Language Learning 44:3, September 1994, pp.

4 1 7 4 4 8

Motivation, Self-confidence, and

Group Cohesion in the Foreign Language
Classroom
Richard Cle'ment
University of Ottawa
Zoltan Dornyei Kimberly A. Noels
Eotvos University University of Ottawa

Defining t h e motivational basis of second a n d foreign

l a n p \ a g e acquisition has been at the center of much re-
search a n d controversy for many years. T h e present study
applied social psychological constructs to the acquisition of
English in the unicultural Hungarian setting. A total of
301 Grade 11 students from t h e region of Budapest an-
swered a questionnaire assessing t h e i r attitude, anxiety,
a n d motivation toward learning English, as well as their
perception of classroom atmosphere a n d cohesion. In
addition, t h e i r teachers rated each of the students on
proficiency and a number of classroom behaviors and
evaluated the relative cohesion of each class group. Factor

This research was supported in part by a grant from the Social Sciences and
Humanities Research Council of Canada to R. Clement and by a grant from
the Runganan Scientific Foundation to 2. Dornyei. We express our gratitude
to the teachers and students of the Budipest Schools who participated in this
study, to Emese Kopphy, E1er.a Mihu, Nelli SzakBcs, and Ildik6 Szigeti for
their assistance in collecting and analyzing the data, and to Peter MacIntyre
for his comments on a previous version of this paper.
Correspondence regarding this article should be addressed to Richard
Ckment, School of Psychology, University of Ottawa, Ottawa, Ontario, K1N
6N5,Canada. E-mail: rxcchQacadvm 1.uottawa.ca

417
418 Language Learn ing Vol. 44, No. 3

and correlational analyses of the results revealed that

xenophilic (M=4.22 on a 1-6 scale), sociocultural (M=3.96),
instrumental (M=3.78),and media-use reasons (M=3.79)
were most strongly endorsed by the students whereas an
identification orientation (M=1.8l)was rejected. Factor
analysis of the attitude, anxiety, and motivation scales
confirmed the existence of attitude-based (integrative mo-
tive) and self-confidence motivational subprocesses and
revealed the presence of a relatively independent class-
room based subprocess,characterized by classroom cohesion
and evaluation. Correlational analyses of these clusters
further revealed that, while all subprocesses were associ-
ated with achievement, self-confidenceand anxiety showed
no relationship to classroom atmosphere. We discuss these
findings in the context of current theories of second and
foreign language acquisition and with reference to their
applied implications.

At the beginning ofthe 199Os,two reviews ofthe literature on

the role of motivation in second language (L2)acquisition (Crookes
& Schmidt, 1991; Skehan, 1991) called for a new approach t h a t
would be more pertinent t o L2 teaching. In Crookes and Schmidt’s
concluding words, “we seek t o encourage a program of research
that will develop from, and be congruent with the concept of
motivation that teachers are convinced are critical for SL success”
(p. 502). Increasing the classroom relevance of motivation re-
search is certainly a worthwhile objective; at the same time, we
contend that achieving such a goal entails an awareness of the
social correlates of pedagogical interventions.
The following study therefore seeks t o reiterate the perti-
nence of a social psychological perspective t o L2 learning (Gardner,
1985; Gardner & Ckment, 1990)within the isolated context of the
foreign language classroom in a unicultural context. Such a
context can be found in Hungary, where in 1990, according to the
official figures ofthe Hungarian Central Statistical Office (Kozponti
Statisztikai Hivatal, 19921, 97.8% of the population were ethnic
Hungarians and the proportion with Hungarian as their mother
tongue was even higher (98.5%). Moreover, less than 9% of the
 Cle'ment, Dornyei, and Noels 419

population claimed t o speak any foreign language. Thus, person-

to-person contact with native Anglophones was minimal and
many students saw English as an ordinary school subject. At the
same time, one must recognize that contact with English language
and culture through the media and through the use of high-
technology devices such as computers was significant and that
English was widely recognized as the lingua franca of interna-
tional communication.

A Social Psychological Perspectiue

The social psychological perspective applied to this partictlar

context borrows from two research traditions. First, following the
work of Gardner and Lambert (1972; Gardner, 19851, it considers
the role of orientations and attitudes as affective correlates of L2
behavior and proficiency. Second, borrowing from Cldment (1980,
19861, i t assesses the role of linguistic self-confidence, including
language anxiety. As a third and novel element, o u r study
integrates aspects of 'group dynamics a s applied t o the classroom
setting.

Orientations and Motivations

Following studies conductedic the United States and Canada,

Gardner and Lambert (1972)suggested that motivation to learna
second language was grounded in positive attitudes toward the L2
community and in a desire t o communicate with and become
similar t o valued members of that community. This latter desire
became known as an integrative orientation, whereas attitudes
became the cornerstone of the integrative motive. Although
interrelated as socially bound prbcesses, these two aspects were
the objects of relatively independent research efforts (cf. Gardner
& Cldment, 1990).
As well as the integrative orientation, Gardner and Lambert
(1972) defined an instrumental orientation associated with a
desire t o learn the L2 for pragmatic gains. Although the original
 Language Learning Vol. 44, No. 3

formulation and measurements may have looked at the two

orientations as opposite ends of a continuum, researchers fourid
(cf. Clement, Gardner, & Smythe, 1977; Gardner & Smythe, 1975)
that both orientations correlated positively and, furthermore,
related positively t o a variety of other attitudinal measures. Thus,
in spite of continued attempts t o present them as antithetical (e.g.,
Soh, 19871, both orientations were shown very early to be posi-
tively related, affectively loaded goals that can sustain motivation.
Nevertheless, studies have observed much variation in the
relationship between the two orientations and indices of motiva-
tion and I 2achievement (e.g.,Gardner & MacIntyre, 1991;Samimy
& Tabuse, 1992). Prompted by the controversy that developed
around the usefulness of these constructs (Gardner 1980; Oller &
Perkins, 19781, Clement and Kruidenier (1983) investigated the
endorsement of reasons for learning second and foreign languages
by various groups of learners differing in t h e degree of
multiculturality of their immediate environments. Their results,
and those obtained in other similar studies (e.g., Moi‘se, Clement,
& Noels, 1990; Noels & Ckment, 1989), demonstrated that one
cannot simply assume cross-cultural pervasiveness of the integra-
tive and instrumental orientations (cf. Skehan, 1991). L2 learning
goals, traditionally lumped together in a general “affective-identi-
fication” kind of integrative tendency, proved to break up into
different orientation clusters, the definition of which varied de-
pending upon the sociocultural setting in which the data were
gathered (e.g., Fahmy & Bilton, 1992; Miiialjevic , 1990).
Most of the groups ’studied with Clkment and Kruidenier’s
(1983)approach had a t least a minimal amount of extracurricular
contact with members of the target language group. In a further
extensioR of Clement and Kruidenier’s approach, Dornyei (1990a)
contended that “foreign language learning” in a classroom setting
could not logically involve attitudes toward the L2 community,
because learners have little or no contact with members of the L2
group. Hungarians would learn English as a lingua franca to link
them with the rest of the world, facilitating trade and travel and
conveying international knowledge and cultural products. Dornyei
 Clement, Uornyei, and Noels 421

