MOTIVATION AND WILLINGNESS TO COMMUNICATE AS

PREDICTORS OF REPORTED L2 USE:

THE JAPANESE ESL CONTEXT
YUKI HASHIMOTO
University of Hawai’i

ABSTRACT

The purpose of this study was to examine affective variables as predictors of reported second
language (L2) use in classrooms of Japanese ESL (English as a Second Language) students. The study
used the socio-educational model and the willingness to communicate (WTC) model as the basis for a
conceptual framework, partially replicating a study by Macintyre and Charos (1996). Descriptive
statistics, reliability of the subscales, correlation, and construct validity (using principal component
analysis) were examined, and a model of L2 communication was tested using structural equation
modeling.
Using Amos version 4.0, structural equation modeling showed that motivation and WTC affect
reported L2 communication frequency in classrooms as hypothesized. Variables underlying WTC
were also examined. Perceived competence and L2 anxiety were found to be causes of WTC, which
led to more L2 use, and L2 anxiety was found to negatively influence perceived competence,
supporting the results of the Macintyre and Charos (1996) study. Although a path from WTC to
motivation was not found to be significant in the original study, it was found to be significant in the
present replication. In addition, a path from perceived competence was found to exert a strong and
direct influence on motivation from a data-driven path.

INTRODUCTION

The use of the target language is one of the main purposes in learning second
languages for many L2 learners, and it has been widely assumed that the use of the target
language is also an indicator of and a necessary condition for successful second language
acquisition (SLA). Researchers have found that the use of the target language plays a
crucial role in SLA (Seliger, 1977; Swain 1995, 1998). However, though many studies
examine affective variables as predictors of proficiency, there are few studies that
examine affective variables as causes of L2 use.
This study examines affective variables as predictors of reported L2 use by Japanese

Second Language Studies, 20(2), Spring 2002, pp. 29-70.

 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 30

ESL students in classrooms. Using the socio-educational model (Gardner, 1985) and the
WTC model (Macintyre, 1994) as the basis for a conceptual framework, motivation and
willingness to communicate were hypothesized to be main causes of the frequency of L2
use in classrooms. This paper will begin by examining the study of L2 motivation, since
motivation is held to be a major affective variable influencing SLA. Other affective
variables which may influence the frequency of L2 use will also be discussed, including
willingness to communicate, L2 anxiety, and perceived competence.

Gardner’s Approach to Motivation

Gardner and Lambert (1959) developed an approach to motivation which has
influenced various studies in L2 motivation to the present day. They made the distinction
between integrative motivation and instrumental motivation. In their definition,
integrative motivation is positive attitudes toward the target language group and a
willingness to integrate into the target language community, whereas instrumental
motivation refers to practical reasons for learning a language, such as to gain social
recognition or to get a better job.
Gardener (1985) established a model of motivation in second language learning
called the socio-educational model. The model is concerned with the role of various
individual differences in the learning of an L2. In the model, two classes of variables,
integrativeness and attitudes toward the learning situation are said to contribute to the
learner’s level of motivation, and these three classes of variables are said to form
integrative motivation.
The Attitude/Motivation Test Battery (AMTB) was developed by Gardner (1985) to
assess various individual difference variables based on the socio-educational model.
Adaptations of the AMTB have been used in many studies of L2 motivation (e.g., Baker
& Macintyre, 2000; Gardner, Day, & Macintyre, 1992; Gardner, Lalonde, Moorcroft, &
Evers, 1987; Gardner & Macintyre, 1991; Gardner & Macintyre 1993; Gardner,
Tremblay, & Masgoret, 1997; Gliksman, Gardner, & Smythe, 1982; Masgoret, Bernaus,
& Gardner, 2001; Tremblay & Gardner, 1995). The AMTB is made up of over 130
items, and its reliability and validity have been supported (Gardner & Gliksman, 1982;
Gardner & Macintyre, 1993).
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 31

The AMTB consists of 11 subtests that can be grouped into five categories (Gardner,
2001, p. 7). Three of the categories, integrativeness, attitudes toward the learning
situation, and motivation have been mentioned above and included in Gardner’s model.
One of the remaining two is instrumental orientation which refers to an interest in
learning the language for pragmatic reasons that do not involve identification with the
other language community. The other is language anxiety, which involves anxiety
reactions when called upon to use the second language (Gardner, 2001, p. 8). Table 1
presents a listing of the constructs assessed in the AMTB, the subtests that define each
construct, and the number of items typically used in each subtest.

Table 1
Constructs and Scales of the AMTB from Gardner (2001, pp. 8-9)

Construct 1: Integrativeness
Subtest 1: Integrative orientation (4 items)
Subtest 2: Interest in foreign languages (10 items)
Subtest 3: Attitudes toward the target language group (10 items)
Construct 2: Attitudes toward the Learning Situation
Subtest 4: Evaluation of the language instructor (10 items)
Subtest 5: Evaluation of the language course (10items)
Construct 3: Motivation
Subtest 6: Motivation intensity (10 items)
Subtest 7: Desire to learn the language (10 items)
Subtest 8: Attitudes toward learning the language (10 items)
Construct 4: Instrumental Orientation
Subtest 9: Instrumental orientation (4 items)
Construct 5: Language Anxiety
Subtest 10: Language class anxiety (10 items)
Subtest 11: Language use anxiety (10 items)

Motivation Beyond Integrative/Instrumental Distinctions

Gardner’s approach outlined above has influenced many studies in L2 motivation.
Although it is clear that Gardner’s theory has made a large contribution to this area, many
studies calling for reconceptualization of motivation have emerged. Gardner’s theory
took the position that learners’ attitudes toward the target language group affect their
success in learning the target language (Baker & Macintyre, 2000, p. 318). Others held
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 32

that integrative motivation is more influential than instrumental motivation. In fact,

instrumental motivation is discussed in very little detail, whereas integrative motivation is
a key concept in the model (Macintyre, MacMaster, & Baker, 2001, p. 464). It was
pointed out that Gardner’s theory puts too much emphasis on the integrative and
instrumental distinctions.
In response to calls for the adoption of a wider vision of motivation, Tremblay and
Gardner (1995) extended Gardner’s construct of L2 motivation by incorporating other
motivational variables into the model. Gardner (2001) acknowledges that there are
factors other than integrative motivation that affect motivation such as instrumental
motivation and attitudes toward a teacher and a course. Although the focus of the model
is on integrative motivation, Gardner (2001) also maintains that there might be other
factors that have direct effects on language achievement such as language learning
strategies, language anxiety, and self-confidence with the language. He states that the
purpose of the model is to focus attention on the role of integrative motivation, rather
than attempting to show all the possible variables (p. 7). Gardner does not currently
claim that integrative motivation is more influential than instrumental or any other type
of motivation, but simply that those who are integratively motivated will probably be
more successful in language learning than those who are not so motivated (Crookes &
Schmidt, 1991, p. 474).
Another argument which has been raised against the Gardner model is that traditional
approaches influenced by the work of Gardner have been almost exclusively social-
psychological, and they have tended to group attitudes and motivation together (Crookes
& Schmidt, 1991, p. 501). Macintyre et al. (2001) provided empirical evidence that
Gardner’s model deals with attitudinal motivation which might be separate from action
motivation. Macintyre tested for the overlap among concepts from four separate research
paradigms: Gardner’s socio-educational model; the model of academic motivation and
learning strategies of Pintrich; the action control model of Kuhl; and McCroskey’s WTC.
Factor analysis revealed that all of the Gardner AMTB variables loaded heavily on a
factor called attitudinal motivation, rather than on two other factors called action
motivation and self-confidence (Macintyre et al., 2001, p. 482). However, Crookes and
Schmidt (1991) acknowledge that language learning takes place within a social context
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 33

and socially grounded attitudes may provide important support or lack of support for
motivation (p. 501). The focus of their arguments was that Gardner’s approach was so
influential that alternative concepts have not been seriously considered (Crookes &
Schmidt, 1991, p. 501; Dörnyei, 1994, p. 274) and that the theory was limited in terms of
the range of possible influences on motivation that exist (Dörnyei, 1994, p. 274).
Schumann’s acculturation model also emphasizes the importance of social-
psychological factors influencing SLA. Schumann (1986) points out that although
instrumental and integrative motivations are useful ways to think about success in second
language learning, motivations are complex constructs that interact with social and other
variables (p. 384). Schumann’s acculturation model predicts that learners will acquire the
target language to the degree they acculturate to the target language group. Motivation is
seen as one of a large number of affective variables contributing to the construct of
acculturation. There are arguments against the acculturation model that the degree of
acculturation does not always positively correlate with the degree of success in SLA
(Schmidt, 1983; Schumann, 1986) and, since the effects of individual affect may be
variable and complex, it is difficult to test the model (Crookes & Schmidt, 1991, p. 477;
Schumann, 1986, pp. 386-387). Since studies undertaken with regard to the acculturation
model did not provide sufficient support for the model, the model is seen as only one
aspect influencing SLA instead of a major causal variable in SLA (Crookes & Schmidt,
1991).
While early studies done in connection with Gardner’s theory supported the
importance of the integrative over the instrumental motivation, the results found in other
studies were contradictory. Oller, Baca, and Vigil (1977) found that subjects (Mexican
Americans in Southeast) who were instrumentally motivated developed resentment
toward the target community (in this case Anglo Americans) as they progressed in the
target language (in this case English). The authors attributed the anti-integrative
motivation of the subjects to the situation in which colonized minority of Mexican
Americans have been oppressed by a powerful political system (p. 182). There are
several other studies which have found negative correlations between attitudes and
language proficiency (e.g., Chihara & Oller, 1978; Oller, Hudson, & Liu, 1977;
Teitelbaum, Edwards, & Hudson, 1975). Gardner (1980) responds to these counter
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 34

