• 沒有找到結果。

This chapter consists of three parts, each involving detailed discussion on

findings of the present study. The first part is devoted to discussion on the positive

and significant correlation between metacognitive awareness and listening

comprehension. Next, possible explanation is offered for the insignificant difference

in metacognition growth between the two groups. Lastly, close analysis is carried out

to uncover the reason why learners receiving metacognitive approach made

significant improvement in listening comprehension.

Metacognitive Awareness as a Predictor for Listening Performance

The findings of the present study suggest that there is a positive, significant

correlation between learners’ metacognitive awareness and their listening

performance. In other words, the higher a learner’ metacognitive awareness is, the

better listening comprehension score s/he achieves. This positive, significant

correlation validated Schraw’s (1998) view: metacognitive skills is facilitative to

cognitive skills. Those who can use metacognitive strategies to regulate their own

learning and to compensate for comprehension gap will have better learning

performance.

In addition, the result indicates that about 12% of learners’ listening

63

comprehension performance can be explained by their metacognitive awareness. The

present study yielded similar result to Vandergrift et al.’s (2006) and Goh and Zeng’s

(2012) studies, where learners’ metacognitive awareness was accounted for 13% and

16% of learners’ listening performance respectively.

After confirming the significant correlation between metacognitive awareness

and listening comprehension, the researcher conducted further analysis of the five

factors underlying MALQ. Results show that two out of the five factors, Directed

attention and Person Knowledge, are significantly associated with listening

performance. It is conceivable to assume that the two have the most to do with

learners’ listening comprehension because of the nature of listening tests and learners’

perception of their current listening ability.

Among the five factors, Directed Attention correlates significantly with listening

performance because of its indispensible role in helping learners focus on essential

information during their listening process. In general, listening comprehension test is

constructed to examine learners’ ability to decipher information from aural input.

However, each listening encounter is transient and irreversible. Therefore, it is crucial

for learners to concentrate on the listening task in order to select and process key

information. Directed Attention, which is operationalized in terms of learners’ ability

to divert attention to key information in the listening text (Vandergrift, 1997), may

64

exert a powerful influence on their listening performance.

As for Person Knowledge, which reflects learners’ understanding of their

strengths and weaknesses, can possibly be an indicator of their actual performance. In

other words, the three items related to Person Knowledge had successfully elicit

learners’ perception of difficulty and anxiety of L2 listening. For example, “I feel that

listening comprehension in English is a challenge for me. (item 8)” “I don’t feel

nervous when I listen to English. (item 15)” The result provides empirical support to

Zhang and Goh (2006), whose findings suggested that learners’ perception of their

listening process positively correlates with their strategy knowledge.

In conclusion, findings in the present study indicated that metacognitive

awareness correlates positively to listening performance, with Directed Attention and

Person Knowledge being the major factors. What is more, although there may be

other powerful variables affecting listening comprehension, metacognitive awareness

was accounted for 12% of the variance in listening performance. As a result, it can be

regarded as one of the most important predictors for listening performance.

The Effect of Metacognitive Approach to Listening Instruction on Metacognitive

Awareness

Despite the fact that most existing studies on metacognitive pedagogical cycle

65

have demonstrated its positive effect on enhancing learners’ metacognitive awareness

(Cross, 2010; Goh, & Taib, 2006; Mareschal, 2007; Vandergrift & Tafaghodtari, 2010),

findings of the present study appear to be distinctly different. While learners in the

control group followed the traditional approach (i.e. listen and answer questions),

learners in the experimental group received eight metacognitive trainings on listening.

However, statistical analysis shows that there is no significant gain in metacognitive

awareness between the two groups. Moreover, the analysis of the five factor

underlying MALQ revealed no significant differences between the two groups, either.

The remarkable difference may result from the following three reasons: 1) the

duration of treatment, 2) the way instruction was given, and 3) the instrument used for

assessment.

First of all, there is a need for longer duration of treatment. It requires sufficient

time for learners to familiarize themselves with target strategies and to adopt theses

strategies for facilitating their learning process. As what Oxford (1994) contended,

strategy training should be carried out over a long period of time rather than being

completed with short invention.

Existing studies showing a rise in learners’ metacognitive awareness during the

listening process have dedicated considerable amount of time for treatment. Take the

following studies using the same instrument, MALQ, for example. Vandergrift and

66

Tafaghoti (2010) spent 13 weeks on the treatment, reporting that less proficient

learners had significant growth in their metacognitive awareness. Similarly, Teng and

Lin (2015) used 12 weeks on the treatment, concluding that learners in the

experimental group significantly outperformed those in the control group on the

MALQ posttest. As for researches whose treatment lasted for less than ten weeks, in

each treatment session, learners were given abundant time to go through the

metacognitive listening process. For example, Cross (2010), Movahed (2014), and

Zeng (2014) devoted 5 to 8 lessons for treatment respectively, but each of their

training session was 45 to 60 minutes long.