hypothesized that, in such contexts, the instrumental orientation

may acquire a special importance, and the individual’s LZ-related
affectively based motivation would be determined by a more
general disposition toward language learning and the values the
L2 conveys rather than ethnocultural attitudes toward the L2
community.
Domyei’s ( 1990a) results, although generally supporting
Clement and Kruidenier’s (1983; see also Gardner, Day, &
MacIntyre, 19921, showed that instrumental goals indeed played
a prominent role in the learning of English up to an intermediate
level. However, learners whose interest in learning English
included sociocultural and nonprofessional reasons demonstrated
the highest degree of desired proficiency; that is, they wished to
master the English lqnguage rather than merely to acquire a
minimal working knowledge of it. Furthermore, the desire to
spend an extended period.of time abroad was related to both
instrumental and sociocultural orientations. More socioculturally
based orientations, therefore, also appear t o be important for
motivation and, together with the instrumental orientation, are
related to a desire for contact with Anglophones.
Thus, even in a context where foreign language learning is
largely an academic matter, student motivation remains socially-
grounded. In Dornyei’s (1990a) study, the students were adult
learners who had voluntarily registered in and paid for English
courses. Whether the orientation pattern obtained by Dornyei-
particularly the prominence.ofthe instrumental orientation-was
a function of this specific sample or of the more general foreign
language learning environmenb remains t o be seen. This is the
first goal of o u r study.

An Integrative Motive

Domyei’s ( 1990b) results concerning orientations could be

interpreted t o show the presence of a strong affective component
because of the relationship t o a desire for contact with foreigners
and Anglophone culture. Gardner (1985; Gardner & Clbment,
422 Language Learning V O ~44,
. NO.3-

1990; also Skehan, 1989) summarized studies that underline the

motivational effects of positive regard toward the L2 community
(known as the integrative rnotiue) not only on L2 achievement but
also on a family of language-related phenomena: persistence in L2
study, level of parental encouragement, L2 maintenance, class-
room participation and seeking contact with members of the L2
group. In fact, a follow-up t o Dornyei's initial study (Dornyei,
1990b)showed that intended contact with target language speak-
ers was sigdcantly related to affectively based motivation. Neither
study (i.e., Domyei, 19QOa,1990b1, however, used a full-fledged
attitudinal test battery of the type used in the North American
studies. Identifying the components of and assessing the impor-
tance of the integrative motive for L2 proficiency in the present
context is, therefore, the second goal of this study.

Self-Confidence and Anxiety

The above discussion presumes important contextual varia-

tions in the structure and importance of affective predispositions.
Although not the only approach concerned with contextual effects
(cf. Gardner, 1985, Ch. 81, Clement's model of second language
acquisition may be particulariy relevant to t h e present situation
(Clement, 1978,1980,1984,1986; Clement & Kruidenier, 1985).
Following Gardner and Lambert (19721, Clement argued that
attitudinal factors were an important motivational basis for L2
acquisition and behavior. Upon noting the results obtained with
groups of students who were in more direct contact with the L2
group, however, he suggested that in such contexts a self-confi-
dence process becomes the most important determinant ofattitude
and effort expended toward L2 learning.
In a multiethnic context, positive attitudes would orient the
individual to seek contact with members of the L2 community. To,
the extent that this contact is relatively frequent and pleasant,
self-confidence in using the L2, operationally defined in terms of
low anxious sffect and high self-perceptions of L2 competence'
would,develop. This being the case, the availability of extracur-
 Clement, Dornyei, and Noels 423

ricular contact provides the conditions for the development of a

motivational process which is relatively independent of the a&tu-
dinal process delineated in previous research and theorizing.
Because the Hungarian context is for the most part unilingual and
unicultural, this self-confidence process would not be expected.
However, this prediction could be erroneous for two reasons.
First, interethnic contact can also be made in a‘more remote
manner, through the media or through travel outside the country.
Second, even though self-confidencewas related t o qualitative and
quantitative aspects ofinterethnic contact (Clkment, 1986;Clkment
& Kruidenier, 1985; Labrie & Clkment, 19861,research and theory
have suggested that anxiety and self-perceptions ofL2 competence
may be determinants of L2 achievement even in contexts where
opportunity t o use the L2 with members of the L2 community is
lacking. Horwitz, Horwitz, and Cope (1986) suggested, for ex-
ample, that certain types of classroom actiyities may promote
language anxiety, particularly thdse that expose the students to
negative evaluations by the teacher or by peers. (Seealso Foss 8z
Reitzel, 1988; Horwitz, 1986; Horwitz & Young, 1991; MacIntyre
& Gardner, 1991a, 1991b; Phillips,T992.) Similarly, Bailey’s
(1983) analysis of the diaries of L2 learners linked the type of
rapport established with the teacher and other students to their
experience of anxiety in the L2 classroom. (See also Cohen &
Norst, 1989;Price, 1991;Young, 1986.) It is therefore possible that
anxiety and, by extension, self-confidence-inthe L2 classroom are
intimately linked t o classroom processes.

The Role of Classroom Dynamics

The above discussion highlights the potqntial relevance of

classroom-related factors, and research has indeed confirmed that
situation-specific factors significantly contribute to L2 motivzkion
in a foreign-language classroom context (Julkunen, 1989). But
how can the complexity of the L2 classroom environment be
operationalized for inclusion into a research paradigm; that is, on
which classroom processes and variables shouldwe focus? Gardner’s
424 Language Learning Vol. 44, No. 3

(1985) socioeducational model and his Attituddvfotivation Test

Battery included the evaluation of the L2 teacher and the L2
course as key components of L2 instruction. Given Honvitz et al.'s
(1986) results on the genesis of language anxiety, the social
processes and dynamics of the classroom probably play a n impor-
tant motivational role.
Group dynamics has been a cqre area of social psychology for
a number ofyears. (For a review, see Brown, 1988;Forsyth, 1990;
Shaw, 1981.) It concerns the scientific analysis ofthe dynamics of
small group behavior, focusing on issues such as group formation
and development, group' structure and group processes. Although
group dynamics has considerable implications for education in
general, for the obvious reason that most institutional teaching goes
on with learners organized in small groups, it has particular rel-
evance to L2 instruction. Current language teachmg methodologies
aim a t developing the learners' communicative competence by pro-
moting classroom interaction between learners as they participate in
communicative events. The quality and quantity of such interaction
is, t o a large extent, a function ofthe social structure and milieu ofthe
class (cf Bar-Tal & Bar-Tal, 1986; Hadfield, 1992; Prabhu, 1992).
One concept central t o the explanation ofmany group-related
phenomena is group cohesion, or "the strength ofthe relatimship
linking the members t o one another and t o the group itself"
(Forsyth, 1990, p. 10). It is taken an index of the level of group
development, directly related t o within-group cooperation and t o
both the quality and quantity of group interaction (cf. Greene,
1989; Shaw, 1981). Evans and Dion's (1991) meta-analysis of
studies addressing the relationship between group cohesion and
group performance found a significant positive relationship be-
tween the two variables, indicating that cohesive groups, on
average, tend t o be more productive than noncohesive groups.
Within the present context, we therefore assumed that cohe-
sion would be closely related t o the evaluation of the learning
environment, and, by extension, t o lower anxiety and higher self-
confidence. Assessing these relationships is the fourth goal of o u r
study.
 Cle'ment, Dornyei, and .Noels 425

In summary, we attempted t o assess the relevance of social

psychological factors, showo t o be influential motivational aspects
in other L2 learning contexts, t o an emerging foreign-language
teaching context marked by the virtual absence of the target
language group. Of particular interest were (a) the emergence of
orientations germane to this context, (b) the role and nature of
affective predispositions in it, (c) the role and nature of self-
confidence,and (d)the relative importance of classroom dynamics.