arguments by stating that the inconsistencies are mainly due to statistical, contextual, and
conceptual factors such as statistical exceptions, sociocultural differences, and
differences in how the affective factors are viewed and measured (pp. 264-268).
Clément and Kruidenier (1983) proposed that contradictory results could be traced to
two factors. First, it is difficult to draw a clear line between instrumental and integrative
motivation, and second the relationship between orientations and achievement in a
second language might vary depending on the context in which the learning takes place
(pp. 274-278). In addition to instrumental orientation, they proposed three other
orientations (the acquisition of knowledge, travel, and friendship) in their study based on
factor analysis suggesting that these four orientations should be considered as
independent orientations in future research in place of the integrative/instrumental
distinction (pp. 286-288). Although these four orientations might have extended
integrative and instrumental distinctions to some extent, it seems these categorizations
still do not address the dynamic and variable nature of motivation. In fact, Gardner and
Macintyre themselves acknowledge that since motivation is dynamic, it is too static and
restricted to employ the old characterization of motivation represented by
instrumental/integrative distinctions (Dörnyei, 1994, p. 274).

Qualitative Approaches to Motivation

Several studies have emerged which directed studies of L2 motivation to focus more
on social context and social identity. Norton Peirce (1995) introduced the conception of
investment, building on Bourdieu’s notion of “cultural capital.” She argues that the
instrumental and integrative distinction does not capture the complex relationship among
power, identity, and language learning. Instead, the notion of investment attempts to
capture the relationship of the language learner to the changing social world (p. 17). She
argues that in the field of SLA, artificial distinctions are drawn between the individual
language learner and the social world. However, motivation must be understood with
reference to social context and in relation to the multiple changing and contradictory
identities of language learners across time and space (p. 26). The term investment refers
to the socially and historically constructed relationship of learners to the target language
and their sometimes ambivalent desire to learn and practice it (Norton, 1997, p. 411).
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 35

Syed (2001) also argues that the notions of multiple and socially constructed identity
need to be addressed in the study of motivation (p. 129). Other researchers also saw the
need for more qualitative approaches to complement the largely quantitative tradition of
research on L2 motivation (Crookes & Schmidt, 1991; Dörnyei, 2001; Ushioda, 2001).
Although L2 motivation research can benefit from the use of qualitative techniques, they
are not without disadvantages in terms of their reliability and generalizability.
Many researchers have questioned the use of self-report questionnaires in studies of
L2 motivation on the ground that they do not always elicit true responses from
participants and they are vulnerable to extraneous influences. Self-reported attitude
measures may also be under the influence of extraneous factors such as the desire to look
good in one’s own eyes (self-flattery), or in the eyes of others (the approval motive), or
simply to be consistent in responding to questions of related content (response set).
Further, it has been suggested that subjects must understand the questions in an attitude
survey in order for them to give self-flattering, socially desirable, and consistent
responses. Therefore, if the questions are phrased in the subject’s native language, they
become a test of intelligence and a rather direct test of first language proficiency. If the
questions on the other hand are phrased in the target language, they become a target
language proficiency measure (Oller, 1981; Oller & Parkins, 1978a; Oller & Parkins,
1978b). Gardner responded that all such claims are based on speculation and lack
empirical support (Gardner, 1980; Gardner & Gliksman, 1982).

Factors Affecting Frequency of the L2 Use

The use of the target language is an end in itself for many L2 learners, and it is
generally believed to be an indicator of and a necessary condition for successful second
language acquisition. In Seliger’s (1977) research with adults studying ESL in the United
States, it was found that students who participate more and thereby elicit more teacher
input exhibit greater gains in L2 proficiency compared to students who play a passive
role in language interaction. Swain (1995, 1998) also emphasized the role of output (i.e.,
production or use) in L2 learning, stating that output is necessary for the development of
production (talking and writing) as input develops only listening and reading
comprehension. According to Swain (1998), output has three functions in L2 learning
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 36

which are noticing, hypothesis testing, and metalinguistic functions. By producing

output, learners notice the gap between the target language and their own interlanguage,
or what they want to say and what they cannot say which may prompt learners to
recognize their linguistic problems. In addition, learners use their output as a way to test
hypotheses about the second language by way of experimenting with new structures and
forms to modify their L2. Finally, output produces metatalk which is language used in
problem solving and for cognitive purposes. Using metatalk, learners become more
aware of noticing, hypothesis testing, and other language learning processes. In contrast,
Day (1984) did not find that the voluntary classroom participation of adult ESL students
in the U.S. was significantly related to proficiency. Although the use of the target
language may not be the only factor affecting the acquisition of the target language, it is
clearly an important condition for successful target language acquisition.
Various affective variables influence the use of the target language in classrooms.
Ely (1986) tested the effects of language class discomfort, language class risktaking,
language class sociability, and strength of motivation, as well as attitude toward the
language class, concern for grade, and language learning aptitude on the classroom
participation of students enrolled in first year university Spanish classes. Data on
classroom participation were collected by classroom observation and other data were
gathered by surveys. Ely hypothesized that the strength of motivation as well as
language class risktaking positively influence classroom participation. On the other
hand, it was posited that language class discomfort has a direct negative influence on
classroom participation as well as an indirect influence through reducing language class
risktaking and language class sociability. It was found that language class risktaking is a
significant positive predictor of classroom participation and language classroom
discomfort influenced classroom participation only indirectly. Other variables did not
have a significant effect on classroom participation.
Several studies have suggested that integrative motivation has a positive influence on
the frequency of the L2 use which in turn affects second language proficiency. In two
investigations (Gardner, Smythe, Clément, & Gliskman, 1976; Gliksman et al., 1982), the
effects of integrative motivation on the frequency of L2 use in classroom were examined
by administering a motivational questionnaire and conducting classroom observation.
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 37

Both studies examined secondary school students in Canada who were enrolled in French
classes. It was hypothesized that integratively motivated students would take every
opportunity to perfect their second-language skills, and they would use the classroom as
an opportunity to use their L2. It was found that integratively motivated students, in
contrast with those not integratively motivated, exhibited a significantly greater number
of several classroom behaviors, including volunteering to answer questions and making
more correct responses. Gardner et al. (1987) employed self-report questionnaires
instead of classroom observations to examine the frequency of the L2 use of secondary
school students who are enrolled in French classes. The findings support the above
studies in that the integrative motivation plays a role in the frequency of the L2 use, and
the frequency of L2 use contributes to individual differences in proficiency (p. 42).
It has been shown that, in addition to attitudes and motivation, anxiety has a large
impact on second language learning (Horwitz, 1986; Horwitz, 2001; Horwitz, Horwitz, &
Cope, 1986; Horwitz & Young, 1991; Macintyre & Gardner, 1989; Macintyre &
Gardner, 1991). Horwitz et al. (1986) identified foreign language anxiety as a situation
specific anxiety which is distinct from other anxieties. The Foreign Language Classroom
Anxiety Scale (FLCAS) developed by Horwitz et al. (1986) is designed to assess three
components of anxiety: communication apprehension, test anxiety, and fear of negative
evaluation. It has been shown that the FLCAS has satisfactory reliability and validity
(Horwitz, 1986). Language anxiety has been shown to correlate negatively with
achievement measures such as language course final grades (Horwitz, 1986) and
performance on a vocabulary learning tasks (Macintyre & Gardner, 1989). Gardner and
Macintyre (1993) found that among attitudes, motivation, and anxiety, measures of both
classroom anxiety and language use anxiety showed the strongest correlations with
several language production measures including a cloze test, a composition task, and an
objective proficiency measure. Gardner and Macintyre (1993) found language anxiety
correlates more highly with the self-ratings of proficiency than with actual performance
on the tests of ability. It was found that anxious students tend to underestimate their
ability and less anxious students tend to overestimate their ability (Macintyre, Noels, &
Clément, 1997). Communication apprehension has also been widely studied, not only in
the field of language education, but also in the field of speech communication (Daly,
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 38

