As shown in Table 1, means and standard deviations of all the participants for the STAI (A-State) at the anticipation, confrontation, and release stages were, 58.61 (11.540), 64.30 (11.129), and 50.80 (14.328) respectively.
4.2 Different Levels of Trait Anxiety and State Anxiety Narrow-banded for the Three Stages
According to the ascending order of the sum of the A-Trait scale, the participants should be divided into an equal number of high-trait-anxious and
low-trait-anxious groups at the median score. However, since there were two students who got the same scores at the median score, they were both classified as
high-trait-anxious speakers. As a result, there were 34 students in the high-trait-anxious group and 32 students in the low-trait-anxious group.
Table 2
Means and Standard Deviations of the High-trait-anxious Group and the
Low-trait-anxious Group for the STAI (A-State) at the Three Narrow-banded Stages
95% Confidence
As can be seen in Table 2, for high-trait-anxious speakers, means and standard deviations for STAI (A-State) at the anticipation, confrontation, and release stages were, 66.06 (8.279), 68.03 (10.630), and 54.44 (13.661) respectively. Means and
standard deviations of low-trait-anxious speakers for STAI (A-State) at the
anticipation, confrontation, and release stages were, 50.69 (8.978), 60.34 (10.391), and 46.94 (14.208) respectively.
In order to know if the differences of state anxiety narrow-banded for the three stages between the two trait-anxious groups show significance, a 3 × 2 factorial ANOVA was used, as can be seen in Table 3 and Table 4.
Table 3
Tests of the Effect of State Anxiety Stages and the Interaction Effect Between Levels of Trait Anticipatory Anxiety and State Anxiety Stages
Source Type III Sum of
Squares df Mean Square F Sig.
Sphericity Assumed 6044.720 2 3022.360 34.300 .000 Greenhouse-Geisser 6044.720 1.600 3777.074 34.300 .000
Huynh-Feldt 6044.720 1.661 3639.953 34.300 .000
State Anxiety Stages
Lower-bond 6044.720 1.000 6044.720 34.300 .000
Sphericity Assumed 664.902 2 332.451 3.773 .026
Greenhouse-Geisser 664.902 1.600 415.467 3.773 .035
Huynh-Feldt 664.902 1.661 400.384 3.773 .034
State Anxiety Stages * Anticipatory
Trait Anxiety Lower-bond 664.902 1.000 664.902 3.773 .056
Sphericity Assumed 11278.724 128 88.115 Greenhouse-Geisser 11278.724 102.424 110.118
Huynh-Feldt 11278.724 106.282 106.121 Error (State
Anxiety Stages)
Lower-bond 11278.724 64.000 176.230
Table 4
Test of the Effect of Levels of Anticipatory Trait Anxiety
Source Type III Sum of
Squares
df Mean Square F Sig.
Intercept 659729.000 1 659729.000 3254.285 .000
Anticipatory Trait Anxiety 5132.030 1 5132.030 25.315 .000
Error 12974.480 64 202.726
First, as shown in Table 3, it revealed that the effect of state anxiety stages
showed significance (df = 1.600; F=34.300, P=.000＜.05). Second, an interaction effect was found between levels of anticipatory trait anxiety and state anxiety stages (df=1.600; F=3.773, P=.035＜.05), as can be seen in Table 3. Third, the effect of levels of anticipatory trait anxiety showed significance (df=1; F=25.315, P=.000
＜.05), as shown in Table 4.
Since the interaction effect showed significance, it is necessary to respectively test the main effect of each of the two variables. As a result, for the main effect of different levels of anticipatory trait anxiety, one-way ANOVA was measured three times for the three stages, as can be seen in Table 5.
Table 5
Test of the Main Effect of Different Levels of Anticipatory Trait Anxiety
Sum of
Within Groups 7076.189 64 110.565 Confrontation
Total 8049.939 65
Between
Groups 928.182 1 928.182 4.784 .032
Within Groups 12416.257 64 194.004 Release
Total 13344.439 65
As shown in Table 5, high-trait-anxious speakers were found to show higher levels of state anxiety than low-trait-anxious speakers at the anticipation (df=1;
F=52.361, P=.000＜.05), confrontation (df=1; F=8.807, P=.004＜.05), and release stages (df=1; F= 4.784, P=.032＜.05).