Method

Participants
The participants were 301 (117 males; 182 females; 2 with
missing gender tlata) secondary-school pupils (in 21 groups) regis-
tered in 11 different schools in various parts of Budapest. Special
care was taken t o select a mixture of schools in terms of both
prestige and location. The participating students were all in the
third (second last) year of their secondary school instruction
(gimnaziurn;ages 17-18) and all studied English a s part of their
official school curriculum, which was the same for all groups (i.e.,
no classes with specialized or intensive EFL syllabi were included).
They evaluated the English proficiency of their mothers and fathers
as vi&ually nil (median=2, on an 8-point scale, in both cases).

Materials

The materials consisted of 8 questionnaire addressed to the

students and a .questionnaire-saddressed t o the teachers. The
scales used in both questionnaires included some items adapted
fro& scales used in previous studies of second language learning
(e.g., Clement, 1986; Dornyei, 1990a) and some items developed
for the present study. Where necessary, we translated scales into
Hungarian. A description of the scales contained in each section,
along with the Cronbach alpha (a)index of internal consistency,
where appropriate, follows.
426 Language Learning .Vol. 44, No. 3

Students’ Questionnaire

Orientations, attitudes, motivation and anxiety. In the first

section,a variety of orientations, attitudes, and anxiety constructs
were assessed through randomly arranged items. Students indi-
cated the extent t o which they agreed or disagreed with the
statement using a 6-point Likert-type scale. The scales were
anchored a t one end by disagree strongly and a t the other end by
agree strongly, such that a high score indicated strong endorse-
ment ofthe item. We reversed negativeljr worded items prior to the
calculation of the scale score.
1. Orientations. .A total of 27 items adapted from relevant
studies (Clement & Kruidenier, 1983; Dornyei, 1990a)were
used t o assess students’ reasons for learning English. They
tapped instrumental, integrative, travel, knowledge, friend-
ship, and sociocultural orientations.
2. Attitude Toward Learning English. Five items, two posi-
tively and three negatively worded, assessed the student’s
affectivereaction toward learning the second language (a=.75).
3. Attitude Toward the British. Five items evaluated the extent
to which the student felt positively toward British cithens
(a=.74).
4. Attitude Toward the Americans. Five items evaluated the
extent to which the student felt positively toward American
citizens (a=.85).
5 . English Use Anxiety. Anxiety experienced while using the
second language outside the classroom was assessed through
four items, two positively and two negatively worded (a=.83).
For the purpose of o u r analyses, the scoring of this scale was
reversed so that a high score means little anxiety.
6 . English Class Anxiety. Five items assessed the extent to
which the student felt anxious during English class (a=.86).
For the purpose of our analyses, the scoring of this scale was
reversed s o that a high score means little anxiety.
7. Satisfaction. The student’s satisfaction with his/her work in
English class and.hidher level of proficiency in English was
 Cle‘ment, Dornyei, and Noels 427

assessed with two 6-point scales anchored a t one end by

absolutely not and a t the other by definitely yes (a=.62).
8. Perceived Group Cohesion. The student assessed the extent
t o which s h e felt that the language class was unified as a
cohesive group using eight items, three positively wordedand
five negatively worded, that were based on measures devel-
oped by Evans, Jarvis, and Dawson (1986)and Rosenfeld and
Gilbert (1989). Examples of these items include “I think that
some people in this group feel left out” and “I am dissatisfied
with my group” (a=.77).
9. Need for Achievement. Four items, two negatively worded
and two positively worded, assessed the extent to which the
students desired to achieve or to perfect their skills in the
second language (a=.63).
10. Motivational Intensity. Four items, two negatively worded
and two positively worded, were used jjo determine the degree
of effort the student exerted when le‘grning English (a=.68).
11. Desired English Proficiency. The student’s desired level of
English proficiency was assessed on a 7-point scale, as fol-
lows:
elementary -intermediate advanced
0 0 0 0 0 0 0
Answers t o this scale were coded from 1 t o 7, with 1 corre-
sponding to the low end of the scale.
Frequency and quality of contact with members of the second
language group. Considering nine different situations (e.g., at
school, on holidays, when corresponding), the participant was
asked t o indicate, on two g-point,scales, first, the frequency, and,
second, the quality, of contact with members of the English-
speaking community. The scales ranged from very rare contact to
very frequent contact, on one hand, and from very unpleasant
contact t o very pleasant contact, on the other hand. A high score
on either scale indicates very frequent (a=.61)and very pleasant
(a=.67)contact with English speakers. 0

English teacher evaluation. The student evaluated the En-

428 Language Learning lol. 44, No. 3

glish teacher using twelve 7-point, bipolar scales that assessed the
teacher’s competence (two items; a=.77), rappoi-t (three items;
a=.84), motivation (two items; a=.71)and teaching styldpersonal-
ity (five items; a=.78)
Bnglish course evaluation. The student evaluated the En-
glish course using seven 7-point, bipolar scales that assessed its
attractiveness (three items; a=.82),difficulty .(one item, reversed
scoring) and relevancdusefulness (two items; a=.83).
Self-evaluation of English language ability. Using four 6-
point scales, participants indicated how well they could use English,
in terms of reading, writing, speaking, and comprehension. A high
score suggests that the individual feels very competent in English
(a=.79).

Teachers’ Questionnaire

We asked the students’ teachers t o use four 7-point scales to

evaluate the quality of each student’s homework in general, how
active and how motivated the student was in class, a s well as the
student’s status among.his/her peers. As well, using two 5-point
scales, the teachers assessed each student’s current level of achieve-
ment in active communication and in passive, theoretical knowledge
ofEnglish. The teachers also provided the student’s English grade
from the previous term.
In addition, t o obtain an alternative index of the group
dynamics of the classroom, we asked the teachers t o rate the
relative cohesion and cooperativeness of each class group on a 7-
point scale.

Procedure

The administration of the questionnaire took place in the

students’ English classes. All participating students received a
letter requesting their cooperation one week prior to the adminis-
tration of the test. Because the questionnaire included sensitive
information (e.g., the students’ evaluation of their teachers), the
’confidentialityof the students’ responses was ensured by sealing
 Cldment, Dornyei, and Noels 429

the completed questionnaires in stamped envelopes. A Hungarian

EFL student from Eotvos University in Budapest then collected
the envelopes. In order to match the students’ responses with the
teachers’ evaluations of the students, a coding-system was devel-
oped based on the students’ seating arrangements.

Results and.Discussion

In order t o fulfill the stated goals of this study, we analyzed

the data using correlational and factor analytic methods. We now
report and discuss the results ofthese analyses in four subsections,
‘onefor each of the four goals: .(a).factorialstructure and level of
endorsement of orientations, (b>the factoiial structure of the
correlations between attitudes, achievement and self-confidence,
(c) the relationships betweep self-evaluation, anxiety, and aspects
of proficiency and classroom behavior, and finally, (d) the interre-
lationship between the preceding constructs and aspects of class
group dynamics..