1991). Although communication apprehension refers to first language anxiety, it is said

that it is conceptually similar to language anxiety in that they both refer to anxiety about
communicating (Daly, 1991; Horwitz et al., 1986).
Much of the research discussed above has demonstrated the influence of affective
variables on achievement and other behavioral measures. A recent addition to the
affective variables coming from the field of speech communication is “willingness to
communicate” (WTC). McCroskey and associates employed the term to describe the
individual’s personality based predisposition toward approaching or avoiding the
initiation of communication when free to do so (McCroskey, 1992, p. 17). WTC was
originally introduced with reference to L1 communication, and it was considered to be a
fixed personality trait that is stable across situations, but when WTC was extended to L2
communication situations, it was proposed that it is not necessary to limit WTC to a trait-
like variable, since the use of an L2 introduces the potential for significant situational
differences based on wide variations in competence and inter-group relations (Macintyre,
Clément, Dörnyei, & Noels, 1998). Macintyre et al. (1998) conceptualized WTC in an
L2 in a theoretical model in which social and individual context, affective cognitive
context, motivational propensities, situated antecedents, and behavioral intention are
interrelated in influencing WTC in an L2 and in L2 use (Figure 1).
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 39

Figure 1. Heuristic model of variables influencing WTC ( Macintyre, Clément, Dörnyei,

& Noels, 1998, p. 547)

Macintyre (1994) developed a path model that postulates that WTC is based on a
combination of greater perceived communicative competence and a lower level of
communication apprehension. The model also postulates that anxiety influences the
perception of competence. Baker and Macintyre (2000) examined the effects of an
immersion versus a non-immersion program on various dependent variables including
perceived competence, WTC, self-reported frequency of communication, communication
anxiety, and motivation of students who have English as their L1 and are studying French
as their L2. It was found that anxiety and perceived competence were key factors in
predicting WTC and self-reported frequency of communication.
Macintyre and Charos (1996) tested a hybrid of Gardner’s socio-educational model
(1985) and Macintyre’s (1994) WTC model to predict the frequency of using the second
language in the daily interactions of Anglophone students taking introductory level
conversational French at adult evening classes. All the paths that were derived from the
Gardner and Macintyre models were replicated. The results confirmed that students who
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 40

have greater motivation for language learning report using the language more frequently
and students who are more willing to communicate are more likely to do so. The
hypothesized variables underlying WTC were also tested. Both language anxiety and
perceived competence influenced WTC, and the predicted effect of anxiety on perceived
communicative competence was also supported. It was shown that perceived
communicative competence has a strong and direct influence on the L2 communication
frequency from a data-driven path. A path from WTC to motivation was also
hypothesized but was not found to be significant.
In the Macintyre and Charos model, it was also hypothesized that personality traits
and social context have an indirect effect on L2 communication frequency through
attitudes, motivation, language anxiety, and perceived competence. Their hypothesis was
based on a study by Lalonde and Gardner (1984) which concluded that personality traits
have an effect on second language achievement indirectly, through motivation and
attitudes. Personality traits were measured using a scale of the “Big-Five” which assesses
five global personality traits: extroversion, agreeableness, conscientiousness, emotional
stability, and intellect. These personality traits influenced motivation and WTC which in
turn affected L2 communication frequency. Social context was measured by a self-report
of the relative concentration of L1 and L2 at home and at work. It was found that having
more opportunities for interaction in L2 affects frequency of L2 use directly and also
indirectly through perceived competence and WTC. These findings support the
suggestions by Macintyre et al. (1998) that context and personality are among the
variables influencing the WTC.
Yashima (2002) investigated variables underlying the WTC in a Japanese English as
a foreign language context using Macintyre’s WTC model and Gardner’s socio-
educational model. Since there is little daily contact with native speakers of English in
the Japanese EFL context, frequency of communication was not included in this model.
Instead, L2 proficiency, attitude toward the international community, confidence in L2
communication, and L2 learning motivation were hypothesized to affect the WTC in the
L2. The hypothesized causes of WTC were replicated. It was shown that a lower level
of anxiety and a higher level of perception of L2 communication competence led to a
higher level of WTC, thus supporting the results of the Macintyre and Charos (1996)
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 41

study. In this model, a combination of relative lack of anxiety and perceived competence
was hypothesized to form the latent variable self-confidence in L2 communication based
on Clément’s model (Clément & Kruidenier, 1985). A data-driven path from motivation
to confidence in L2 communication was significant. A hypothesized direct path from
motivation to WTC was not significant.

Purpose of the Study

The purpose of the present study is to examine the relationships among L2 learning
and L2 communication variables using the WTC model and the socio-educational model
as the basis for a framework and to extend the models by testing their ability to predict
language use in the Japanese ESL formal (classroom) context. Six research questions
motivate the present study:
1. How is the reported willingness to communicate of Japanese ESL students, as
measured by the WTC scale, related to reported frequency of L2 use in classrooms?
2. How is the reported motivation of Japanese ESL students, as measured by the mini-
AMTB, related to reported frequency of L2 use in classrooms?
3. How is the reported perceived competence of Japanese ESL students related to
reported frequency of L2 use in classrooms?
4. To what degree are perceived competence and communication apprehension causes
of WTC, as hypothesized by Macintyre (1994) and found by Macintyre and Charos
(1996)?
5. To what degree is communication apprehension a cause of perceived competence, as
hypothesized by Macintyre (1994) and found by Macintyre and Charos (1996)?
6. To what degree is WTC related to motivation, as hypothesized by Macintyre and
Charos (1996)?
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 42

METHOD

Participants
The participants were 56 Japanese undergraduate and graduate students attending the
University of Hawaii at Manoa (UHM) in Honolulu. All of the participants spoke
English as their second language and Japanese as their first language. To enter
undergraduate programs at the UHM, students must have a TOEFL score of at least 500.
Some graduate students are required to have TOEFL scores as high as 620. Since this
study was done anonymously, gender, major, and class standing were not classified.

Materials
Measures of language learning affect. A short version of the Attitude/Motivation
Test Battery (the mini-AMTB) was employed. The mini-AMTB has recently been
introduced to reduce administration time while maintaining the basic conceptual structure
of the original version. This “Guilfordstyle” instrument measured the eleven variables in
the original AMTB (see Table 1) using single-item indicators each on a 7-point rating
scale. Several studies have successfully employed the mini-AMTB (e.g., Baker &
Macintyre, 2000; Gardner & Macintyre, 1993; Macintyre & Charos, 1996; Macintyre &
Noels, 1996; Masgoret et al., 2001). In spite of the potential problems with single-item
measures, Gardner and Macintyre (1993) have shown that this instrument has acceptable
concurrent and predictive validity. Since the original AMTB was written with regard to
attitudes toward learning French and French Canadians, it was modified to refer to
attitudes toward learning English and English speakers. The five subscales on this
measure are as follows:
1. Integrativeness (α = .86 in Macintyre & Charos, 1996). This measures the degree to
which respondents were learning English for the purpose of interacting and
communicating with members of the second language community. Integrativeness
was measured with three single-item measures of integrative orientation, attitude
toward the target language group, and interest in foreign languages.
2. Attitudes toward learning situation (α = .89 in Macintyre & Charos, 1996). This was
measured by two items, attitude toward the language teacher and attitude toward the
course.
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 43

