psychology of the language learner: Individual differences in
Dörnyei, Z. (2014). The psychology of the language learner: Individual differences in