Chapter 1 Introduction
1.5 Organization of the Study
The thesis is organized as follows. Chapter One provides the background and
motivation of the present study on young learners’ listening proficiency in a
communicative test with authentic input from NESTs. Chapter Two reviews relevant
papers to the present research topic including (a) general backgrounds of listening
comprehension and authenticity, (b) the effectiveness of English-native teachers’
authentic input in a co-teaching module, and (c) three primary approaches in listening
assessment. Chapter Three describes the details of the methodology in the present study.
Results and discussions are presented in Chapter Four. Finally, Chapter Five
summarizes and concludes the major findings. Limitations and suggestions for future
research are also provided at the end of the chapter.
7
Chapter 2
Literature Review
The present study investigated the influence of NESTs on younger learners’
listening performance on communicative tests. To explore the stated research questions,
the relevant literature reviewed in this chapter is separated into three sections: Section
2.1 summarizes the general backgrounds of listening comprehension and focuses on the
aspect of authenticity; Section 2.2 explores the effectiveness of English-native teachers’
authentic input in a co-teaching module; Section 2.3 reviews listening assessment with
three primary approaches to language testing; and finally, major findings and
limitations from previous studies will be summarized in Section 2.4.
2.1 Listening Comprehension
2.1.1. Listening Comprehension Process
Listening comprehension has been under great interest and been investigated from
many aspects for decades among educational researchers. Most researchers in recent
years focus on the cognitive aspect of listening comprehension, which mainly examines
the process of the comprehension (e.g., Kurita, 2012; O’Malley & Küpper, 1989; Rost,
1990; Rubin, 1994; Vandergrift, 1999, 2007). There are two commonly adopted models
of listening comprehension: Anderson’s (1985) three stages of language comprehension
and top-down processing versus bottom-up processing. Anderson (1985) proposed that
8
comprehension is an interrelated and recursive process which involves perceptual
processing, parsing, and utilization. In the first stage, perceptual processing, attention
is focused on the encoding of the spoken message. In the second stage, parsing, words
and messages are utilized to build up meaningful mental representations. The third stage
is utilization, in which the listener relates the mental representation to the existing
knowledge and it is the stage where comprehension takes place if the mental
representation and the existing knowledge are matched.
Based on the aforementioned listening comprehension model, O’Malley and
Küpper (1989) conducted an empirical study to examine whether L2 learners process
information as proposed by Anderson and the strategies used while comprehending
academic texts with a group of high school students in an ESL context using “think
aloud” protocol. The participants were asked to listen to different types of listening
passages including lectures and short stories and think aloud in either Spanish or
English to explain how they understood the listening passages. The result showed that
the listening comprehension presented by the participants was consistent with the three
stages proposed by Anderson. Furthermore, the strategies used by effective and
ineffective listeners were different and were applied in various ways depending on the
phase of comprehension.
9
Another model of listening comprehension deals with the idea of top-down
processing and bottom-up processing. Top-down processing refers to the application of
background knowledge in understanding the meaning of a message, while in the
bottom-up processing, the understanding of each of the linguistic characteristics from
the incoming input is used as the basis for understanding the whole message. It is
generally agreed by many researchers that listening comprehension occurs when both
top-down and bottom-up processes continuously interact (Vandergrift, 2007). For
example, Chien and Wei (1998) investigated the strategies used by different proficiency
levels of college EFL learners in listening comprehension with two sets of authentic
listening materials in recall tasks. The result showed that to match the new information
with what is already known (top-down processing) and to infer meaning between words
or phrases (bottom-up processing) are two most effective cognitive strategies for
understanding the aural input. Chang and Read (2006) also evaluated the influence of
listening strategies on the listening performance of college EFL learners. The listening
comprehension test they applied was a section of TOEIC listening comprehension test
which was recorded by native-speaking ESL teachers with diverse accents. They found
that the most effective type of support was providing information about the topic, which
is more of a top-down processing; vocabulary instruction was the least useful form of
support, which is more of a bottom-up processing.
10
Despite the fact that all the empirical studies mentioned above explored listening
comprehension strategies used by learners with different tests or tasks and put much
emphasis on the process of comprehension, the listening constructs, that is, the
definition of listening competence and the listening tasks, in the aforementioned studies
were neglected. As Chapelle (1999) pointed out that there are two ways to define a
construct: (1) to define the competence, that is, the abilities that we believe the
test-takers should acquire and (2) to define the tasks that we think test-test-takers should perform.