On the other hand, for the main effect of state anxiety stages, one-way ANOVA
for dependent samples was tested two times for high-trait-anxious speakers and low-trait-anxious speakers respectively, as can be seen in Table 6, Table 7.
Table 6
Test of the Main Effect of State Anxiety Stages for High-trait-anxious Speakers
95% Confidence Interval for
Lower Bound Lower Bound
Confrontation -1.971 1.118 .087 -4.246 .304
Note. Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons：Least Significant Difference (equivalent no adjustments).
b. Anticipatory Trait Anxiety = high-trait-anxious speakers
As shown in Table 6, high-trait-anxious speakers were found to show higher level of state anxiety at the anticipation and confrontation stages than at the release stage although no significance was detected between the anticipation and
confrontation stages.
Table 7
Test of the Main Effect of State Anxiety Stages for Low-trait-anxious Speakers
95% Confidence Interval for Differencea (I) timing (J) timing Mean
Difference (I-J) Std. Error Sig. a
Lower Bound Lower Bound
Confrontation -9.656* 2.164 .000 -14.070 -5.243
Note. Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons：Least Significant Difference (equivalent no adjustments).
b. Anticipatory Trait Anxiety = low-trait-anxious speakers
Low-trait-anxious speakers were found to have their highest level of state anxiety at the confrontation stage, but no significance was found between the anticipation and release stages, as can be seen in Table 7.
Figure 1 shows the comparisons of the state anxiety narrow-banded for the three stages between high-trait-anxious and low-trait-anxious groups.
Figure 1. The distributions of state anxiety for high-trait-anxious speakers and
low-trait-anxious speakers at the three narrow-banded stages.Although high-trait-anxious speakers showed higher levels of state anxiety at all the three narrow-banded stages, the two groups revealed similar patterns of state anxiety during the three stages, as shown in Figure 1.
4.3 Gender Difference and State Anxiety Narrow-banded for the Three Stages There were 30 males and 36 females participating in this study. As for the male group, it consisted of 11 seventh graders, 15 eighth graders, and four ninth graders.
The female group comprised 18 seventh graders, 13 eighth graders, and five ninth
graders.
Table 8
Means and Standard Deviations of the Male Group and the Female Group for the STAI (A-State) at the Three Narrow-banded Stages
Gender Mean Std. Deviation N
As can be seen in Table 8, for male speakers, means and standard deviations for STAI (A-State) at the anticipation, confrontation, and release stages were, 55.70 (12.570), 63.13 (12.439), and 49.43(14.668) respectively. Means and standard deviations of female speakers for STAI (A-State) at the anticipation, confrontation, and release stages were, 61.03 (10.154), 65.28 (9.982), and 51.94 (14.144)
respectively.
Table 9
Tests of the Effect of State Anxiety Stages and the Interaction Effect Between Gender Difference and State Anxiety Stages
Source Type III Sum
of Squares
df Mean Square F Sig.
Sphericity Assumed 6015.948 2 3007.974 32.507 .000 Greenhouse-Geisser 6015.948 1.631 3688.043 32.507 .000
Huynh-Feldt 6015.948 1.694 3551.103 32.507 .000
State Anxiety Stages
Lower-bond 6015.948 1.000 6015.948 32.507 .000
Sphericity Assumed 99.282 2 49.641 .536 .586
Greenhouse-Geisser 99.282 1.631 60.864 .536 .550
Huynh-Feldt 99.282 1.694 58.604 .536 .557
State Anxiety Stages * Gender
Difference
Lower-bond 99.282 1.000 99.282 .536 .467
Sphericity Assumed 11844.344 128 92.534 Greenhouse-Geisser 11844.344 104.397 113.455 Huynh-Feldt 11844.344 108.423 109.242 Error (State
Anxiety Stages)
Lower-bond 11844.344 64.000 185.068
Table 10
Test of the Effect of Gender Difference
Source Type III Sum of Squares df Mean Square F Sig.