3. Motivation (α = .65 in Macintyre & Charos, 1996). Motivation was measured with
three single-item measures of the desire to learn English, motivational intensity, and
attitude toward learning English.
4. Instrumental orientation. This was measured by one item of instrumental orientation.
5. Language anxiety (α = .48 in Macintyre & Charos, 1996). This was measured by two
items, one assessing English classroom anxiety and the other measuring English use
anxiety.
Communication-related variables. The following four measures were adapted to
refer to communication using English. Each of the measures presents 12 communication
contexts involving four communication contexts: (a) public speaking, (b) formal
meetings, (c) small groups, and (d) dyads, and each of these is applied to three types of
receivers (strangers, acquaintances, and friends).
1. Willingness to communicate in English (α = .97 in Macintyre & Charos, 1996). This
study used the WTC scale from McCroskey (1992). Twenty items assessed the
percentage of time respondents would choose to communicate in each type of
situation (when completely free to do so) using a probability estimate scale between
0% and 100%. Eight of the items are fillers (items 1, 2, 5, 7, 10, 13, 16, and 18), and
12 items are scored as part of the scale (McCroskey, 1992). A few changes were
made in the questionnaire to make it more appropriate for the respondents in this
study. This instrument was shown to have strong content validity, and there is some
support for its construct and predictive validity (McCroskey & Richmond, 1990, p.
73).
2. Perceived competence in English (α = .98 in Macintyre & Charos, 1996). Twelve
items from Macintyre and Charos (1996) assessed the average percentage of time
(ranging from 0% to 100%) that respondents felt competent in using English to speak
in 12 situations. This instrument was modified to refer to classroom contexts in this
study.
3. Frequency of communication in English (α = .97 in Macintyre & Charos, 1996).
Items from the perceived competence scale were adapted to measure the frequency of
communication in English for each of the 12 situations using a 7-point scale.
4. Communication anxiety in English (α = .92 in Yashima, 2002). The 12 items for
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 44

communication apprehension or anxiety used by Yashima (2002) assessed the

average percentage of nervousness (ranging from 0% to 100%) that respondents felt
in communicating in English in 12 situations. This instrument was modified to refer
to classroom contexts in this study.
Items regarding instrumental orientation and communication anxiety in English were
not included in the Macintyre and Charos (1996) study, but they were added in the
present study. Instrumental orientation was added because it plays an important role in
the socio-educational model, and it was included in the original mini-AMTB by Gardner
and Macintyre (1993). Macintyre and Charos (1996) used two language anxiety items
from the mini-AMTB to measure L2 anxiety, but here, it seemed more appropriate to
assess L2 anxiety with more items on a separate instrument. Therefore, in this study, L2
anxiety was measured with 12 communication anxiety items. All scales were translated
into Japanese. Back-translation was used to ensure the accuracy of the translation. The
English version of the questionnaire is shown in Appendix.

Procedures
Students participated in this study voluntarily and received a movie ticket as
compensation for their participation. They were presented with a consent form and told
that the data would be collected anonymously and kept confidential. Respondents were
given as much time as required to complete the questionnaire.

RESULTS

The SPSS version 11.0 statistical program was used to analyze descriptive statistics
and reliability, and to do principal components analysis. Amos version 4.0 was used to
test the hypothesized model using structural equation modeling.
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 45

Descriptive Statistics
Descriptive statistics are numerical representations of how participants performed on
a test or questionnaire (Brown, 1996). These descriptive statistics are averages for each
participant of all the items in the corresponding measures. The variable labels represent
each of the measures as follows. AMTB represents a brief version of the
Attitude/Motivation Test Battery, WTC is the willingness to communicate scale, PC is
perceived competence, ANXIET refers to communication anxiety, and FREQ is
frequency of communication. The statistics include the number of participants (N),
number of items (k), mean (M), standard deviation (SD), minimum (MIN), and maximum
(MAX), Median (MDN), mode (MODE), and skewness (SKEW). It should be noted that
AMTB and FREQ are based on a 7-point scale, whereas other measures are based on a
probability estimate scale ranging from 0% to 100%.

Table 2
Descriptive Statistics

VARIABLE N k M SD Min Max Mdn Mode Skew

AMTB 56 11 5.31 .79 3.36 6.64 5.41 5.36 -.61
WTC 56 12 54.75 16.79 17.50 89.17 53.33 48.33 .00
PC 56 12 66.15 18.42 12.92 94.17 69.38 45.83 -.59
ANXIET 56 12 43.59 18.07 6.25 95.83 42.92 23.33 .68
FREQ 56 12 4.42 1.09 1.25 7.00 4.42 4.83 -.32

The mean, median, and mode are indicators of the central tendency of the scores. The
standard deviation, as well as the minimum and maximum scores, are indicators of the
dispersion of scores around the mean. In these measures the dispersion appears to be
fairly broad in all cases. Skewness characterizes the degree of asymmetry of a
distribution around its mean (Brown, 1997). If a distribution of scores is skewed, that
means it is probably non-normal because of a high number of high or low scores. In such
cases, the skewness statistic will vary widely from .00 with a positive value indicating the
possibility of a positively skewed distribution or with a negative value indicating the
possibility of a negatively skewed distribution. Values of 2 standard errors of skewness
(ses) or more are considered to be skewed to a significant degree (Brown, 1997; Brown,
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 46

Cunha, Frota, & Ferreira, 2001, p. 263).

Since the ses in this study is .327, two times the standard error of the skewness is
.654. As the skewness statistic for ANXIET is .68, which is slightly higher than .654, it
can be assumed that the distribution of communication anxiety scale is significantly
skewed. Since the sign of the skewness statistic is positive, the distribution is positively
skewed. In other words, for the Japanese students involved here, the distribution for
ANXIET was not normal, but instead, was somewhat positively skewed, which means
there was a high number of low scores on the measure of communication anxiety. Other
skew statistics fell within the range between -.654 and + .654 which indicates that there
are no other significant skewness problems.

Reliability

The reliability coefficients indicate the degree to which the results on a scale can be
considered internally consistent, or reliable (Brown, 1996, p. 192). The Cronbach alpha
was used in this study. It can range from .00 to 1.00.

Table 3
Cronbach’s Coefficient Alpha Reliability Estimates and Standard Error of Measurement

Variable k Alpha sem

AMTB 11 .83 .33
WTC 12 .85 6.50
PC 12 .95 4.12
ANXIET 12 .90 5.71
FREQ 12 .92 .31

Table 3 shows that all the Cronbach alpha estimates are reasonably high. They can be
interpreted as the percent of consistent variance in the students’ answers. For example,
the reliability of .83 for the AMTB can be said to indicate that the scale is 83%
consistent, or reliable (Brown et al., 2001, p. 264). Another way of looking at the
consistency of a set of scores is called the standard error of measurement (sem). The sem
can be interpreted as a band around a student’s score within which that student’s score
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 47

would be expected to fall repeatedly if they were to fill out the instrument repeatedly
(Brown, 1996, p. 206; Brown et al., 2001, p. 265). For instance, the sem of 6.50 for the
WTC indicates that a participant who has a total score of 50 on that scale can be expected
to score within a band of one sem plus (50+6.50 = 56.50) or minus (50-6.50 = 43.5) 68 %
of the time if the participant were to fill out the instrument time and again. The sem may
be easier to interpret than a reliability coefficient because it is expressed in terms of raw
score bands rather than percent-of-reliability terms. A scale that has a small sem is more
consistent than one with a large sem (Brown, 1996, p. 208). Considering that AMTB and
FREQ are on a 7-point scale and other scales are based on a probability estimate scale,
the sem can be said to be fairly narrow for all five scales.

Correlation
Table 4 shows a correlation matrix for the five main variables in this study: AMTB,
WTC, PC, ANXIET and FREQ. All correlations except that between ANXIET and WTC
were significant at p < .05. As expected, FREQ correlated significantly with the other
four variables. It was expected that ANXIET would correlate significantly with WTC,
but there was no significant correlation. There was a significant negative correlation
between ANXIET and PC and between ANXIET and AMTB suggesting that lower L2
anxiety is associated with higher L2 perceived competence and higher motivation. The
AMTB was positively correlated with WTC and PC indicating that higher motivation is
related to higher willingness to communicate and higher perceived competence.

Table 4
Correlation Matrix

AMTB WTC PC ANXIET FREQ

AMTB 1.00
WTC .39* 1.00
PC .59* .26* 1.00
ANXIET -.33* -.05 -.46* 1.00
FREQ .50* .36* .38* -.27* 1.00
* p < .05
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 48

Validity
Construct validity. Principal components analysis was performed (with VARIMAX
rotation) on the responses to the five scales to investigate the degree to which the
instrument was measuring what it claims to measure. Examining the Eigen values above
1.00, the scree plot, theory, and the interpretability of the rotated factors, a six factor
solution was determined to be best. These six factors accounted for 62% of the variance.
The loadings for each of the variables on six factors are shown in Table 5. The asterisks
indicate loadings of .30 or higher, and the bold-faced type indicates the highest loading
for each variable. Communalities are presented in the column furthest to the right. The
communalities indicate the total proportion of variance that the six factors account for in
each variable (Brown, Robson, & Rosenkjar, 2001). At the bottom of the table, a row is
presented which indicates the proportion of variance in the overall solution accounted for
by each factor. For example, the proportion of variance accounted for by the first factor
is .15, which represents 15% of the variance in the overall solution.

