This chapter presents analyses of the data collected through the GEPT listening comprehension tests, background information questionnaire, English listening anxiety questionnaire, and Multiple Stimulus Tolerance of Ambiguity Test questionnaire. The findings are illustrated in the following four sections. The first section presents the background information of the participants. The second section reports the relationships of English listening performance to listening anxiety and tolerance of ambiguity. The third section reveals the predictors for high school students’ English listening
comprehension. With the summary of the results of the present study, the last section demonstrates the conclusion of this chapter.
Results of Background Information Questionnaire
The number of valid questionnaires collected from the participants was 223.
According to the background information questionnaires, the participants in the current study included 104 (46.6%) male students and 119 (53.4%) female students. As shown in Table 5, over half of the participants had learned English for eleven to fifteen years (58.7%) or for six to ten years (39.4%). Only one participant had learned English for more than fifteen years (0.4%), and three within five years (1.3%).
Table 6
Participants’ Reported Years of English Learning
Items n %
No more than five years 3 1.3
Six to ten years 88 39.4
Eleven to fifteen years 131 58.7
More than fifteen years 1 0.4
35
Note. N = 223
Table 6 reveals the participants’ perception of the most difficult English skill.
Based on the participants’ self-report, English writing (38%) was considered the most difficult skill among the four skills, followed by speaking (35%), listening (18%), and reading (9%).
Table 7
Participants’ Perception of the Most Difficult Skill
Items n %
Listening 41 18
Speaking 78 35
Reading 19 9
Writing 85 38
Note. N = 223
In terms of the materials participants adopted for their English listening practice outside of the classroom, as can be seen in Table 7, 59 participants (26%) tended to choose English online resources such as Voice of America and VoiceTube as their English listening material, 49 participants (22%) listened to programs that went with English magazines, 18 participants (8%) took English radio broadcast as the source of listening practice, and 19 participants trained their listening ability through other materials such as English songs or having conversations with their foreign friends or teachers. However, it is worth noticing that 121 participants (54%) reported that they had no extra listening practice outside of the classroom.
36
Table 8
Reported Materials for English Listening Practice Outside of the Classroom
Items Frequency %
No extra practice 121 54
English magazines 49 22
English radio broadcast 18 8
English online resources 59 26
Others 19 9
Note. The materials that each participant reported might be more than one. Therefore, the sum of items chosen by each group may exceed the number of participants (N
= 223).
With regard to the time spent weekly on English listening outside the classroom, as presented in Table 8, almost half of the participants (47%) practiced listening for less than half an hour, and only 19 participants (8%) practiced English listening for more than one hour per week on average.
Table 9
The Time Allocated for English Listening on a Weekly Basis Outside the Classroom
Items n %
Less than an hour 104 47
One to two hours 50 22
Two to four hours 35 16
Four to six hours 15 7
Six to eight hours 16 7
More than eight hours 3 1
Note. N = 223
37
Results of Listening Anxiety Questionnaire
As mentioned in the previous chapter, the potential score of the listening anxiety questionnaire, SLLAS, adopted in the current study ranges from 9 to 45, with 3 to 15 points from each of the three dimensions, namely, cognitive anxiety, somatic anxiety, and avoidance behavior. As can be seen from Table 9, the mean score of the scale was 23.87. While the score indicates that the high school students did not seem to have a high level of listening anxiety, it should be noted that the mean score of cognitive aspect of listening anxiety (7.49) greatly outscored that of somatic (4.45) and behavioral (4.02) aspects.
Table 10
Descriptive Statistics of Listening Anxiety Questionnaire
Variable Min Max Mean Score SD
LAS 11 45 23.87 6.23
Cognitive 2 10 7.49 1.90
Somatic 2 10 4.45 1.89
Behavioral 2 10 4.02 2.00
Note. LAS = listening anxiety total; N = 223
Relationships of English Listening Performance to Listening Anxiety and Tolerance of Ambiguity
Table 10 reports the descriptive statistics of the participants’ listening performance on the GEPT listening comprehension test and their responses to the whole listening anxiety questionnaire as well as tolerance of ambiguity questionnaire. As shown in Table 10, the mean score of the GEPT listening comprehension test was about 32 (out of 45), and the mean scores of their listening anxiety and tolerance of ambiguity were
38
about 24 (out of 45) and 65 (out of 100) respectively. With regard to the mean scale score, the score of tolerance ambiguity (2.95) was slightly higher than that of listening anxiety scale (2.65).
Table 11
Descriptive Statistics of the Questionnaires
Variable Min Max
Note. LS = listening score; LAS = listening anxiety score; TAS = tolerance of ambiguity score; NA = not applicable; N = 223
With a view to answering the first research question, the relationships between English listening performance and the two learner factors, listening anxiety and tolerance of ambiguity, were explored. Pearson product-moment correlation analyses were conducted, and the correlation coefficients are presented in Table 11. Table 11 showed that both listening anxiety and tolerance of ambiguity correlated significantly with the English listening score, and the two learner factors were also significantly correlated with each other. Compared with the correlation found between tolerance of ambiguity and English listening comprehension score (r = .26, p < .001), English listening anxiety showed a significant, negative, yet higher correlation (r = -.50, p
< .001) with English listening. In other words, whereas participants with a higher level
of tolerance of ambiguity tended to have a better performance on the listening
comprehension test, participants that experienced higher listening anxiety were inclined to acquire lower listening comprehension scores. The correlation between English
39
listening anxiety and tolerance of ambiguity was also negative (r = -.32, p < .001).