Intercept 654948.002 1 654948.002 2386.664 .000
Gender
Difference 543.638 1 543.638 1.981 .164
Error 17562.872 64 274.420
In order to know whether the differences of state anxiety narrow-banded for the three stages between different genders show significance, a 3 × 2 factorial ANOVA was measured, as can be seen in Table 9 and Table 10.
First, as shown in Table 9, it indicated that the effect of state anxiety stages showed significance (df = 1.631; F=32.507, P=.000＜.05). Second, no interaction effect was found between gender difference and state anxiety stages (df=1.631;
F=.536, P=.550＞.05), as can be seen in Table 9. Third, the effect of gender difference also showed no significance (df=1; F=1.981, P=.164＞.05), as shown in Table 10.
Table 11
Test of the Main Effect of State Anxiety Stages for the Male Group
95% Confidence Interval for Differencea (I) timing (J) timing Mean Difference
(I-J) Std. Error Sig. a
Lower Bound Upper Bound
Confrontation -7.433* 2.207 .002 -11.948 -2.919
Note. Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons：Least Significant Difference (equivalent no adjustments).
b. Gender = male
Since no interaction effect was found between the two variables and merely the main effect of state anxiety stages showed significance, one-way ANOVA for
dependent samples was tested two times for females and males respectively, as shown in Table 11 and Table 12.
As can be seen in Table 11, male speakers were found to show their highest level of state anxiety at the confrontation stage, but no significance was detected between the anticipation and release stages.
Table 12
Test of the Main Effect of State Anxiety Stages for the Female Group
95% Confidence Interval for Differencea (I) timing (J) timing Mean Difference
(I-J) Std. Error Sig. a
Lower Bound Upper Bound
Confrontation -4.250* 1.441 .006 -7.176 -1.324
Note. Based on estimated marginal means
*. The mean difference is significant at the .05 level.
a. Adjustment for multiple comparisons：Least Significant Difference (equivalent no adjustments).
b. Gender = female
Female speakers were found to have their highest state anxiety at the confrontation stage, second at the anticipation stage, and the lowest at the release stage, as shown in Table 12. The differences of state anxiety at the three stages within the female group were all significant.
Figure 2 displayed the comparisons of the distributions of state anxiety at the three narrow-banded stages based on gender difference. There is difference and similarity between the two state anxiety patterns at the three stages which were presented by the two sexes.
Figure 2. The distributions of state anxiety for male speakers and female
speakers at the three narrow-banded stages.As shown in Figure 2, both males and females showed similar patterns of state anxiety narrow-banded for the three stages, with females’ state anxiety higher than that of males’ at all the three stages.
4.4 Summary
There were three main findings in this study. First, the overall means and standard deviations of state anxiety narrow-banded for the three stages showed the sensitization pattern. That is, all the participants tended to have their state anxiety rise at the pre-speaking period, to climax at the confrontation stage, and then to decline in the last stage. Second, as for the relationship between different levels of trait anxiety and state anxiety narrow-banded for the three stages, high-trait-anxious speakers’
levels of state anxiety were all significantly higher than those of low-trait-anxious speakers at the three stages. Besides, both high- and low-trait anxious speakers showed the same sensitization pattern. Third, males and females also revealed the
sensitization pattern. However, no significant difference was showed based on gender difference even though females’ levels of state anxiety was higher than those of males’ at the three stages.
CHAPTER FIVE
DISSCUSSION AND CONCLUSION 5.1 Discussion
This section includes the discussion of the overall means of the state anxiety narrow-banded for the three stages, the comparison between the high-trait-anxious group and low-trait-anxious group, and the comparison between the male group and female group.