Table 5
VARIMAX Rotation of the Six Factor Solution

Variable / Components 1 2 3 4 5 6 h2
Motivation
AMTB1 .38* .44* -.01 .11 .04 .24 .41
AMTB2 .21 .22 .05 -.04 .31* .50* .45
AMTB3 .48* .31* -.10 .16 .14 .36* .51
AMTB4 .07 .16 .03 .45* -.04 .40* .39
AMTB5 .04 .26 .00 .19 .20 .48* .38
AMTB6 .43* .10 .00 .06 .14 .52* .49
AMTB7 .22 .31* .03 .12 -.02 .55* .46
AMTB8 .39* .08 .19 .14 .41* .22 .44
AMTB9 .25 .40* -.08 .13 -.22 .51* .55
AMTB10 .35* .07 -.36* .08 .11 -.11 .29
AMTB11 .46* .17 -.37* -.24 .27 -.07 .51
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 49

L2 Willingness to Communicate
WTC3 .00 .03 .12 -.15 .84* .10 .75
WTC4 .25 .25 -.06 .63* .01 .14 .55
WTC6 .11 .04 -.16 .69* .28 .08 .59
WTC8 .29 .15 -.15 .25 .52* -.11 .47
WTC9 .16 .01 -.04 .72* -.06 .20 .59
WTC11 .16 .11 -.05 .75* .39* .07 .76
WTC12 .10 .54* -.06 .03 .23 -.01 .35
WTC14 -.10 .13 .09 .13 .82* .16 .75
WTC15 .06 .10 -.15 .65* .38* -.08 .62
WTC17 .06 .14 -.15 .30* .55* -.18 .47
WTC19 .12 .06 -.04 .84* .06 -.13 .74
WTC20 -.02 .10 .17 .19 .83* .13 .78
L2 Perceived Competence
PC1 .61* -.15 -.25 -.38* .27 .30* .76
PC2 .78* .26 -.16 .16 .08 .30* .82
PC3 .79* .10 -.16 .29 .07 .11 .76
PC4 .83* .20 -.27 .11 -.06 .10 .83
PC5 .78* .07 -.12 .27 .08 -.07 .72
PC6 .78* .25 -.20 .15 .12 .14 .76
PC7 .80* .26 -.21 .07 -.11 .09 .79
PC8 .61* .09 -.29 -.06 -.05 .58* .80
PC9 .79* .17 -.17 .34* -.10 .10 .82
PC10 .79* .13 -.23 -.10 -.03 .02 .70
PC11 .76* .01 -.19 .33* -.06 -.06 .73
PC12 .68* .07 -.25 -.14 .06 .48* .78
L2 Anxiety
ANXIET1 .14 .17 .54* .52* -.08 -.41* .78
ANXIET2 -.35* -.15 .64* -.05 -.04 .02 .56
ANXIET3 -.06 -.26 .78* -.24 .09 -.11 .75
ANXIET4 -.16 -.08 .75* .05 .10 .06 .62
ANXIET5 -.24 -.06 .70* -.20 .14 -.09 .61
ANXIET6 -.17 -.18 .81* -.13 -.11 -.11 .77
ANXIET7 -.27 -.02 .55* .12 .16 .16 .44
ANXIET8 .12 -.10 .47* .36* .04 -.51* .64
ANXIET9 -.25 -.10 .78* -.18 -.06 -.06 .72
ANXIET10 -.21 .04 .72* .15 -.02 .03 .58
ANXIET11 -.15 -.13 .68* -.32* .14 -.18 .65
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 50

ANXIET12 -.01 -.10 .53* .45* -.11 -.49* .75

L2 Communication Frequency
FREQ1 .00 .76* -.06 -.13 .25 .14 .69
FREQ2 .24 .59* -.09 .07 -.04 .24 .48
FREQ3 .18 .72* .03 .25 -.09 .14 .64
FREQ4 .27 .63* -.15 .28 .02 -.08 .57
FREQ5 .20 .61* -.05 .21 -.25 .30* .62
FREQ6 .15 .73* -.08 .25 .06 .23 .68
FREQ7 .08 .69* -.20 -.17 .16 -.24 .63
FREQ8 -.03 .67* -.01 -.02 .11 .32* .56
FREQ9 .08 .75* -.12 .05 .04 .09 .59
FREQ10 .20 .72* -.02 -.11 .14 -.26 .65
FREQ11 .10 .67* -.17 .29 -.19 .20 .65
FREQ12 -.02 .69* -.14 .03 .25 .23 .61
Proportion of Variance .15 .13 .11 .10 .07 .07 .62
* loadings above .30
[bold] highest loading for each variable

Examining Table 5, you will notice that all of the items for L2 perceived competence
load most heavily on component one, the items for L2 communication frequency loads
most heavily on component two, and that all the items for L2 anxiety except for ANXIET
8 load most heavily on component 3. Two other scales (motivation and L2 willingness to
communicate) present more complex patterns of loadings. Six of the items (WTC 4, 6, 9,
11, 15, and 19) of WTC load most heavily on component four, and five of the items
(WTC 3, 8, 14, 17, and 20) load most heavily on component five. Three items (WTC 11,
15 and 17) load on both components four and five. When items loading most heavily on
component four are closely examined, it appears that they are all about communicating in
informal situations. For example, WTC 9, which loads most heavily on component four,
asked participants’ willingness to communicate with a friend while standing in line. In
contrast, the four items loading most heavily on component five (WTC 3, 14, 17, and 20)
appear to be about communication in formal situations. For instance, WTC 3 asks
participants’ willingness to speak in public to a group of strangers. WTC 8, which loads
most heavily on component five, asks about participants’ willingness to communicate in
a small group of strangers. It seems that the interpretation of WTC 8 as formal or
informal can differ depending on the context. Interestingly, WTC 12, which asks about
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 51

willingness to talk with a stranger while standing in line, does not load on either
component four or five. It appears that WTC 12 is measuring something different from
other items on the WTC.
The AMTB has the most complex patterns of loadings. Five items (AMTB 2, 5, 6, 7,
and 9) of the AMTB load most heavily on component six. AMTB 10 loads most heavily
on component three, an L2 anxiety component, and also loads heavily on component one,
a perceived competence component. These loadings can easily be interpreted because
AMTB 10 asks about language class anxiety. Similarly, AMTB 11 loads most heavily on
component one and also loads heavily on component three. These loadings can easily be
explained because AMTB 11 asks about language use anxiety. It is clear that these two
items of the motivational scale are more closely related to perceived competence and L2
anxiety than to motivation. The loadings for AMTB 1, AMTB 3, AMTB 4, and AMTB 8
are not so easily interpretable. AMTB 1, which asks about integrative orientation, loads
most heavily on an L2 communication frequency component and also loads heavily on an
L2 perceived competence component. AMTB 3, which asks about attitudes toward the
target language group, also loads on both an L2 perceived competence component and an
L2 communication frequency component along with a motivation component. Since both
AMTB 1 and AMTB 3 are subscales of a construct called integrativeness, this pattern
might suggest that integrativeness is related to L2 perceived competence and
communication frequency. AMTB 4, which asks about the evaluation of a language
instructor, loads heavily on both informal willingness to communicate and motivational
components. AMTB 8, which asks about attitudes toward learning a language, loads
heavily on L2 perceived competence and formal willingness to communicate
components. These loadings are interpretable since the attitudes toward learning a
language would involve L2 perceived competence and willingness to communicate in
formal situations. These complex patterns of loadings probably may simply indicate that
motivation is a complex variable which is influenced by other variables.

Structural Equation Modeling

Structural equation modeling (SEM) is also known as analysis of covariance
structures, or causal modeling (Arbuckle & Wothke, 1999). SEM is a statistical
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 52

methodology that takes a confirmatory hypothesis-testing approach to the analysis of a

structural theory. The hypothesized model can be tested statistically to determine the
extent to which it is consistent with the data. If goodness of fit is adequate, the model
argues for the plausibility of assumed relations among variables, but if the goodness of fit
is not adequate, the tenability of such relations is rejected (Byrne, 2001). Amos is short
for Analysis of Moment Structures, and it is one of the widely used programs for SEM.
The models were tested using Amos version 4.0 in this study. Figure 2 is a portion of the
model from Macintyre and Charos (1996). The figure describes the relationships among
L2 learning and L2 communication variables in French as a second language context in
Canada.