However, either way was applied in the abovementioned studies. When using listening
tests to conduct research, the construct validity should be of primary concern
(Vongpumivitch, 2007). In terms of the listening assessment, further literature review
will be presented in the later section. Apart from the listening constructs, the materials
that learners process during the comprehension in both the empirical studies and the
real-life situations are also of importance considered by many researchers (e.g., Buck,
1995; Field, 2010; Goh, 2002; Rubin, 1994; Vandergrift, 2004) which were also
overlooked in the above-reviewed empirical studies.
2.1.2. Authentic Input for Listening Comprehension
Despite the fact that the major focus of the recent listening comprehension studies
has been laid on the process of the comprehension, there is a body of research on
listening comprehension about how learners interact with authentic input (Rubin, 1994).
11
Some researchers examined the text characteristics such as the text type of the listening
passage (e.g., Shohamy & Inbar, 1991), the acoustic variables (i.e. speech rate) of the
authentic input (e.g., Griffiths, 1991); others from the aspect of interlocutor
characteristics such as the gender of the speaker (e.g., Markham, 1988).
The term “authentic” was defined by Rogers and Medley (1988, p. 468) as
“language samples that reflect a naturalness of form, and an appropriateness of cultural
and situational context that would be found in the language as used by native speakers.”
As Vandergrift (2007, p. 200) pointed out that “exposure to authentic-type texts and
natural speech rate is preferred by L2 learners and can be beneficial for listening
development.” Such statement coincides with Rogers and Medley’s (1988) study that
they believed students should experience the language which is used for real
communication by native speakers. It is necessary to incorporate authentic materials in
the activities developed for language learning. After all, the ultimate goal for language
learning is to facilitate L2 learners to comprehend the target language in real-life
situations (Rogers & Medley, 1988). This might also explain why all the empirical
studies reviewed adopt authentic listening materials when examining listening
comprehension.
However, not all studies present such point of view. For example, Teng (2001)
examined the effects of modified sentences and the speech rate of listening texts on the
12
listening comprehension of a group of 168 EFL freshmen in college. The results showed
that students performed better when they listened to the tests that were syntactically
modified. Besides, students performed better in the slower version of the input instead
of the original version. The researcher concluded that the use of modified listening
passages should be applied in the listening tests. It seems that mixed statements have
been claimed considering the use of authentic or modified listening materials when
examining listening comprehension. Thus, the present study intends to investigate how
the authentic input of NESTs in a co-teaching model influences younger learners’
listening proficiency.
2.2 Co-teaching Native-English-Speaking Teachers (NESTs) and
Non-Native-English-Speaking Teachers (NNESTs)
As the effectiveness of NESTs as an authentic input in language learning
classrooms has long been a controversial issue, the advantages and disadvantages of
NESTs and NNESTs have been under great interest for decades among educational
researchers (e.g., Barratt & Kontra, 2000; Benke & Medgyes, 2005; Florence Ma, 2012;
Herbert & Wu, 2009; Medgyes, 1992). The findings suggested that NESTs and NNESTs
have their own strengths and weakness. NESTs are valued for their authenticity and
positive personal traits (Barratt & Kontra, 2000). They are good at teaching
conversation classes, providing perfect linguistic models, getting learners to speak,
13
being friendly (Benke & Medgyes, 2005) and creating relaxing and lively classroom
atmosphere (Poon & Higginbottom, 2000). However, NESTs are mostly criticized for
being unable to identify language learner’s common problems (Barratt & Kontra, 2000),
lack of expertise in English language teaching (Chang, 2007), and their
non-examination-oriented teaching styles (Florence Ma, 2012). In contrast to NESTs,
findings pointed out that NNESTs are appreciated for understanding students’ learning
difficulties and needs (Florence Ma, 2012) and serve as models of successful language
learners (Medgyes, 1992). Nonetheless, it is suggested that they are usually poor in
pronunciation and tend to have traditional and textbook-based teaching styles (Medgyes,
1992). In short, it seems that NESTs and NNESTs make up for each other’s
disadvantages, which “make both groups of teachers serve equally useful purposes in
their own terms” (Medgyes, 1992, p. 349). Thus, having NESTs and NNESTs to
co-teach can help consolidate their qualities and make a learning environment which
benefits students’ English learning.