Orientations

Following Clement and ICruidenier (1983) and Dornyei

(1990a), we first factor analyzed the orientation items in order to
delineate clusters that would define orientations in this particular
context. The ratio of observations t o items was more than satisfac-
tory (11.15:l; see Tabachnick 8z Fidell; 1989, p. 603). We used
exploratory factor analysis (EFA) because the dual origin of the
items did not permit the predictiin of a factor structure a d usage
of confirmatory factor analysis (CFA). We therefore applied
SPSSx Maximum-likelihood extraction, followed with Oblimin
rotation, to the correlations between the responses to the orienta-
tion items2 Upon application of the Scree Test (Cattell, 19661, five
factors were retained, which accounted for 41.5% of the variance.
Although significant, the chi-square-to-degrees-of-freedom ratio
(442 68/226= 1.96)suggested that the .solution provides an acceptable

fit t o the data (Carmines & McIver, 1981). The pattern matrix,
430 Language Learning Vol. 44, No. 3

Table l a
Factor Analysis Summary of Orientation Items:
Oblimin Rotated Factor Pattern Matrix, Communalities (h2),and
E igen u a I ues

Factor
1 2 3 4 5 h2

Make friends' with foreigners .67

Meet foreigners .48
Know new foreign people .46
Will help when traveling .36
Keep in touch with foreign friends .38
Would like to travel .4 1
Learn many foreign languages .27
Without English, less travel ' .21
Thinkhehave as UWUS people - .81 - - - .71
Be similar t o U W S people - .78 - - - .60
Know various culturedpeoples - - .76 - - .71
Learn about English world - - .76 - - .55
Understand English nations - - .75 - - .61.
Know culturedworld events - - .72 - - .64
Know life of English nations - - .56 - - .46
Is part of b a n g educated - - - .70 - .51
To be more knowledgeable - - - .55 - .31
Without it-difficult to succeed - - - .48 - .24
To broaden my outlook - -3 .35 .38 - .38
May need latedjob, studies - - - .35 - .18
It is expected of me - - - .32 - .19
To take the State Language Exam .15
To understand filmdvideos - - - - .69 .53,
To understand pop music - - - - .54 .38
To read bookwmagazines - - - - .42 .35
To spend time abroad - - - -: .35 .26
Do not want bad marks - - - - .31 .19
Eigenvalue 5.97 1.53 1.62 1.17 0.92
Factor l=Xenophilic; Factor 2=Identification;Factor 3=Sociocultural;Factor
4=Instrumental-Knowledge; Factor 5=English Media
 Clement, Uornyei, and Noels 431

Table l b
Factor Analysis Summary of Orientation Items:
Factor Correlation Matrix

Factor
Factor 1 2 3 4 5

1. Xenophilic 1.00
2. Identification .16 1.00
3. Sociocultural .51 .24 1.00
4. In strumen tal-Kn ow 1edge 15 .07 .02 1.00
5. English Media .19 .ll .07 .23 1.00

sorted by factor, using a loading greater than .30 as a criterion of

factor salience (cf. Tabachnick & Fidell, 19891,appears in Table la.
As can be seen, Factor 1 receives appreciable loadings from
seven variables, the three most prominent ofwhich relatelearning
English t o making friends with foreigners. This factor therefore
corresponds t o a friendship orientation similar to that reported by
Clement and Kmidenier (1983) with the difference that this
tendency would concern foreigners in general rather than only
Anglophones. Therefore,.it seems best t o label this factor as
reflecting a Xenophilic orientation, similar to but more general
than, the “active sociocultural” dimension reported by Dornyei
( 1990a).
Factor 2 receives loadings from two variables which are
comparatively more focused. Both concern identification and
similarity with the American and the British. This factor.is
therefore best labeled as an Identification orientation.
The five variables loading onsFactor 3 relate learning English
to an interest in cultural aspects of the English‘world. Factor 3 is
similar t o the sociocultural orientations described by ‘bothClkment
and Kruidenier (1983) and Dornyei (1990a)and for that reason we
also label it as a Sociocultural dimension.
Factor 4 is defined by six items suggesting that being more
educated and knowledgeable is related t o success in work and
studies. It is therefore similar t o the instrumental orientations
432 Language Learning Vol. 44, No. 3

found by Clkment and Kruidenier (1983) and Dornyei (1990a),

with the provision that knowledge is an important component of
pragmatic attainment. Therefore we label this factor as a n
Instrumental-Knowledge dimension.
Finally, Factor 5 is mostly defined by variables relating to the
use of English in the media. It is therefore labeled a s an English
Media factor. Factor 5 is similar to! although more general than,
the “reading for nonprofessional purposes” and the “passive socio-
cultural’’ dimensions identified by Dornyei (1990a).
The results show similarities and differences with those
reported previously. *he factor pattern collapses, on the one hand,
the instrumental, and knowledge orientations and, on the other
hand, the friendship and travel orientations obtained by Clement
and Kruidenier (1983). This solution reflects the particular con-
straints and context of English acquisition for these Hungarian
students. As a major language oftrade in the New Europe, English
as knowledge is perceived t o have pragmatic consequences. For
these Hungarian secondary-school students, pragmatic rewards
may appear quite remote; the wish t o prepare for a bright career
is related to getting higher qualifications, and thus to obtaining
knowledge. S G l a r l y , the unicultural Hungarian context only
permits crosscultural friendships in the context of travel. Finally,
the relative isolation of Hungarians may also be the reason for the
emergence of a sociocultural orientation. As i t was for Clkment
and Kruidenier’s (1983) students of Spanish, the English culture
may be different enough t o o u r Hungarian students t o justify a n
interest for its distinctive culture and people.
One striking difference from the solution obtained by Dornyei
(1990a) is the emergence here of an “identification” orientation.
We originally included the two items defining this factor in a n
integrative orientation scale together with friendship items. Ap-
parently, respondents did not concur with the theoretical
composition of this orientation and isolated the identification from
the affective component.
Two additional analyses contribute to the interpretation of
the identification factor. First, because we used an oblique rota-
 Clkment, Dornyei, and Noels 433

tion to rotate the factors, the correlation between them may be

indicative of relationships between the corresponding orienta-
tions. As can be seen in Table lb, the xendphilic orientation
appears t o be positively correlated t o the sociocultural orientation,
which is itself related t o a lesser degree with the identification
orientation. Thus, the cultural, friendship, and identification
components are related, as would be expected under the definition
of an integrative orientation. A modest relation also obtains
between the Instrumental-knowledge and the English media ori-
entations.
The second analysis considers the level of endorsement of
each orientation. We computed these by taking the average
endorsementl(minimum=l;maximum=6)for those items shownto
load on each factor (Gorsuch, 1983). We included items that cross-
loaded onto two factors in the index of the factor that they defined
most highly. The results show a relatively strong endorsement of
all orientations except one.. The xenophilic (M=4.22), media
(M=3.79),sociocultural (M=3.96)and instrumental (M=3.78)ori-
entations are endorsed strongly whereas identification (M=1.81)
receives only minimal support. Thus, despite its relation to
endorsed orientations, identification as a goal for learning English,
is rejected here. This may be a symptom of perceived profound
cultural divergence. The English language and the Anglophones
may appear culturally and amicably interesting, but there is a
definite limit t o the extent of the desired rapprochement. This
limit may be related t o how foreign the English language and
culture are considered. Even though English may evidence consid-
erable vitality as a new European language of trade (viz., o u r
instrumental Orientation), the Hyngarian cultural context may
still foster less-than-permeable intergroup boundaries (cf. Giles &
Byrne, 1982; Giles & Johnson, 1987) so that identification is
-simply not perceived as possible.
In summary, the results show that the Hungarian situation
is peculiar in terms of the orientations it promotes. The absence
of Anglophones from the immediate environment would appear to
sustain distal friendships (through travel), an interest in English
434 Language Learnihg VoL. 44, No. 3

culture as a foreign phenomenon, and an instrumental orientation

based on the acquisition af knowledge and media usage rather
than on the achievement of pragmatic outcomes. This relative
isolation may be common t o many foreign language learning
situations but, in this case, it is coupled with rather strong cultural
and linguistic tradihons which would result in the rejection of
identification as an outcome of language learning.