Figure 2. Model of L2 communication applied to French as a second language situation

in Canada (Macintyre & Charos, 1996, p. 12)

The major elements of Gardner’s (1985) model and Macintyre’s (1994) model are
shown in this model. The model shows that language anxiety reduces perceived
communicative competence, and both of these variables influence willingness to
communicate. Both willingness to communicate and L2 motivation contribute to the
extent of the L2 communication frequency. A path from willingness to communicate to
motivation is proposed based on Clément’s (1980) model. In Figure 2, the path from
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 53

willingness to communicate to motivation has been deleted, since it was not significant.
This path was based on speculations about the relations among the variables and had not
been tested before. The dotted path from perceived competence to L2 communication
frequency is a data-driven path and therefore considered to be tentative. Solid paths
indicate originally hypothesized paths.
Figure 3 presents the process of in this study of revising the model of L2
communication applied to a Japanese ESL classroom context. The base model is a
replication of Macintyre and Charos’s model. To determine the goodness-of-fit between
the hypothesized model and the sample data, in other words, to test the appropriateness of
the model, the goodness-of-fit statistics in Amos are examined.
Looking at Table 6, chi-square for base model was 20.76 with 3 degrees of freedom
which was significant. In this case, a non-significant finding is an indication of
goodness-of-fit. Other fit indexes are also provided since chi-square is considered to be
of limited value especially with small samples (Byrne, 2001, p. 81). GFI indicates
goodness-of-fit index and AGFI indicates adjusted goodness-of-fit index with values
close to 1.00 being indicative of good fit. The AGFI differs from the GFI in that it
adjusts for the number of degrees of freedom in the specified model (Byrne, 2001, p. 82).
As shown in Table 6, for the base model, both GFI (0.89) and AGFI (0.44) are indicative
of a poor fit of the model to the data. For the fit index labeled CFI (comparative fit
index), values larger than .95 are considered representative of a good-fitting model.
RMSEA represents the root mean square of approximation. Values less than .05 indicate
a good fit. The expected cross-validation index (ECVI) has no determined appropriate
range of values, but the model having the smallest ECVI value demonstrates the best fit.
Table 6 shows that CFI (0.69), RMSEA (0.33), and ECVI (0.81) are indicative of an ill-
fitting model. Therefore, all the fit statistics indicate poor fit for the base model. Note
that there are other fit statistics but only those mentioned above are presented in the table.
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 54

Base Model

Revision 1

Revision2
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 55

Figure 3. The process of revising the model of L2 communication applied to Japanese

ESL classroom context

Table 6
Step-By-Step Procedure for Revising the Model to Add and Delete Data-Driven Paths

Model χ² df χ²/df GFI AGFI CFI RMSEA ECVI

Base model 20.76 3 6.92 .89 .44 .69 .33 .81
Revision 1: Delete path from
perceived competence to
frequency of
communication 21.56 4 5.39 .89 .57 .69 .28 .80
Revision 2: Add path from
perceived competence to
Motivation 2.31 3 .77 .99 .92 1.00 .00 .48

In Revision 1, shown in Figure 3 and Table 6, a data driven path from perceived
competence to L2 communication frequency has been deleted. Perceived competence
was found to exert a direct and strong influence on the frequency of communication with
beginning French as a second language students in Canada by Macintyre and Charos
(1996). In the base model, perceived competence influenced L2 communication
frequency weakly with a standardized regression weight of .13. Since this path was data
driven, which is exploratory and tentative, the path was deleted. Examining Table 6, chi-
square for Revision 1 is 21.56 with 4 degrees of freedom which is significant. Other fit
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 56

statistics also indicate a poor fit of the model to the data.

Areas of misfit in the model can be identified by examining two types of information
which are standardized residuals and the modification indexes. The essence of SEM is to
determine the fit between the hypothesized model and the sample and any discrepancy
between the two is captured by the residual covariance matrix (Byrne, 2001, p. 88). The
matrix of standardized residuals gives estimates of the number of standard deviations the
observed residuals are from the zero residuals that would exist if model fit were perfect;
values larger than 2.58 are considered to be large (Byrne, 2001, p. 89). In examining the
standardized residual values of Revision 1, a residual value of 3.567 was found for the
covariance between perceived competence and motivation. This was the only value that
exceeded the cut point of 2.58. From this, it can be said that the only statistically
significant discrepancy lies in the covariance between these two variables.
The modification indexes (MIs) are another way to detect model misspecification.
The MIs can be conceptualized as a chi-square statistic with one degree of freedom and
the value of the MIs indicate the extent to which chi-square would be reduced by adding
an additional path having the highest modification index value (Byrne, 2001, p. 90). In
reviewing the regression weights section, the MI value between motivation and perceived
competence was 15.115 with an expected parameter change value of 0.021. Based on the
standardized residuals and the modification indexes of Revision 1, a path from perceived
competence to motivation was added which resulted in Revision 2. Table 6 shows that
chi-square for Revision 2 is 2.31 at 3 degrees of freedom, which was not significant.
Other fit statistics also indicate a very good fit of the model with a GFI of .99, a CFI of
1.00, and an RMSEA of 0.00. Therefore, Revision 2 represents the final model in this
study.
DISCUSSION

Significant positive paths were obtained leading from willingness to communicate

and motivation to L2 communication frequency. These paths indicate that students who
have greater motivation for language learning and who are more willing to communicate
report using the language more frequently in the classroom. Although a path from
perceived competence to L2 communication frequency was found to be significant by
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 57

Macintyre and Charos (1996), the path was not significant with these particular Japanese
ESL students even though it was expected that higher perceived competence would lead
to more frequent L2 use in classroom. The Macintyre and Charos study was conducted
with beginning students whose actual proficiency was low; perhaps perceived
competence did not influence L2 use as much with more advanced students. This
suggests that merely perceiving that one has the ability to communicate can affect the
frequency of L2 use with beginning students but not with more advanced students.
Perceived competence and L2 anxiety were found to be causes of WTC supporting a
hypothesis proposed in Macintyre (1994) and supported in the study by Macintyre and
Charos (1996). L2 anxiety was found to exert a strong and direct negative influence on
perceived competence, supporting the Macintyre (1994) hypothesis and results found in
the Macintyre and Charos (1996) study. Although a path from L2 WTC to motivation
was not found to be significant by Macintyre and Charos, it was found to be significant in
this study indicating that willingness to communicate has motivational properties. The
largest single effect was obtained from perceived competence to motivation. This path
suggests that increased perceived competence will lead to increased motivation which in
turn affects frequency of L2 use in the classroom. This suggests that perceived
competence or self-confidence in an L2 is a positive indicator of motivation. Since
adding additional paths is regarded as exploratory, and data-driven, this path needs to be
replicated and should be further investigated.
There are several limitations to this study. First of all, the sample size was limited (n
= 56). In addition, the frequency of communication was measured using self-report
questionnaires. As discussed earlier, there are some problems involved in the use of self-
report questionnaires in L2 motivational studies. Also, this study may be generalizable
only to Japanese students.
Nonetheless, this study has some implications for teachers. One is that by increasing
perceived competence and reducing language anxiety, the willingness to communicate
may lead to more language use in the classroom increases. Creating a less threatening
atmosphere to reduce anxiety and encouraging students to increase perceived competence
may be effective in increasing willingness to communicate and frequency of L2 use in
classrooms with Japanese ESL students. Perceived competence had a direct and strong
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 58

influence on motivation, which in turn affected L2 communication frequency in the

classroom. It may be especially important with the Japanese ESL students to increase
perceived competence.
The following questions may prove useful for future research in line with this study:
1. What relationships would be found in comparisons between intention to behave and
actual behavior?
2. Would similar results be obtained if frequency of L2 use were extended to use outside
of classroom context?
3. Would similar results be obtained if this study were replicated with Japanese ESL
students at different levels of proficiency?
4. How would other factors such as gender, personality, and context affect the frequency
of L2 use with Japanese ESL students?
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 59

REFERENCES

Arbuckle, J. L., & Wothke, W. (1999). Amos 4.0 user’s guide. Chicago: Small Waters.
Baker, S. C., & Macintyre, P. D. (2000). The role of gender and immersion in
communication and second language orientations. Language Learning, 50(2), 311-
341.
Barraclough, R. A., Christophel, D. M., & McCroskey, J. C. (1988). Willingness to
communicate: A cross-cultural investigation. Communication Research Reports, 5,
187-192.
Brown, J. D. (1997). Questions and answers about language testing statistics: Skewness
and kurtosis. Shiken: JALT Testing & Evaluation SIG Newsletter, 1(1), 16-18.
Brown, J. D. (1996). Testing in language programs. Upper Saddle River, NJ: Prentice
Hall Regents.
Brown, J. D., Cunha, M. I. A., Frota, S. de F. N., & Ferreira, A. B. F. (2001). The
development and validation of a Portuguese version of the motivated strategies for
learning questionnaire..In Z. Dörnyei & R. Schmidt (Eds.), Motivation and second
language acquisition (Technical Report #23, pp. 257-275). Honolulu: University of
Hawaii, Second Language Teaching and Curriculum Center.
Brown, J. D., Robson, G., & Rosenkjar, P. R. (2001). Personality, motivation, anxiety,
strategies, and language proficiency of Japanese students. In Z. Dörnyei & R.
Schmidt (Eds.), Motivation and second language acquisition (Technical Report #23,
pp. 361-398). Honolulu: University of Hawaii, Second Language Teaching and
Curriculum Center.
Byrne, B. M. (2001). Structural equation modeling with Amos: Basic concepts,
applications, and programming. Mahwah, NJ: Lawrence Erlbaum Associates.
Chan, B., & McCroskey, J. C. (1987). The WTC scale as a predictor of classroom
participation. Communication Research Reports, 4, 47-50.
Chihara, T., & Oller, J. W. (1978). Attitudes and attained proficiency in EFL: A
sociolinguistic study of adult Japanese speakers. Language Learning, 28, 55-68.
Clément, R., Gardner, R. C., & Smythe, P. C. (1980). Social and individual factors in
second language acquisition. Canadian Journal of Behavioural Science, 12, 293-302.
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 60