Co-teaching is defined as “two or more professionals delivering substantive
instruction to a diverse, or blended, group of students in a single physical space” (Cook
& Friend, 1995, p. 2). It is a commonly applied teaching model in East Asian native
English teacher programs, including Hong Kong NET (Native-speaking English
Teacher) scheme, EPIK (English Program in Korea) program, Japanese JET (the Japan
14
Exchange and Teaching) scheme, and etc. Taiwan, as well, utilized co-teaching in
different programs which include NESTs, such as MOE’s FET (Foreign English
Teachers) program, Fulbright’s ETA program, King Car’s Schweitzer Program, etc. As
findings suggested that co-teaching provides more opportunities for students to use the
target language (Barratt & Kontra, 2000; Carless & Walker, 2006), creates a more
pleasant learning environment (Buckley, 2000), and helps facilitate learners with low
English proficiency (Storey et al., 2001).
Although co-teaching seems to help facilitate learners with low English
proficiency, studies directly assessing the effects of authentic input provided by the
NESTs in a co-teaching model are still very limited (e.g., Lin, 2017; Wu, 2015). Wu
(2015) explored the effectiveness of different co-teaching models between NESTs and
NNESTs on sixth graders’ English listening and speaking performances in an EFL
context. There were three types of co-teaching in the study: (1) NEST dominated the
teaching while NNEST facilitated, (2) station-teaching, and (3) team-teaching. The
participants were given a pre-test and followed by a post-test after 10 weeks of different
co-teaching instruction. The result showed that NEST dominated the teaching while
NNEST facilitated and station-teaching both had significant effects on students’
listening and speaking performances. The team-teaching model had a greater effect on
students’ speaking abilities; however, it only had a limited effect on students’ listening
15
abilities. In general, NESTs in co-teaching models have a positive effect on language
learners. Lin’s (2017) study also supported such a finding. The study investigated the
influence of NESTs and NNESTs’ co-teaching on Taiwanese students’ listening
comprehension. The results indicated co-teaching by NESTs and NNESTs greatly
impacted young EFL learners’ listening. Such co-teaching influenced not only learners’
listening comprehension, but also their performance on different questions types such
as those requiring English listening and reading at the same time. Learners with longer
co-teaching experience performed better on question types requiring English listening
and reading. In short, it seemed that the authentic input provided by the NESTs did have
a positive effect on learners in a co-teaching model.
2.3 Listening Assessment
Listening comprehension is a complex and active process (Vandergrift, 1999). A
number of theories have been established to explain the construct for such process and
researchers has been turning these theoretical notions into actual practice, that is, the
listening tests. According to Buck (2001), there are three major approaches to language
testing: the discrete-point, integrative and communicative approaches. The approaches
will be reviewed in the following sub-sections.
2.3.1 Discrete-point Testing
Discrete-point testing is based on the theory proposed by Lado (1961). He assumed
16
that it is possible to identify the isolated elements of language (i.e., grammar,
vocabulary, and pronunciation) and to test each of these items separately. He believed
that by testing a representative item of the language can estimate learners’ language
proficiency. Typical discrete-point tests include phonemic discrimination tasks,
response evaluation tasks, and paraphrase recognition tasks (Buck, 2001).
However, the opponents of the discrete-point tests pointed out that the actual
function in communication might not be evaluated through discrete-point testing (e.g.,
Carroll, 1961; Oller, 1975). Oller (1975) claimed that the assessment of the individual
items in a language is no better than the language as a whole for an item cannot
represent the overall proficiency of a learner’s language ability. Both Carroll and Oller
advocated that integrative test is a better approach to assessing language learning.
2.3.2 Integrative Testing
In contrast to discrete-point tests, integrative tests are intended to evaluate L2
learners’ language abilities as a whole. Oller (1973, p. 37) explained that “whereas
discrete items attempt to test knowledge of language one bit at a time, integrative tests
attempt to assess a learner’s capacity to use many bits all at the same time.” The
theoretical notion associated with integrative testing is that listening is to process texts
in real time, that is, to interpret the semantic meaning. Some well-known integrative
tests are the close test, dictation, and gap-filling (Buck, 2011). The problem with
17
integrative tests is somewhat similar to discrete-point tests in the way that both types
of the test tend to evaluate language as an individualized item as in discrete-point tests
or as an isolated event as in integrative tests. When language processing is considered
to be an isolated event, learners merely get the fundamental linguistic information in
the message without relating to the context.