Attitudes, Self-confidence, and Achievement

We now turn t o the examination of the role of attitudes and

self-confidence in the present context. In order t o remain method-
ologically consistent with previous work (e.g., Gardner, 1985;
Clement & Kruidenier, 19851, we entered indices of effort, atti-
tudes, anxiety, self-evaluation, course and teacher evaluation, and
English achievement into a factor a n a l y ~ i sWe
. ~ also included the
orientation indices, as derived in the previous analysis. Again, the
ratio of observations t o items was more than satisfactory (11.15:l;
see Tabachnick & Fidell, 1989, p. 603). Following the Scree test,
we extracted five factors, using the maximum-likelihood proce-
dureandrotatedthemviaVarimax. Theresultingmatrix.accounted
for 48.9%of the variance. Although the chi-square is statistically
significant, the chi-square-to-degreesoffreedom ratio P745/27~=2.39)
suggests that this represents an acceptable fit (Carmines& McIver,
1981).
The solution appears in Table 2. Using a cutoff criterion of .30
(Tabachnick 8z Fidell, 1989, p. 639), Factor 1 is'defmed by six
variables all pertaining t o the student's evaluation of the English
teacher and course. It therefore seems best labeled as an Evalu-
ation of the Learning Environment dimension.
Factor 2 receives appreciable loadings from nine variables,
including indices of contact, anxiety, self-evaluation of compe-
tence, and motivation. Students who show little anxiety when
using English evaluate their own proficiency relatively positively,
are satisfied with their current level of English proficiency and
report positive and frequent contact with English. They are also
 Clement, Dornyei, and Noels 435

Table 2
’
Attitude, Achievement, and Motivation Indices: Varimax Rotated
Factor Matrix, Communalities (h2),and Eigenvalues

Factors
1 2 3 4 5 h2

Teacher’s style .94 - - - - .89

Teacher’s competence .80 - - - - .64
Course attractiveness .79 - - - - .66
Teacher’s motivation .72 - - - - .55
Course usefulness .64 - - - - .47
Teacher’s rapport .61 - - - - .42
Anxiety in class (reversed) - .81 - - -I .73
English use anxiety (reversed) - .77 - - - .63
Self-eval. of proficiency - .66 - - - .52
FreqGency of contact - .50 - - - .30
Attitudes t/w learning English - .49 - .46 - .53
Quality of contact - .46 - .39 - .43
Satisfaction with com pe tence - .41 - - - .32
Course difficulty (reversed) - .37 - - - .23
Desired E ngl i sh pro fi cien cy - .35 - - - .22
Student’s motivation - - . .87 - - .83
Student’s homework - - .75 - - .61
Student’s activity - - .73 - - .72
English achievement - - .69 - - .66
Student’s status - - .50 - - .39
Attitudes toward the British - - - .71 - .55
Motivation a1 in tensity - - - .59 - .44
Attitudes toward Americans - - - .51 - .31
Student’s need t o achieve - - - .39 - .18
Identification orientation - - - .36 - .15
Xenoph ilic orientation - - - - .68 .77
English media orientation - - - - .55 .31
Sociocultural orientation - - - - .51 .51
Instrumen tal-kn owledge - - - - .38 .18
Eigenvalue 6.32 3.35 2.16 1.48 0.84
Factor l=Xenophilic; Factor 2Jdentification; Factor 3=Sociocultural;Factor
4=Instrumental-Knowledge;Factor 5=English Media
436 Language Learning Vol. 44, No. 3

motivated to learn English and find the course easy. In keeping

with previouseesearch and theorizing where similar factors were
obtained (e.g. Clement, 1978,19861,this factor seems best labeled
as a Self-Confidence d i t h English dimension.
Factor 3 evidences appreciable loadings from five variables,
all concerned with the student’s English achievement and the
teacher’s rating of the student’s classroom performance. It there-
fore seems appropriately labeled as a Student Achievement and
Performance factor.
Factor 4 receives appreciable loadings from five variables,
including indices of attitudes towarh Anglophones, motivation
(motivational intensity and need t o achieve),and identification. In
the light of previous research and theorizing (Gardner, 19851,
particularly regardmg the co-occurrence of attitude and identifica-
tion indices, this factor would correspond t o an Integrative Motive.
Finally, Factor 5 is defined by four of the five orientation
scales and therefore seems best labeled as an Orientation dimen-
sion, with the provision that it does not include the identification
orientation which, as seen above, was associated with the integra-
tive motive.
In spite of a markedly different context, the results obtained
here are similar t o those obtained in North American studies. The
integrative motive emerged with a definition which indeed corre-
sponds closely to Gardner’s (1985)discussion of the construct. The
self-confidence-with-English cluster was also composed of van-
ables typi ally associated with the construct (e.g., Clement, 1980,
d
1986;C16 ent, Gardner & Smythe, 1977,1980).It should be aoted
that contrary t o Au’s (1987)critique, self-evaluations ofproficiency
are here associated with self-confidence(Factor 2) and not achieve-
ment as reported by teachers (Factor 314Furthermore, English-class
anxiety and course difficulty, which are associated in Gardner’s
(1985) socioeducational model with the learning environment, are
here associated wkh self-confidence. However, o u r factorial struc-
ture did not show crossloadings of variables defining the teaching
environment (Factor 1)with those defining self-confidence(Factor2).
We now turn to a more detailed examination of these relationships.
 Clkment, Dornyei, and Noels 437

Anxiety, Self-Confidence, and the Classroom

Because of the unexpected absence of a substantial relation

between self-confidence and the student's evaluation of tRe teach-
ing environment, we conducted a more detailed examination of
correlations in order t o delineate the relation between the compo-
nents of these constructs. Focusing on the anxiety indices also
provides a more common basis for compaiing these results and

Table 3
Correlations Between Components of Self-Confidence and
Classroom Variables

EUA" ECA" SEP sc

Student's evaluation of:
Teacher style .09 .02 .08 .ll
Teacher rapport .12* .ll .10 .15*
Teacher competence .06 -. 05 .02 .02
Teacher motivation .12* -. 02 .10 .ll
Course attractiveness .14* .08 .ll .15*
Course difficulty .26** .34** .28** .35**
Course usefulness .08 .07 .13* .12*
Teacher's rating of student's:
Communicative skills .36** .45** .49** .52**
Passive skills .22** .32** .33** .35**
Last grade .30** .38** .43** .43**
Quality of homework .ll .20** .16* .18*
Activity in classroom .37** .46** .43** .48**
Apparent motivation .33** .38** .36** .41**
Student status .30** .42** .33** .41**
Attitude and effort index .36** .37** .37** .45**
Frequency of contact .46** .33** .39** .45**
Quality of contact .45** .37** .40** -.48**
yeverse d
EUA=English Use Anxiety; ECA=English Class Anxiety; SEP=Self-Evalua-
tion of Proficiency; SC-Self-confidence
n=301
*p<.05, two-tailed. **pc.OOl, two-tailed.
438 Language Learning Vol. 44, No. 3