Clément, R., & Kruidenier, B. G. (1983). Orientations on second language acquisition:

The effects of ethnicity, milieu, and their target language on their emergence.
Language Learning, 33, 273-291.
Clément, R., & Kruidenier, B. G. (1985). Aptitude, Attitude, and motivation in second
language proficiency: A test of Clément’s model. Journal of Language and Social
Psychology, 4, 21-37.
Crookes, G. & Schmidt, R. (1991). Motivation: Reopening the research agenda.
Language Learning, 41(4), 469-512.
Daly, J. (1991). Understanding communication apprehension: An introduction for
language educators. In E. K. Horwitz & D. J. Young (Eds.), Language anxiety: From
theory and research to classroom implications (pp. 3-13). Englewood Cliffs, NJ:
Prentice Hall.
Day, R. R. (1984). Student participation in the ESL classroom or some imperfections in
practice. Language Learning, 34(3), 69-102.
Dörnyei, Z. (1990). Conceptualizing motivation in foreign language learning. Language
Learning, 40, 45-78.
Dörnyei, Z. (1994). Motivation and motivating the foreign language classroom. The
Modern Language Journal, 78(3), 266-272.
Dörnyei, Z. (2001). Teaching and researching motivation. Harlow, England: Pearson
Education.
Ely, C. M. (1986). An analysis of discomfort, risktaking, sociability, and motivation in
the L2 classroom. Language Learning, 36(1), 1-25.
Gardner, R. C., & Lambert, W. E. (1959). Motivational variables in second language
acquisition. Canadian Journal of Psychology, 13, 266-272.
Gardner, R. C. (1980). On the validity of affective variables in second language
acquisition: Conceptual, contextual, and statistical consideration. Language Learning,
30, 255-270.
Gardner, R. C. (1985). Social psychology and second language learning: The role of
attitudes and motivation. London: Edward Arnold.
Gardner, R. C. (2001). Integrative motivation and second language acquisition. In Z.
Dörnyei & R. Schmidt (Eds.), Motivation and second language acquisition
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 61

(Technical Report #23, pp. 422-459). Honolulu: University of Hawaii, Second

Language Teaching and Curriculum Center.
Gardner, R. C., Day, J. B., & Macintyre, P. D. (1992). Integrative motivation, induced
anxiety, and language learning in a controlled environment. Studies in Second
Language Acquisition, 14(2), 197-214.
Gardner, R. C., & Gliksman, L. (1982). On “Gardner on affect”: A discussion of validity
as it relates to the Attitude/Motivation Test Battery: A response from Gardner.
Language Learning, 32, 191-200.
Gardner, R. C., Lalonde, R. N., Moorcroft, R., & Evers, F. T. (1987). Second language
attrition: The role of motivation and use. Journal of Language and Social
Psychology, 6, 29-47.
Gardner, R. C., & Macintyre, P. D. (1991). An instrumental motivation in language
study: Who says it isn’t effective? Studies in Second Language Acquisition, 13(1), 57-
72.
Gardner, R. C., & Macintyre, P. D. (1993). On the measurement of affective variables in
second language learning. Language Learning, 43(2), 157-194.
Gardner, R. C., Smythe, P. C., Clément, R., & Gliskman, L. (1976). Second-language
learning: A social psychological perspective. The Canadian Modern Language
Review, 32, 198-213.
Gardner, R. C., Tremblay, P. F., & Masgoret, A-M. (1997). Towards a full model of
second language learning: An empirical investigation. The Modern Language
Journal, 81, 344-362.
Gliksman, L., Gardner, R. C., & Smythe, P. C. (1982). The role of the integrative motive
on students’ participation in the French classroom. The Canadian Modern Language
Review, 38, 625-647.
Horwitz, E. K. (1986). Preliminary evidence for the reliability and validity of a foreign
language anxiety scale. TESOL Quarterly, 20, 559-562.
Horwitz, E. K. (2001). Language anxiety and achievement. Annual Review of Applied
Linguistics: Language and Psychology, 21, 112-126.
Horwitz, E. K., Horwitz, M. B., & Cope, J. (1986). Foreign language classroom anxiety.
The Modern Language Journal, 70, 125-132.
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 62

Horwitz, E. K., & Young, D. J. (1991). Language anxiety: From theory and research to
classroom implications. Englewood Cliffs, NJ: Prentice Hall.
Lalonde, R. N., & Gardner, R. C. (1984). Investigating a causal model of second
language acquisition: Where does personality fit? Canadian Journal of Behavioral
Science, 16, 224-237.
Macintyre, P. D. (1994). Variables underlying willingness to communicate: A causal
analysis. Communication Research Reports, 11, 135-142.
Macintyre, P. D., Baker, S. C., Clément, R., & Conrod, S. (2001) Willingness to
communicate, social support, and language-learning orientations of immersion
students. Studies in Second Language Acquisition, 23, 369-388.
Macintyre, P. D., & Charos, C. (1996). Personality, attitudes, and affect as predictors of
second language communication. Journal of Language and Social Psychology, 15, 3-
26.
Macintyre, P. D., Clément, R., Dornyei, Z., & Noels, K. A. (1998). Conceptualizing
willingness to communicate in a L2: A situational model of L2 confidence and
affiliation. The Modern Language Journal, 82(4), 545-562.
Macintyre, P. D., & Gardner, R. C. (1989). Anxiety and second language learning:
Toward a theoretical clarification. Language Learning, 39, 251-275.
Macintyre, P. D., & Gardner, R. C. (1991). Methods and results in the study of anxiety
and language learning: A review of literature. Language Learning, 47, 265-287.
Macintyre, P. D., MacMaster, K., & Baker, S. C. (2001). The convergence of multiple
models of motivation for second language learning: Gardner, Pintrich, Kuhl, and
McCroskey. In Z. Dörnyei & R. Schmidt (Eds.), Motivation and second language
acquisition (Technical Report #23, pp. 461-492). Honolulu: University of Hawaii,
Second Language Teaching and Curriculum Center.
Macintyre, P. D., & Noels, K. A. (1996). Using socialpsychological variables to predict
the use of language learning strategies. Foreign Language Annals, 29(3), 373-
386.
Macintyre, P. D., Noels, K. A., & Clément, R. (1997). Biases in self-ratings of second
language proficiency: The role of language anxiety. Language Learning, 47, 265-287.
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 63

Masgoret, A-M., Bernaus, M., & Gardner, R. C. (2001). Examining the role of attitudes
and motivation outside of the formal classroom: A test of the mini-AMTB for
children. In Z. Dörnyei & R. Schmidt (Eds.), Motivation and second language
acquisition (Technical Report #23, pp. 281-295). Honolulu: University of Hawaii,
Second Language Teaching and Curriculum Center.
McCroskey, J. C. (1992). Reliability and validity of the willingness to communicate
scale. Communication Quarterly, 40, 25-26.
McCroskey, J. C., Fayer, J., & Richmond, V. P. (1985). Don’t speak to me in English:
Communication apprehension in Puerto Rico. Communication Quarterly, 33, 185-
192.
McCroskey, J. C., & Richmond, V. P. (1990). Willingness to communicate: Differing
cultural perspectives. Southern Communication Journal, 56, 72-77.
Norton, B. (1997). Language, identity, and the ownership of English. TESOL Quarterly,
31(3), 409-429.
Norton Peirce, B. (1995). Social identity, investment, and language learning. TESOL
Quarterly, 29(1), 9-31.
Oller, J. W. (1981). Research on the measurement of affective variables: Some remaining
questions. In R. W. Andersen (Ed.), New dimensions in second language acquisition
research (pp. 14-27). Rowley, MA: Newbury House.
Oller, J. W. (1982). Gardner on affect: A reply to Gardner. Language Learning, 32, 183-
189.
Oller, J. W., Baca, L., & Vigil, F. (1977). Attitudes and attained proficiency in ESL: A
sociolinguistics study of Mexican Americans in the Southwest. TESOL Quarterly, 11,
173-183.
Oller, J. W., Hudson, A. J., & Liu, P. F. (1977). Attitudes and attained proficiency in
ESL: A sociolinguistic study of native speakers of Chinese in the United States.
Language Learning, 27, 1-27.
Oller, J. W. & Perkins, K. (1978b). Intelligence and language proficiency as sources of
variance in selfreported affective variables. Language Learning, 28, 85-97.
Oller, J. W. & Perkins, K. (1978a). A further comment on language proficiency as a
source of variance in certain affective measure. Language Learning, 28, 417-423.
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 64