5.1.1 The overall means of the state anxiety narrow-banded for the three stages Comparing the overall means of state anxiety narrow-banded for various stages in the two studies, this study revealed the sensitization pattern (58.61, 64.30, and 50.80) while Lewis et al.’s study (2007) showed the habituation pattern (56.01, 53.34, 44.44, and 37.27). That is, in this study, all the participants’ state anxiety tended to rise at the anticipation stage, climax at the confrontation stage, and decline at the release stage. On the other hand, in Lewis et al.’s study (2007), their participants tended to have their highest state anxiety at the anticipation stage and then to descend in the following stages.
According to previous studies (Lewis et al., 2007; Sawyer & Behnke, 2002), the overall levels of anxiety in the sensitization pattern were found to be higher than those in the habituation pattern. Sawyer and Behnke (2002) also indicated that the sensitization pattern occurs when speakers are confronting a strong stimulus while a weak stimulus tends to produce the habituation pattern. Besides, the contexts of the two studies may make a difference to the results. First, the participants in Lewis et al.’s study were undergraduate students who spoke English as a native or second language while the participants of this study were EFL learners who studied English in a junior high school. Second, Lewis et al. conducted their study in an informal speech while this study was examined in a formal English speech contest which might
arouse more stress for the speakers. As a result, the contradictory results between the two studies suggest that the task in Lewis et al.’s study (2007) seems to be an easier one for their participants while the task in this study offers more challenge for these participants.
Additionally, the majority of native or ESL learners were found to show the habituation pattern since their learning environment provided them enough support to overcome their difficulty and to gradually help reduce their anxiety (Sawyer &
Behnke, 2002). The overall sensitization pattern in this study suggests that most students seem to need more pedagogical guidance from the teachers, especially at the confrontation stage.
5.1.2 High-trait-anxious group and low-trait-anxious group
The results of this study indicate that there is difference and similarity between the two trait anxious groups in their state anxiety narrow-banded for the three stages.
First, the two trait anxious groups differ in their magnitude of state anxiety at the three stages. That is, high-trait-anxious speakers’ state anxiety at the three stages (66.06, 68.03, and 54.44) was all higher than that of the low-trait-anxious group (50.69, 60.34, and 46.94). This finding is similar to that in Lewis et al.’s study (2007).
Moreover, likeness is shown in the state anxiety patterns of the two trait anxious groups at the three stages. Compared to Lewis et al.’s study (2007), which showed the habituation pattern in the high-trait-anxious group and the sensitization pattern in the low-trait-anxious group, the sensitization pattern was found in both groups in this study. The difference may result from the different research procedures in the two studies. In Lewis et al.’s study (2007), the experiment was conducted in an informal speech without any preparation time immediately before the presentation.
Nevertheless, this study was done in a formal English speech contest, with 20 minutes for all students to prepare for their speech right before standing on the stage. Since
preparation has been reported as a means to reduce students’ anxiety (Lewis et al., 2007), it may be the time for preparing for the speech that provides all the students in this study some time to think of strategies to deal with their anxiety. As a result, both the high-trait-anxious and low-trait-anxious speakers seem to become calmer at the anticipation stage and have their highest anxiety at the confrontation stage, resulting in the sensitization pattern.
5.1.3 Male group and female group
In this study, both genders revealed the same sensitization pattern. Besides, no significant difference was found among the three stages based on gender difference although females were found to exhibit higher levels of anxiety at all stages than males. These findings are half consistent with Behnke and Sawyer’s study (2000), in which the two sexes revealed the same psychological pattern, with females’ anxiety levels significantly higher than those of males at all stages.
In fact, there were some differences between Behnke and Sawyer’s study and this study. First, this study examined the three stages: anticipation, confrontation, and release stages. Behnke and Sawyer’s study (2007), however, observed three different stages during a public speech: That is, the moment when the public speech was
assigned in class, the moment during preparation, and the moment immediately before the speech. Second, this study was conducted in a formal English speech contest while the task in Behnke and Sawyer’s study (2000) was an informal speech. Third, the participants of this study were EFL learners at the junior high school level. The participants in their study, on the other hand, were undergraduates who spoke English as a native or second language.