those of other studies. Table 3 therefore presents selected corre-

lations between the anxiety indices, self-evaluation and a composite
self-confidence indee aggregating anxiety and self-e~aluation,~
and different aspects of the evaluation of the teacher, the course,
student’s attitude and effort toward learning English, inter-group
contact and a teacher’s ratings of the student. ,We obtained the
attitude and effort index by first standardizing and then averaging
the motivational intensity and attitude toward learning English
scales.
As can be seen, very few of the students’ evaluations of the
teacher and the course correlate consistently and highly (pc.01)
with anxiety or self-evaluation. However, a notable exception is
the evaluation of course difficulty which, (as was shown in the
factor analysis)was related t o the self-confidencefactor. As for the
teachers’ rating of the students, both indices of proficiency (com-
municative and passive skills) are significantly related t o anxiety
and self-evaluation, with indices of communication proficiency
showing the strongest relationships. Furthermore, all aspects of
classroom participation and involvement in learning English
‘(withthe single exception of quality of homework) are related to
the anxiety indices. Finally, all indxes are correlated with the
attitude and effort index and aspects of interethnic contact. For
the latter, similarly to the proficiency and classroom-behavior
indices, the aggregation of anxiety and self-evaluation into a single
self-confidenceindex produces relatively enhanced relationships.
These results therefore buttress previous hypotheses linking
anxiety and self-evaluation with extracurricular contact (e.g.,
Clement & Kruidenier, 1985). They also support Phillips’ (1992)
conclusion regarding the relation between anxiety, attitudes, and
motivation towards learning the second language. Not only is
anxiety related t o motivation, but it is also related t o indices of
prokiency and more so t o communication proficiency, as sug-
gested by Young (1991)and Horwitz et al. (1986). It is also related
to what Crookes and Schmidt (1991) might consider teacher-recog-
nized symptoms ofmotivation and engagement:quality ofhomework,
apparent motivation, activity, and status in the classroom.
 Clement, Dornyei, and Noels 439

Our results, however, fail t o support current suggestions that

the source of anxiety lies in the student’s perception of certain
aspects of the classroom situation. Within the limits of the
measures used here, anxiety is linked neither to perceptions of the
teacher nor t o the course (with the exception of course difficulty).
Following o u r theoretical conjecture regarding the role of group
dynamics, that relationship may, however, be established through
class group cohesion. We now turn t o an examination of that
aspect within the context of the relationship Between the major
constructs.

Motivation and Group Dynamics

To assess the relation between the constructs definedthrough

the previous analyses, we calculated aggregated indices, using the
results of the previous factor. solution (Table 2). Thus we calcu-
lated indices for English Achievement, Evaluation of the Learning
Environment, Integrative Motive, Self-confidence with English,
and Orientations by first standardizing and then sllmming the
variables defining the corresponding factors (Gorsuch, 1983). We
did not include the two motivational indices, Motivational Inten-
sity and Attitude Toward Learning English in the scores derived
from their respective factors but, as for the previous analysis,
aggregated them t o constitute a single Attitude and Effort index.
In addition t o these six variables, two cohesion scores were consid-
ered: one derived from the student’s perception of classroom
cohesion and one corresponding t o the teacher’s evaluation of the
cooperation and cohesion of each class group. Thus, we calculated
correlations among eight variables.
The matrix resulting from this operation is appears in Table 4.
Focusing on the most important relations (i.e., those withp<.Ol),
English achievement is related significantly t o self-confidence,the
evaluation of the learning environment and the motivational
indices. In turn, the attitude and effort index is also related to self-
confidence, the learning environment, and a cluster of affectively
based attitudes and motivation. Following OUT conjecture, cohe-
440 Language Learning Vol. 44, No. 3

Table 4
Correlations Between Attitudes, Motivation, Self-Confidence, and
Aspects of the Classroom

E-A LE OR SC IM CS CT

Attitude & Effort Index .38** .32** .24* .47** .40** .ll .09
English Achievement . .32** .18* .53** .17* .OO,’ .26**
Learning Environment .03 .12* .08 .30** .62**
Orientations .23* .40**-.04 .02
Self-confidence .21* .02 .10
Integrative Motive .07 -.11
Cohesion (Student’s) .20*
Cohesion (Teacher’s) 1.00
EA=English Achievement; LE=Learning Environment; OR=Orientation;
SC=Self-Confidence; IM=Integrative Motive; CS=Cohesion (student’s rat-
ing)’ CT=Cohesion (teacher’s rating)
n=301
*pc.05, two-tailed. **pc.Ol, two-tailed.

sion was associated t o a positive perception of the learning envi-

ronment. Contrary to expectations, however, self-confidencewas
associated neither with evaluation of the learning environment
nor with the cohesion indices.
There appears t o be a close association between the evalua-
tion of the environment, the teacher’s appraisal of the group, and
the students’ evaluation of their own learner group. Thus the
tasks, the teacher, and the learner group are perceived as interde-
pendent aspects of classroom reality, significantly affecting student
L2 learning behavior. Gardner’s ( 1983) socioeducational model
has already incorporated attitudes toward the learning situation
(evaluation of the teacher and the course) as a component of the
integrative motive. Whereas o u r results do support the presence
of a learning environment cluster and its relationship to student
behavior and achievement, they show it to be distinct from the
integrative motive.
 Ge'ment, Dornyei, and Noels 441

Conceptualizing Foreign Language Classroom Motivation

Together with the factor analytic solutions, the results sup-

port the existence of a tricomponent motivational complex (see
Figure 1) in L2 learning within a foreign language classroom
environment. - The first component, the integrative motive, is
associated with a number of orientations, including the instw-
mental-knowledge orientation. I t is to be hoped that this latter
result, coupled with Gardner's (1985) extensive discussion, will
put an end t o the misleading use of a simplistic integrative-
instrumental dichotomy.
Even though the integrative motive emerged here a s a well-
defined cluster, unlike the results in North American studies
(Gardner, 19831, it showed little relationship t o the evaluation of
the teaching environment. Such results suggest that in this
context the language teacher and the L2 course are not perceived
as linked t o the L2 group and its members. The more obvious
impact of the present situation on the social determinants of L2
acquisition appears t o be the relative independence of social-
psychological processes, which in a multicultural context, have
been more closely associated with one another.
The second major component of L2 motivation is self-confi-
dence. Self-confidence influences L2 proficiency both directly and
indirectly through the students' attitude toward and effort ex-
panded on leariing English. The indirect course of influence was
the one Clkment and Kruidenier (1985)hypothesized and tested.

Motivation

Linguistic Foreign Language

Self-confidence Behavior and Competence

Appraisal of
Classroom Environment

Figure 1. Schematic representation of a tricomponential approach.