The respective effect of the three dimensions of the SLLAS on listening comprehension performance was also analyzed. As can be seen in Table 11, the behavioral dimension showed the highest negative correlation with listening comprehension scores (r = -.58, p < .001), followed by the cognitive dimension (r = -.32, p < .001) and somatic dimension (r = -.21, p < .001). It is noteworthy that the behavioral
dimension (r = -.34, p < .001) and cognitive dimension (r = -.28, p < .001) also showed significant and negative correlation with tolerance of ambiguity, which indicates that participants that had lower cognitive and behavioral anxiety tended to have higher tolerance of ambiguity. However, no significant correlation was found between the somatic dimension and tolerance of ambiguity.
Table 12
Correlation Matrix of the Variables
Variables LS LA LA-C LA-S LA-B TA
Listening Score (LS) 1
Listening Anxiety (LA) -.50*** 1
Cognitive (LA-C) -.33*** .73*** 1
Somatic (LA-S) -.22*** .74*** .40*** 1
Behavioral (LA-B) -.59*** .73*** .30*** .31*** 1
Tolerance of Ambiguity (TA) .26*** -.32*** -.28*** -.10 -.34*** 1 Note. ***p < .001; N = 223
Predictors for English Listening Comprehension
In order to answer the second research question, stepwise multiple regression analysis was applied, with English listening performance being the dependent variable,
40
English listening anxiety and tolerance of ambiguity being independent variables. As presented in Table 12 and Table 13, only listening anxiety was included as a significant predictor of listening performance (t = -8.52, p < .001, = -.50 ), and the value of the coefficient of determination (R2) suggests that 25% of the variance in English listening comprehension performance could be explained by one’s listening anxiety.
Table 13
Results of Stepwise Multiple Regressions of English Listening Ability on Listening Anxiety and Tolerance of Ambiguity
Model R R2 Adjusted R2 F p
Durbin-Watson
1 .50 .25 .24 72.61 .000 1.99
Note. Predictors: (Constant) Listening Anxiety; Dependent Variable: Listening score
Table 14
Analysis of Coefficients of Stepwise Multiple Regressions of English Listening Ability on Listening Anxiety and Tolerance of Ambiguity
Model B SE Beta t p Tolerance VIF
1 (constant) 46.76 1.77 26.47 .000
Listening Anxiety
-.61 .07 -.50 -8.52 .000 1.00 1.00
Note. Dependent Variable: Listening score
A further examination on the three aspects of listening anxiety revealed that, as demonstrated in Table 14 and Table 15, only cognitive (t = -3.01, p < .01) and behavioral dimensions (t = -9.52, p < .001) could explain one’s English listening performance, with the behavioral dimension being a better predictor ( = -.53) than the cognitive dimension ( = -.16). From the data in Table 14, the value of the coefficient of determination (R2) suggests that 36% of the variance in English listening
41
comprehension ability could be explained by the behavioral and cognitive dimensions of listening anxiety (F = 64.11, p < .001).
Table 15
Results of Stepwise Multiple Regressions of English Listening Ability on Listening Cognitive, Behavioral, and Somatic Anxiety
Model R R2 Adjusted R2 F p
Durbin-Watson
1a .58 .34 .33 114.95 .000
2b .60 .36 .36 64.11 .000 1.92
Note. Dependent Variable: Listening score
a. Predictors: (Constant) Behavioral Listening Anxiety
b. Predictors: (Constant) Behavioral Listening Anxiety, Cognitive Listening Anxiety.
Table 16
Analysis of Coefficients of Stepwise Multiple Regressions of English Listening Ability on Listening Cognitive, Behavioral, and Somatic Anxiety
Model B SE Beta t p Tolerance VIF
1 (constant) 41.15 .93 44.06 .000
Behavioral -2.22 .20 -.58 -10.72 .000 1.00 1.00
2 (constant) 45.48 1.70 26.67 .000
Behavioral -2.03 .21 -.53 -9.52 .000 .91 1.09
Cognitive -.68 .22 -.16 -3.01 .003 .91 1.09
Note. Dependent Variable: Listening score
Summary of Chapter Four
The results of the current study are summarized as follows. To begin with, listening was not considered the most difficult language skill by most participants in the current study, who self-reported having very little listening practice after class. For those who
42
had extra practice, online resources such as YouTube were best preferred as the material for listening practice. Based on the participants’ self-report, they appeared to have a moderate level of English listening anxiety and tolerance of ambiguity.
As for the results concerning the correlations among the variables, both English listening anxiety and tolerance of ambiguity significantly correlated to English performance. The correlation between English listening anxiety and English listening performance was negative and stronger than that between tolerance of ambiguity and English listening performance, which was in a positive direction.
While English listening anxiety, tolerance of ambiguity, and listening performance correlated with one another, the results showed that only English listening anxiety (especially its behavioral and cognitive dimensions) could serve as a significant predictor for one’s listening performance.
43