Thus, the reasons why gender difference showed on significance in the levels of state anxiety in this study may result from the above differences. That is, this study was conducted in a formal speech contest in Taiwan, which was a foreign language
environment. The foreign language context and the task of the English speech contest in this study would probably offer much difficulty and the same pressure for both male and female students. As a result, there was no significant difference showed based on gender difference.
5.2 Conclusion
This section contains the summary of findings in this study, some pedagogical implications from previous studies, this researcher’s insights from this study, some limitations of this study, and several suggestions for future studies.
5.2.1 Summary of findings
The result of the present study shows that only the sensitization pattern was found in the state anxiety distribution narrow-banded for the three stages, regardless of different levels of anticipatory trait anxiety or gender difference. That is, the speakers in each group all tended to ascend their anxiety at the anticipation stage, with their anxiety peaking at the confrontation stage and then descending at the release stage. This finding suggests that the preparation process in the English contest may help all students lower their anxiety before presenting their speeches. Therefore, their anxiety was relatively lower at the pre-speaking period, and they were calmer to come up with strategies to conquer their anxiety. On the other hand, the overall means of the state anxiety narrow-banded for the three stages also revealed the sensitization pattern. This result implies that the speech task in this study seems to be a strong stimulus for most speakers. Thus, their first underestimated anxiety at the anticipation stage started to rise to the climax at the confrontation stage when they were faced with the audience. This result also indicates that the majority of speakers in this study may need more help from their teachers, especially when they are first confronted with the audience.
5.2.2 Pedagogical implications
In fact, many pedagogical strategies have been developed in order to promote the habituation effect and minimize the sensitization effect (Behnke & Sawyer 2004;
Sawyer & Behnke, 2002). According to Whitworth and Cochran (1996), behavioral intervention or therapy, cognitive based intervention, and communicative skills trainings were the three orientations for the treatments of public speaking anxiety. In their study, systematic desensitization, cognitive modification, and skills training respectively represented for each of the orientations and were indicated as the three most popular treatments of public speaking anxiety (Whitworth & Cochran, 1996).
They found that the speakers treated with the combined treatments were found to have significantly reductions in public speaking anxiety than those receiving one of the treatments alone. Sawyer and Behnke (2002) also suggested that exposure therapy, which is synonymous with skills training, might be combined with traditional
cognitive-behavioral method to help reduce students’ speaking anxiety. Besides, strategies, for example, such as having students present their speeches to small groups of audience (Behnke & Sawyer, 2004), offering positive feedback to students in the beginning of their speech performance (Behnke & Sawyer, 2004), or “arranging speaking assignments in ascending order of complexity or difficulty” (Sawyer &
Behnke, 2002, p.117) are all useful and practical for anxiety reduction.
5.2.3 Insights from the study
There are some insights which this researcher has gained from this research.
First, the result in this study that the students truly experienced the fluctuations of state anxiety during a public speech, reminds her of the influence of public speaking anxiety. To help her students perform well in their speeches in the future, she should pay more attention to how to reduce students’ anxiety during a public speech. Second, this study showed that the students still received much challenge from the context of
the formal English contest and the foreign language environment in Taiwan as well.
In order to lower students’ anxiety, this researcher can not only offer more tasks with different levels of difficulty for her future students but also can arrange these tasks in an order from the easier ones to the more difficult ones. Third, this study also proved that preparation may be a useful way to help reduce students’ anxiety. This result reminds the researcher to provide some preparation time for her students before giving their speech in their future training process. Besides, it is also important that pedagogical interventions need to be implemented at the right time. For the students in this study, the first minute of the speech contest seems to be the most
anxiety-triggering moment, at which they probably need their teachers’ support most.
This result suggests that this researcher should provide some strategies for her students to deal with their highest anxiety when they are first facing the audience.
5.2.4 Limitations of the study
Although this researcher has taken this study under careful inspections, there were still some limitations to the research. First, since this was only a small-scale
Although this researcher has taken this study under careful inspections, there were still some limitations to the research. First, since this was only a small-scale