Scovel, T. (1978). The effect of affect: A review of the anxiety literature. Language
Learning, 28, 129-142.
Schmidt, R. (1983). Interaction, acculturation, and the acquisition of communicative
competence: A case study of an adult. In S. Wolfson & E. Judd (Eds.),
Sociolinguistics and language acquisition. Rowley, MA: Newbury House.
Schumann, J. H. (1986). Research on the acculturation model for second language
acquisition. Journal of Multilingual and Multicultural Development, 7(5), 379-392.
Seliger, H. W. (1977). Does practice make perfect?: A study of interaction patterns and
L2 competence. Language Learning, 27, 263-278.
Steinberg, F. S., & Horwitz, E. K. (1986) The effect of induced anxiety on the denotative
and interpretive content of second language speech. TESOL Quarterly, 20, 131-136.
Swain, M. (1998). Focus on form through conscious reflection. In C. Doughty and J.
Williams (Eds.), Focus on form in classroom second language acquisition (pp. 64-
81). Cambridge: Cambridge University Press.
Swain, M., & Lapkin, S. (1995). Problems in output and the cognitive processes they
generate: A step towards second language learning. Applied Linguistics, 16(3), 371-
391.
Syed, Z. (2001). Notions of self in foreign language learning: A qualitative analysis. In Z.
Dörnyei & R. Schmidt (Eds.), Motivation and second language acquisition
(Technical Report #23, pp. 127-148). Honolulu: University of Hawaii Second
Language Teaching and Curriculum Center.
Teitelbaum, H., Edwards, A., & Hudson, A. (1975). Ethnic attitudes and the acquisition
of Spanish as a second language. Language Learning, 25, 255-266.
Tremblay, P. F., & Gardner, R. C. (1995). Expanding the motivation construct in
language learning. The Modern Language Journal, 79, 505-518.
Ushioda, E. (2001). Language learning at university: Exploring the role of motivational
thinking. In Z. Dörnyei & R. Schmidt (Eds.), Motivation and second language
acquisition (Technical Report #23, pp. 93-125). Honolulu: University of Hawaii,
Second Language Teaching and Curriculum Center.
Yashima, T. (2002). Willingness to communicate in a second language: The Japanese
EFL context. The Modern Language Journal, 86(1), 54-66.
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 65

APPENDIX
SURVEY QUESTIONNAIRE FORM

DIRECTIONS: Please indicate your opinion after each statement by putting an X that
best describes the extent to which you believe the statement applies to you.
1. If I were to rate my feelings about learning English in order to interact with members
of the second language community, I would say it is:
Weak____:____:____:____:____:____:____Strong
2. If I were to rate my interest in foreign languages, I would say that it is:
Very Low____:____:____:____:____:____:____Very High
3. If I were to rate my attitude toward members of the second language community, I
would say that it is:
Unfavorable____:____:____:____:____:____:____Favorable
4. If I were to rate my attitude toward my second language instructor, I would say that it
is:
Unfavorable____:____:____:____:____:____:____Favorable
5. If I were to rate my attitude toward my second language course, I would say that it is:
Unfavorable____:____:____:____:____:____:____Favorable
6. If I were to rate how hard I work at learning English, I would characterize it as:
Very Little____:____:____:____:____:____:____Very Much
7. If I were to rate my desire to learn English, I would say that it is:
Very Low____:____:____:____:____:____:____Very High
8. If I were to rate my attitude toward learning English, I would say that it is:
Unfavorable____:____:____:____:____:____:____Favorable
9. If I were to rate how important it is for me to learn English for employment, I would
say that it is:
Very Low____:____:____:____:____:____:____Very High
10. If I were to rate my anxiety in my second language class, I would rate myself as:
Very Calm____:____:____:____:____:____:____Very Nervous
11. If I were to rate my anxiety when speaking English, I would rate myself as:
Very Calm____:____:____:____:____:____:____Very Nervous
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 66

DIRECTIONS: Below are 20 situations in which a person might choose to communicate

or not to communicate. Please presume that you have completely free choice to initiate or
avoid communication. Please indicate in the space at the left the percentage of times you
would choose to communicate in English in each type of situation.
0 %= never, 100 %= always
1. Talk with an acquaintance in an elevator.
2. Talk with a stranger on the bus.
3. Speak in public to a group (about 30 people) of strangers.
4. Talk with an acquaintance while standing in line.
5. Talk with a salesperson in a store.
6. Talk in a large meeting (about 10 people) of friends.
7. Talk with a janitor/resident manager.
8. Talk in a small group (about 5 people) of strangers.
9. Talk with a friend while standing in line.
10. Talk with a waiter/waitress in a restaurant.
11. Talk in a large meeting (about 10 people) of acquaintances.
12. Talk with a stranger while standing in line.
13. Talk with a shop clerk.
14. Speak in public to a group (about 30 people) of friends.
15. Talk in a small group (about 5 people) of acquaintances.
16. Talk with a garbage collector.
17. Talk in a large meeting (about 10 people) of strangers.
18. Talk with a librarian.
19. Talk in a small group (about 5 people) of friends.
20. Speak in public to a group (about 30 people) of acquaintances.
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 67

DIRECTIONS: Below are 12 situations in which you might need to communicate.

People’s abilities to communicate effectively vary a lot and sometimes the same person is
more competent to communicate in one situation than in another. Please indicate how
competent you believe you are in communicating in English in each of the situations
described below. Indicate in the space provided at the left of each item your estimate of
your competence.
Presume 0 %= completely incompetent and 100 % = completely competent
1. Speak in public to a group (about 30 people) of strangers.
2. Talk with an acquaintance.
3. Talk in a large meeting (about 10 people) of friends.
4. Talk in a small group (about 5 people) of strangers.
5. Talk with a friend.
6. Talk in a large meeting (about 10 people) of acquaintances.
7. Talk with a stranger.
8. Speak in public to a group (about 30 people) of friends.
9. Talk in a small group (about 5 people) of acquaintances.
10. Talk in a large meeting (about 10 people) of strangers.
11. Talk in a small group (about 5 people) of friends.
12. Speak in public to a group (about 30 people) of acquaintances.
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 68

DIRECTIONS: Below are 12 situations in which you might need to communicate.

Please indicate how nervous you believe you will feel about communicating in English in
each of the situations described below. Indicate in the space provided at the left of each
item the percentage of time you would feel nervous.
Presume 0 %= I would never feel nervous and 100 %= I would always feel nervous
1. Speak in public to a group (about 30 people) of strangers.
2. Talk with an acquaintance.
3. Talk in a large meeting (about 10 people) of friends.
4. Talk in a small group (about 5 people) of strangers.
5. Talk with a friend.
6. Talk in a large meeting (about 10 people) of acquaintances.
7. Talk with a stranger.
8. Speak in public to a group (about 30 people) of friends.
9. Talk in a small group (about 5 people) of acquaintances.
10. Talk in a large meeting (about 10 people) of strangers.
11. Talk in a small group (about 5 people) of friends.
12. Speak in public to a group (about 30 people) of acquaintances.
 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 69

DIRECTIONS: Below are 12 situations in which you might need to communicate.

Please indicate how frequent you believe you will communicate in an English classroom
in each of the situations described below. Indicate by putting an X that best describes the
extent of your estimate of your frequency of communication.
1. Speak in public to a group (about 30 people) of strangers.
Never____:____:____:____:____:____:____Many, many times
2. Talk with an acquaintance.
Never____:____:____:____:____:____:____Many, many times
3. Talk in a large meeting (about 10 people) of friends.
Never____:____:____:____:____:____:____Many, many times
4. Talk in a small group (about 5 people) of strangers.
Never____:____:____:____:____:____:____Many, many times
5. Talk with a friend.
Never____:____:____:____:____:____:____Many, many times
6. Talk in a large meeting (about 10 people) of acquaintances.
Never____:____:____:____:____:____:____Many, many times
7. Talk with a stranger.
Never____:____:____:____:____:____:____Many, many times
8. Speak in public to a group (about 30 people) of friends.
Never____:____:____:____:____:____:____Many, many times
9. Talk in a small group (about 5 people) of acquaintances.
Never____:____:____:____:____:____:____Many, many times
10. Talk in a large meeting (about 10 people) of strangers.
Never____:____:____:____:____:____:____Many, many times
11. Talk in a small group (about 5 people) of friends.
Never____:____:____:____:____:____:____Many, many times
12. Speak in public to a group (about 30 people) of acquaintances.
Never____:____:____:____:____:____:____Many, many times

(THE END. THANK YOU VERY MUCH)

 Hashimoto – Motivation and Willingness to Communicate as Predictors of L2 Use 70

Yuki Hashimoto
Department of Second Language Studies
1890 East-West Road
Honolulu, HI 96822

[email protected]