However, in the present study, given the tight schedule of high school learners’

regular courses, the 8 treatments were completed in haste, each lasting for roughly 20

minutes. The lack of time for treatment may be one of the major reasons why learners

receiving metacognitive approach showed no significant improvement in their

metacognitive awareness.

Second, in addition to lack of time, how metacognitive instruction was carried

out may influence its effect on learners’ metacognitive awareness. According to

Veenman et al. (2006) and Wenden (1991), metacognitive knowledge usually reflects

learners’ deep-rooted beliefs that are usually resistant to change. As what Brown et al.

(1983) suggested, metacognition belongs to statable knowledge, which is already

67

stored in learners’ long-term memory, different considerably from transient

knowledge, which emerges temporarily in their learning process.

Although researches have tried to “make mental processes explicit” (Goh, 2008,

p.200), it remains unclear that under what circumstances can explicit instruction have

significant effect on learners’ deep-rooted knowledge. Until now, researchers are still

exploring principles which make metacognitive instruction successful. For example,

Vandergrift (1999) promoted the use of listening comprehension checklists to raise

students’ awareness of their listening process. He also advocated the importance of

guided reflection after the completion of every listening task. By providing listeners

questions prompts like “Did it help you to listen for key words?” or “Could you

understand the message without understanding every word?”, learners will be aware

and reflect on their use of the key metacognitive strategies. Besides, Goh and Taib

(2006) recommended a variety of methods such as including reflection and discussion,

teacher modeling, integrated sequence with alternative focus on either text or process

to ensure the effectiveness of instruction. Also, Veenman et al. (2006) argued there are

three key principles that make metacognitive instruction effective: integration into the

subject matter, clearly informing learners about its usefulness, and prolonged training.

Result in the present study suggests that more studies are needed to find out how

to maximize the effect of explicit instruction on learners’ metacognitive knowledge.

68

Since research on metacognitive awareness about listening is relatively new, further

studies on metacognitive approach to listening should be conducted on how to help

learners develop “a habit of mind that is metacognitively alert and responsive to

learning” (Goh, 2008, p.200) with effective instruction.

Last but not least, using different instruments to assess metacognitive awareness

may yield different results. Although MALQ has been proven to be a valid, reliable

questionnaire (Vandergrift et. al., 2006) widely used for assessing metacognitive

awareness, changes in learners’ metacogntivie knowledge and regulation remains a

dynamic process (Flavell, 1979), which is too diverse and individualized to reflect on

the 21 items of MALQ.

To compensate for the insufficiency of quantitative method, several researchers

have included qualitative methods to explore different facets of learners’

metacognitive awareness. These qualitative methods may not be generalizable;

however, they can provide evidence for observable differences in learners’

metacognitive awareness. Researchers that had incorporated qualitative methods such

as reflective journal (Vandergrift, 2002, 2003), audio and video recording (Cross,

2009b), self report (Goh & Taib, 2006), and stimulated recall (Mareschal, 2007) have

uncovered supportive evidence of improved metacognitive awareness.

After the eight trainings of metacognitive pedagogical sequence, learners in the

69

experimental group made significant progress in their listening comprehension,

scoring five more points on average in the GEPT posttest. Since findings in the

present study show that metacognitive awareness positively and significantly

correlates with listening performance, learners’ gain in listening comprehension may

partially result from growth in metacognition. Moreover, analysis of the experimental

group’s notes provided evidence for learners’ continuous reflections upon the listening

process, suggesting that the metacognition of learners in the experimental group may

have changed to a certain extent. However, the changes might be too subtle to reach

the significant level based on responses from MALQ.

The Effect of Metacognitive Approach to Listening Instruction on Listening

Comprehension

Findings in the present research are consistent with the results of most of

previous research, showing that there is positive effect of metacognitive approach on

listening performance (Bozorigian, 2014; Goh & Taib, 2006; Vandergrift &

Tafaghodtari, 2010; Zeng, 2007, 2014) and that the gain of less proficient learners is

significantly higher than that of more proficient learners (Goh, 2008; Goh & Taib,

2006; Vandergrift, 1997; Vandergrift & Tafaghodtari, 2010).

The result of the present study has validated the positive effect of metacognitive

70

approach to listening in comparison with the listen-and

answer-comprehension-question convention in traditional listening classes. While

learners in the experimental group scored 5 more points on the GEPT posttest, the

performance of the control group remained nearly the same.