442 Language Learning Vol. 44, No. 3

The direct course indicates, in addition, that aspects of contact

with the L2 and its users provide the student with skills readily
applicable in the classroom.
The third component corresponds t o the classroom environ-
ment. Group cohesion is associated with a positive evaluation of
the learning environment, thus forming a broader “perceived
classroom situation” cluster. A s has been seen, this motivational
dimension emerges a s a motivational subsystem independent of
integrative motivation and self-confidence.
Although it is possible t o conceptualize these components as
independent contributors t o proficiency, the present results also
suggest that the classroom context and extracurricular contact
activities may jointly affect language self-confidence-as would be
expected from an ihteractive model of language learning. On the
one hand, good classroom atmosphere promotes student involve-
ment and activity while moderating anxiety and promoting
self-confidence. On the other hand, the student brings into the
classroom a level of self-confidence and anxiety related to extra-
curricular experiences with the language, the quality and quantity
of which would then influence classroom behavior, achievement
and anxiety. Accordingly, being active in class means believing
that one is able t o use English outside the classroom. Even in the
present case, where there is minimal contact with the second
language community, this “real” world would therefore be a n
important source of “bad” experiences with actual consequences,
which challenge the students and generate anxiety (MacIntyre &
Gardner, 1989).
Obviously, given the impossibility of making strong causal
inferences from correlational results: this interactive interpreta-
tion would need to be tested for its causal sequence and its
generality. It is, however, in line with previous results and would
theoretically extend Clement’s ( 1980) model of the language learn-
ing process.
 Clement, Uornyei, and Noels 443

Pedagogical Implications

Our results confirm the relevance of a social psychological

approach t o the understanding of L2 motivation: even in the
Hungarian situation, where contact with L2 speakers was. re-
stricted, socially grounded factors were related to the students'
attitude and effort, classroom behavior and achievement. Fur-
thermore, application of the social psychological constructs of
group dynamics prcved useful in describing the L2 classroom
environment. Thus, o u r results give empirical grounding to a
recent development in L2 methodology, whereby group dynamic
activities are incorporated into the L2 syllabus in order to foster
various aspects of group development and enhance group dohe-
sion, with the aim of creating an environment more conducive to
learning (Hadfield, 1992). Our results suggest, in addition, that
the task of the foreign language teacher and researcher is also to
curb and use influences which extend beyond the'school context.
While recognizing the necessity of looking a t the student through
the eyes of the teacher as we have done here, it seems imperative
that we all look a t the task of learning another language through
the eyes of the student. That, we suggest reveals foreign language
acquisition to be a complex social process.
Revised version accepted 13 May 1994

Notes

'Defined operationally, the self-confidence construct includes two compo-

nents (Clement & Kruidenier, 1985): anxiety as the affective aspect and self-
evaluation ofjgoficiency as the cognitive component. Although the functional
relationship between the two components and behavior would,Seem to be
complex (Noels & Ckment, 1994),,their high intercorrelation justifies treat-
ing the two aspects as feeding into the same construct.
T h e maximum-likelihood procedure was chosen here because of its superior-
ity overothermethods(e.g., principal component,principal axislforproviding
an index of the goodness of fit of the solution to the data. It should be noted
that, in this case, because of the impossibility of correlating error variances
between the items, the goodness-of-fit index should not be expected to reflect
as good a fit as with CFA techniques where this is possible. The Oblimin
rotation procedure was chosen because it does not assume the independence
444 Language Learning Vol. 44, No. 3

offactors. In this case, rotation ofthe factorsfollowing thevarimaxprocedure

did, however, produce virtually identical results. We reported the obliquely
rotated solutionhere because of the rather high correlation between Factors
1and 3. Finally, for both analyses, we computed multiple solutions, extract-
ing different numbers of factors and using dfferent rotation techniques. We
decided upon the solution presented in this paper by considering the eigenval-
ues associated with and the percentage of variance accounted for by each
factor, a n examination of the Scree plot, and evidence of simple structure.
3Asfor the factor analysis of the orientations, the solution generatedhere was
achieved via maximum-likelihood EFA. In this case, however, to be consis-
tent with previous studies, we applied Varimax rotation. The matter of the
relationship betweer, the major constructs studied here is dealt with in a
subsequent section of this paper.
’In fact, before Au’s (1987)article, the question of whether the self-evaluation
component belongs more readily with anxiety or achievement was tested by
Clement and Kruidenier (19851.a~a feature of the measurement model of
their LISREL study. In none of the groups studied was there any indication
that associating self-evaluation measures to indices of proficiency would
provide a better fit to the model.
me correlation between the two anxiety indices was .69 (pc.001) and both
correlated .56 (p.001) with the self-evaluation score.
6Application of multiple-regression or causal-modeling techniques to these
data would not have permitted stronger causal inferences (cf. Cliff, 1983;
Muliak, 1987). These apprdaches would, in this case, entail only mathemati-
cal transformations of the correlation matrices presented here.

References

Au, S. Y. (1987). A critical appraisal of Gardner’s social-psychological theory

of second-language (L2) learning. Language Learning, 38, 75-100.
Bailey, K. M. (1983).Competitiveness and anxiety in adult second language
learning: Looking at and through the diary studies. In H. W. Seliger & M.
H. Long (Eds.), Classroom oriented research in second language acquisi-
tion (pp. 67-102). Rowley, MA: Newbury House.
Bar-Tal, Y., & Bar-Tal, D. (1986). Social psychological analysis of classroom
interaction. In R. S. Feldman (Ed.), The social psychology of education:
Current research and theory (pp. 132-149). Cambridge: Cambridge Uni-
versi ty Press.
Brown, R. (1988). Group processes; Dynamics within and between groups.
Oxford: Blackwell.
C a d n e s , E. G., & McIver, S. P. (1981).Analyzing models with unobserved
variables: Analysis of covariance structures. In G. W. Bohrnstedt & E. F.
Borgatta (Eds.), Social measurement: Current issues (pp. 65-115). Beverly
Hills, CA: Sage.
 Clkment, Dornyei, and Noels

Cattell, R. B. (1966). The scree test for the number of factors. Multivariate
Behavioral Research, 1, 245-276.
Clement, R. (1978). Motivational characteristics of Francophones learning
English. Quebec: Universib2 Laval, International Centre for Research on
Bilingualism.
Clement, R. (1980). Ethnicity, contact and communicative competence in a
second langffage. In H. Giles, W. P. Robinson, & P. M. Smith (Eds.),
Language: Social psychological perspectives: Selected papers h r n the first
International Conference on Social Psychology and Language held a t the
University ofBristo1, England, July 1979(pp. 147-154)*0xford: Pergamon.
Clbment, R. (1984). Aspects socio-psychologiques de la communication inter-
ethnique et del’identite sociale [Social psychological aspectsofinterethnic
communication and social identity]. Recherches sociologiques, 15, 293-
312.
Clement, R. (1986). Second language proficiency and acculturation: An
investigation of the effects of language status and individual characteris-
tics. Journal of Language & Social Psychology, 5,271-290.
Clement, R., Gardner, R. C., &‘Smythe, P. C. (1977). Motivational variables
in secondlanguage acquisition: A study offiancophoneslearning English.
Canadian Journal of Behavioural Science, 9, 123-133.
Clement, R., Gardner, R. C., & Smythe, P.C. (1980). Social and individual
factors in second language acquisition. Canadian Journal ofBehavwura1
Science, 12, 293-302.
Clbment, R., & Kruidenier, B. G. (1983). Onentations in second language
acquisition: I. The effects ofethnicity, milieu, and target language on their
emergence. Language Learning, 33,273-291.
Clement, R., & Kruidenier, B. G. (1985). Aptitude, attitude and motivation in
second language proficiency: A test of Clement’s model. Journal of Lan-
guage & Social Psychology, 4, 21-37.
Cliff, N. (1983). Some cautions concerning the application ofcausal modelling
methods. Multivariate Behavioral Research, 18, 115-126.
Cohen, Y., & Norst, M. J. (1989). Fear, dependence, and loss of self-esteem:
Affective barriers in second language learning among adults. RECL
Journal, 20,61-77.
Crookes, G., & Schmidt, R. W. (1991). Motivation: Reopening the research
agenda. Language Learning, 41,469-512.
Dornyei, Z. (1990a). Conceptualizing motivation in foreign-language learn-
ing. Language Learning, 40,45-78.
Dornyei, 2. (1990b, April). Analysis of motivation components in foreign
language learning. Paper presented at the 9th World Congress of Applied
Linguistics, Thessaloniki, Greece. (ERIC Document Reproduction Service
No. ED 323 810)
446 Language Learning VOl. 44, No. 3