2.3.3 Communicative Testing
With the rise of communicative teaching, communicative testing was developed
in response to the trend. Communicative testing is based on the theoretical notion that
language proficiency is seen as communicative competence (Hymes, 1972). That is, the
test should be less concerned with how much a test-taker knows about the language,
but more about how he or she can apply such language to communicate. According to
Widdowson (1978), communicative tests should test the use of the language for its
communicative function, rather than usage. Carroll (1980) also pointed out that the use
of language is the objective while the mastery of the usage of that language is a means
to achieve that objective. In communicative tests, it is important to identify whether
learners can use the target language to communicate in real-life situation instead of
knowing the correct usage of the grammar. In other words, a communicative test tests
the “use of language in ordinary situations” (Morrow, 1979, p 148).
The most salient characteristics of communicative tests include (1) authentic texts
18
(2) communicative purpose and (3) the authentic tasks used in the tests. In terms of
authentic text, Brindley (1998, p. 175) suggested that the test should include
conversations, narratives, directions and talkback exchanges. For example, Buck (1990)
conducted a research study which used several communicative tests. In one of the tests,
students were told to imagine that they were going to England for a vacation, and they
called the Tourist Information center to ask for the tours in London. Students listened
to the man who worked at the center and based on the passage to fill in the missing
information, a total of 12 questions, in the grid. The questions asked include “What is
the speakers’ recommendation of a full day bus tour?” and “When will the Thames
River tour depart?” In respect of communicative purpose, “the language use takes place
for a definite communicative purpose” (Buck, 2001, p. 89). That is, test-takers should
well understand the purpose of listening. Possible tasks reflecting communicative
purposes could be asking for and giving personal details such as names, ages, etc., or
identifying and describing simple objects such as shape, size, weight, color, etc.
(Cambridge English Certificate Handbook for Teachers, 2018). As to the authentic tasks
used in the test, some people argued that listeners do not perform any task in real-life
situations; however, the idea is that the test tasks should reflect the real-world listening
situation (Buck, 2001). For example, the test-takers were asked to tell a simple story
which happens in everyday lives with the help of pictures and matched a list of
19
illustrated words or names with a set of pictures, such as a picture of parents, a picture
of uncle, and a picture of daughter, by writing the letter of the correct picture in a
corresponding box. (Cambridge English Qualifications Hand book for Teachers, 2018).
Given the interest in communicative language teaching nowadays, it is not
surprising to find that many of the standardized tests such as TOEIC, TOEFL Junior
tests and Cambridge English: Young Learners tests aim at evaluating learners’
communitive ability. Weir (1993) provided a comprehensive taxonomy of
communicative listening sub-skills. He categorized the sub-skills in four aspects (see
Table 1). The four aspects included (1) direct meaning comprehension, such as listening
for gist, listening for specifics, (2) inferred meaning comprehension, such as making
inferences and deductions, relating utterances to their social and situational contexts,
(3) contributory meaning comprehension, such as understanding phonological features
and discourse markers, (4) listening and taking notes.
20
Table 1. A Taxonomy of Communicative Listening sub-skills (Weir, 1993).
Direct meaning comprehension - listening for gist
- listening for the main idea(s) or important information; and distinguishing that from supporting detail, or examples
- listening for specifics, including a recall of important details
- determining a speaker’s attitude or intention towards a listener or a topic Inferred meaning comprehension
- making inferences and deductions
- relating utterances to their social and situational contexts - recognizing the communicative function of utterances
- deducing the meaning of unfamiliar lexical items from context Contributory meaning comprehension
- understanding phonological features
- understanding grammatical notions such as comparison, cause, result, degree etc.
- understanding discourse markers
- understanding the main syntactic structure of clauses or idea units - understanding cohesion, especially reference
- understanding lexical cohesion, especially lexical set membership and collocations
- understanding lexis Listening and taking notes
- ability to extract salient points to summarize the text - ability to select relevant key points
21
As each approach has its supporters and opponents, there is no panacea for testing
learners’ language proficiency for the process of language is too complex. Farhady
(1982) also indicated in his study that there might be some other factors such as test
takers variables including their characteristics, educational and family backgrounds,
which were not directly related to language proficiency but should also be taking into
consideration when applying tests on L2 learners. However, given the interest and trend
in communicative language teaching nowadays, the present study adopted a
communicative testing approach and tests so as to investigate the listening
comprehension of L2 learners of younger students throughout Taiwan with the
authentic input of NESTs in a co-teaching model.
2.4 Synthesis
As more and more native speaker-oriented programs have been gradually
incorporated into Taiwanese’s educational settings, a new issue considering the
incorporated into Taiwanese’s educational settings, a new issue considering the