The difference may result from how metacognitive approach to listening change

the way learners approach a listening text (Goh, 2008). With explicit instruction and

repetitive practice in the metacognitive pedagogical sequence, learners became aware

of their listening process, developing ability to use strategies to regulate their listening

process. Taking control over their own listening process, learners started paving the

way for self-regulated learning (Wenden, 1998). These reflective, self-redirected

learners performed better on the listening test since metacognition was confirmed to

be one of the significant factors for listening performance (Al-Alwan et al., 2013; Goh

& Zeng, 2012; Vandergrift & Tafaghodtari, 2010).

Although metacognitive approach has a significant effect on learners’ listening

comprehension, analysis of MALQ shows no significant differences between the two

groups. In other words, in addition to changes in metacognitive awareness, there may

be other critical factors involved in enhancing learners’ listening performance.

To identify the underlying factors, further analysis on GEPT tests was performed,

reporting a significant difference in Short Dialogue section between the two groups.

71

On the GEPT posttest, the experimental group scored eight more points, which was

two times more than the gain of the control group.

The experimental group’s significant growth in Short Dialogue may shed light

on the positive effect of metacognitive approach to listening. According to LTTC’s

(1999) research on preliminary examination of intermediate level of GEPT, Short

Dialogue is the most difficult part in the listening test, with the lowest correct

response rate. It requires learners to demonstrate a variety of ability to reach

comprehension (see Table 26).

Table 26

Goals of Assessment for Different Sections in the Intermediate Level GEPT Listening Test

Section Text type Goals of assessment

I Picture

description

To assess learners’ listening ability to comprehend daily life vocabulary and high frequency patterns

II Question or statement response

To assess learners’ listening ability to comprehend daily life-related questions or direct speech and to respond properly

III

Short dialogue

To assess learners’ listening ability to comprehend daily life conversations. Learners have to demonstrate their ability to a. grasp the topic and main idea of the dialogue

b. figure out the context of the dialogue, for example, the purpose, the location, and the relationship between the interlocutors

c. comprehend the important details in the dialogue

d. be able to make inferences, read between the lines, and make predictions based on the information obtained

Note. Adapted from LTTC (1999)

72

Aside from enhancing learners’ ability to regulate their listening, metacognitive

approach’s primary focus on “process” rather than “product” can possibly account for

the difference. Compared with product-oriented conventional approach in listening

classrooms, the process-based metacognitive approach engaged learners in situations

similar to real-life listening─“where there is no specific purpose for listening other

than to understand the main idea and as much supporting detail as possible

(Vandergrift & Tafaghodtari, 2010, p.488).”

It’s common for learners who have to answer comprehension questions to put

heavy emphasis on the outcome, focusing exclusively on key information for getting

the questions correct. In contrast, learners in the experimental group have been used

to reflecting upon the whole listening process for text reconstruction, making

endeavor to comprehend as much information as possible (Goh & Taib, 2006). Thus,

the obtained information with rich details as well as main idea can successfully help

them figure out the context and make inferences in the relatively challenging Short

Dialogue section.

In addition to providing empirical support for metacognitive approach’s positive

effect on listening comprehension, the present study validated findings in previous

studies that less proficient learners benefited more than the more proficient ones from

this approach (Goh & Taib, 2006; Vandergrift, 1997; Vandergrift and Tafaghodtari,

73

2010). In the present study, there were 37 students in the experimental group, with 17

more proficient learners and 20 less proficient learners based on GEPT pretest result.

Results show that while the more proficient ones gained 2 more points, the less

proficient learners scored 8 more points on the GEPT posttest.

According to Oxford (2002), researchers have to explore the strategies used by

language learners, and then teach these strategies to novice learners to facilitate their

language learning. Since more proficient learners are generally the ones who can use

metacognitive strategies for comprehension (Bacon, 1992; O’Malley & Chamot, 1990;

Vandergrift, 1997), they may not benefit as much as the less proficient ones from the

metacognitive approach. As for the less proficient learners, they profited considerably

from the strategy instruction because they were guided to experience the

comprehension process of their more proficient peers. Adopting these metacognitive

strategies employed by their more proficient peers, the less proficient learners not

only became “better listeners” but “effective language learners” (Nunan, 2004, p.240).

As what Anderson (2002) contended, “Strong metacognition empowers second

language learners” (p.5). To sum up, findings in the present study suggest that

metacognitive approach to listening has a positive and significant effect on learners’

listening performance. Moreover, the particular focus on process rather than product

may potentially influence the way learners approach the listening text, enabling them

74

to make better inferences based on the obtained information. Last but not least,

metacognitive approach appears to be more beneficial to less skill learners, helping

them to move closer to “ the profile of “a good language learner.’” (Oxford, 2002,

p.129).

75

相關文件