Evans, N. J., & Jarvis, P. A., Dawson, C. (1986). The group attitude scale: A
measure of attraction to groups. Small Group Behavior, 17, 203-216.
Evans, C. R., & E o n , K. L. (1991). Group cohesion and performance: A meta-
analysis. Small Group Research, 22, 175-186.
Fahmy, J. J., & Bilton, L. (1992). The sociocultural dimension of TEFL
education. Journal of Multilingual and Multicultural Development, 13,
269-289.
Forsyth, D. R. (1990). Group dynamics (2nd ed.). Pacific Grove, CA: BrookdCole.
FOSS, K. A., & Reitzel, A. C. (1988). A relational model for managing second
language anxjety. T E S O L Quarterly, 20, 559-562.
Gardner,R. C.(1980).On thevalidityofaffectivevnriablesin secondlanguage
acquisition: Conceptual, contextual, and statistical considerations. Lan-
guage Learning, 30,255-270.
Gardner, R. C. (1983).Learning another language: A true social psychological
experiment. Journal of Language & Social Psychology, 2, 219-239.
Gardner, R. C. (1985). Social psychology and second language learning: The
role of attitudes and motivation. London: Arnold.
Gardner, R. C., & ClBment, R. (1990). Social psychological perspectives on
second language acquisition. In H. Giles & W. P. Robinson (Eds.), Hand-
book of language and social psychology (pp. 495-517). Chichester: John
Wiley & Sons.
Gardner, R. C., Day, J. B., & MacIntyre, P. D. (1992).Integrative motivation,
induced anxiety, and language learning in a controlled environment.
Studies i n Second Language Acquisition, 14, 197-214.
Gardner, R. C., & Lambed, W. E. (1972).Attitudesand motivation in second-
language learning. Rowley, MA: Newbury House.
GardnLr, R. C., & MacIntyre, P. D. (1991).An instrumental motivation in
language study: Who says it isn’t effective? Studies in Second Language
Acquisition, 13, 57-72.
Gardner, R. C., & Smythe, P. C. (1975). Second 1anguageacquisition:A social
psychological approach (ResearchLBulletin No. 332). London, Ontario: The
University of Western Ontario, Department of Psychology.
Giles, H., & Byme, J. L. (1982). An intergroup approach to second language
acquisition. Journal of Multilingual and Multicultural Development, I ,
1740.
Giles, H., & Johnson, P. (1987). Ethnolinguistic identity theory: A social
psychological approach to language maintenance. International Journal
of the Sociology of Language, 68, 69-99.
Gorsuch, R. L. (1983). Factor analysis (2nd ed.). Hillsdale, NJ: Erlbaum.
Greene, C. N. (1989). Cohesion and productivity in work groups. Small Group
Behavior, 20, 70-86.
Hadfield, J. (1992). Classroom dynamics. Oxford: Oxford University Press.
 Cle‘ment, Dornyei, and Noels 447

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.,Horwitz,M. B., &Cope, J.(l986). Foreignlanguageclassroom
anxiety. The Modern Language Journal , 70, 125-132.
Horwitz, E. K., & Young, D. J. (1991). Language anxiety: From theory and
research to classroom implications. Englewood Cliffs, N J Prentice Hall.
Julkunen, K. (1989). Situation- and task-speckc motivation in foreign-
language learning and teaching (University of Joensuu Publications in
Education, No. 6). Joensuu, Finland: University of Joensuu.
Kozponti Statisztikai Hivatal (Central Statistical Omce). (1992).Az 1990. evi
nepszamlalas; 3. kotet; osszefoglalo adatek [The 1990 census: Vol. 3.
Summary data]. Budapest: Author.
Labrie, N., & ClCment, R. (1986).Ethnolinguistic vitality, selfcofi+nce a n d
second language proficiency: An investigation. Journal of Muftilingual
and Multicultural Development, 7, 269-282.
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. (1991a). Investigating language class
anxiety using the focused essay technique. The Modernhnguage Journal,
75,296-304.
MacIntyre, P. D., & Gardner, R. C. (1991b). Methodsandresults in the study
of anxiety and language learning: A review of the literature. Language
Learning, 41,85-117.
Mihaljevik, J. (1990).Researchon motivation for learning Englishasaforeign
language: Aproject in progress. Studia Romanica etAngelicaZagrabiensa,
25, 151-160.
Moi‘se, L. C., ClCment, R., & Noels, K. A. (1990). Aspects motivationnels de
l’apprentissage de l’espagnol a u niveau universitaire [Motivational as-
pects of learning Spanish at the university level]. Canadian Modern
Language Review, 46,689-705.
Mulaik, S. A. (1987). Toward a conceptualisation of causality applicable to
experimentation and causal modelling. Child Development, 58, 18-32.
Noels, K. A., & CICment, R. (1989). Orientations to learning German: The
effects of language heritage on second language acquisition. Canadian
Modern Language Review, 4 5 , 2 4 5 2 5 7 .
Noels, K. A., & Clbment, R. (1994). Second language anxiety: A socio-
communicative model of affect, cognition and behaviour. Unpublished
manuscript .
Oller, J. W., Jr., & Perkins, K. (1978). Intelligence and language proficiency
as sources of variance in self-reported affective variables. Language
Learn ing,~28,85-9 7.
448 Language Learning Vol. 44, No. 3

Phillips, E. M. (1992). The effects of language anxiety on students' oral test

performance and attitudes. The Modern Language Journal, 76,14-26.
Prabhu, N . S. (1992). The dynamics ofthe language lesson. TESOL Quarterly,
26,225-241.
Price, M. L. (1991). The subjective experience of foreign language anxiety:
Interviews with highly anxious students. In E. K. Horwitz & D. J. Young
(Eds.), Language anxiety: From theory and research to classroom implica-
tions (pp. 101-108). Englewood Cliffs, NJ: F'rentice Hall.
Rosenfeld, L. B., & Gilbert, J. R. (1989). The measurement ofcohesion and its
relationship to dimensions of self-dsclosure in classroom settings. Small
Group Behavior, 20,291-301.
Samimy, K. K., & Tabuse M. (1992). Affective variables and al es s commonly
taught language: Astudy in beginning Japanese classes. Languagebarn-
ing, 42,377-398.
Shaw, M . F . (1981). Group dynamics: Thepsychology ofsmallgroup behavior
(3rd ed.). New York: McGraw-Hill.
Skehan, P. (1989). Individual differences in second-language learning. Lon-
don: Arnold.
Skehan, P. (1991).Individual difference in second language learning. Studies
in Second Language Acquisition, 13, 275-298.
Soh, K. (1987). Language use: A missinglink? Journal of Multilingual and
Multicultural Education, 8 , 4 4 3 4 4 9 .
Tabachnick, B. G., & F'idell, L. S. (1989). Using multivariate statistics (2nd
ed.). Philadelphia,' PA: Harper & Row.
Young, D. J. (1986). The relationship between anxiety and foreign language
oral proficiency ratings. Foreign Language Annals, 12,439-448.
Young, D. J. (1991). Creating a low-anxiety classroom environment: What
does language anxiety research suggest? The Modern Language Jo'urnal,
75,426-437.