An Exploration of Chinese EFL
Students’ Emotional Intelligence
and Foreign Language Anxiety
KAIQI SHAO
Hangzhou Normal University
School of Foreign Languages
Wenlan Street 1
Hangzhou, ZJ 310036, China
Email: shaokaiqi@live.cn
WEIHUA YU*
Hangzhou Normal University
School of Foreign Languages
Wenlan Street 1
Hangzhou, ZJ 310036, China
Email: weihuayu0806@163.com
ZHONGMIN JI
Hangzhou Normal University
School of Foreign Languages
Wenlan Street 1
Hangzhou, ZJ 310036, China
Email: jzmrain@yahoo.com
This study examined 510 Chinese students’ emotional intelligence (EI) and English classroom learning
anxiety at three universities in Hangzhou, People’s Republic of China. Results obtained from the Trait
Emotional Intelligence Questionnaire–Short Form (TEIQue–SF) and the Foreign Language Classroom
Anxiety Scale (FLCAS) showed the following results: (a) More than half of the Chinese college students
possessed a middle to high level of EI and at least one third experienced language anxiety in English
class; (b) Moderate to relatively strong associations were found among students’ EI, foreign language
anxiety (FLA), English achievement, and self-rated English proficiency; (c) FLA had a significant and
partial mediating effect on EI in predicting students’English achievement; and (d) FLA also significantly
and partially mediated the relationship between EI and self-rated English proficiency. Results and
implications are discussed in the context of the importance of emotional factors affecting the learning of
English.
Key
word
word文档格式规范word作业纸小票打印word模板word简历模板免费word简历
s: emotional intelligence; foreign language anxiety; second language learning; English
proficiency; Chinese EFL students
EMOTIONAL INTELLIGENCE (EI) IS THE
ability to perceive emotions, to access and
generate emotions to assist thought, to under-
stand emotions and emotional knowledge, and
to reflectively regulate emotions to promote
emotional and intellectual growth (Mayer &
Salovey, 1997). As a theoretical construct, it
incorporates all possible feelings and emotional
skills into a unified framework (Goleman, 2001).
Accordingly, it is theorized as allowing an
individual to identify and regulate negative
emotions and also to generate and utilize positive
emotions to facilitate thinking (Ciarrochi &
Mayer, 2007). Finally, it determines a person’s
level of anger, frustration, and motivation
(Goleman, 1995).
With regard to language learning, MacIntyre
(2002) stated that “to some extent language
learning itself is prone to create intense emotion”
(p. 67). Furthermore, to him, emotion “just
might be the fundamental basis of motivation,
one deserving far greater attention in the
*Denotes Corresponding Author
The Modern Language Journal, 97, 4, (2013)
DOI: 10.1111/j.1540-4781.2013.12042.x
0026-7902/13/917–929 $1.50/0
© 2013 The Modern Language Journal
language learning domain” (pp. 45–68). Scovel
(2000) concurs in this assessment of the impor-
tance of emotion, finding that “emotions might
well be the factor that most influences language
learning, and yet is the least understood by
researchers in second language acquisition
(SLA)” (p. 140).
Specifically, negative emotions such as anxiety,
fear, stress, and anger can compromise learners’
optimal learning potential and largely reduce
their language learning capacity. By contrast,
positive emotions such as self-esteem, empathy,
motivation, and enjoyment can put learners in an
optimal state for language learning and greatly
facilitate the language learning process (Arnold&
Brown, 1999; Schumann, 1994; Stevick, 1995).
Second language learners with a higher level of EI
are better able to control impulses, manage stress,
and maintain a positive attitude in the face of
challenges and frustrations during the acquisition
process (Pishghadam, 2009). In essence, EI claims
to predict how learners will react to the demands
of various second language learning and use
contexts, which is crucial for their successful
acquisition of the second language.
Research devoted to themeasurement of EI has
led to two distinct models (Petrides &
Furnham, 2001). The ability model of EI employs
a purely cognitive measure that aims to elicit test
takers’ maximal performance on certain emo-
tional information processing tasks. By contrast,
the trait model of EI is concerned with behavioral
dispositions and self-perceived abilities as mea-
sured through self-reports that strongly reflect
personality variables. Trait EI theory considers EI
as a constellation of emotion-related self-percep-
tions and dispositions. Such a theoretical propos-
al is consistent not only with mainstream theories
of personality but also with the bulk of the
available evidence from multiple studies in
different domains, such as life satisfaction,
rumination, and coping styles (Petrides & Furn-
ham, 2006; Petrides, Pita, & Kokkinaki, 2007).
Extending their theoretical work into empirical
research, Petrides and Furnham (2001, 2003)
developed a research instrument, the Trait
Emotional Intelligence Questionnaire (TEIQue).
It has repeatedly been used in research and
overwhelmingly found to be a highly valid and
reliable instrument (Barlow,Qualter, & Stylianou,
2010; Freudenthaler et al., 2008; Mavroveli
et al., 2008; Mikolajczak et al., 2007; Mikolajczak
& Luminet, 2008; Swami, Begum, & Petrides,
2010).
Based on the extensive research that docu-
ments the importance of EI for learning overall, a
key question pertains to the nature of foreign
language anxiety (FLA). At themost general level
it is characterized as the feeling of tension and
apprehension associated with second language
learning contexts, particularly in relation to
speaking and listening tasks (MacIntyre &
Gardner, 1994). Importantly, like MacIntyre
and Gardner, Horwitz, Horwitz, and Cope
(1986), too, see FLA as situation-specific
and describe it as a distinct complex of self-
perceptions, beliefs, feelings, and behaviors
related to classroom language learning that arises
from the uniqueness of the language learning
process. It differs from the anxiety experienced
in other fields of learning because of the socio-
cultural and linguistic demands that it imposes
on learners. In fact, an entire theoretical
construct, the affective filter hypothesis, was
posited by Krashen (1980) to recognize the
impact of anxiety on language learning. Accord-
ing to Krashen, when second language learners’
anxiety is high, their affective filter is raised and
they become less able to process language input,
fail to take in the available target language
messages, and do not progress in their language
acquisition. Along similar lines,MacIntyre (1995)
proposed that language anxiety causes divided
attention and therefore diminishes levels of
cognitive performance that further impair lin-
guistic performance. In turn, Horwitz (1986)
noted that language learners with high levels of
anxiety tend to be self-conscious, have difficulty
concentrating, and fear making mistakes. As a
result, they will avoid communicative situations,
study for extended times, become forgetful, and
even experience a mental block (Horwitz
et al., 1986; Tobias, 1979). Echoing those
observations, Broidy (2005) argued that students
with higher language anxiety are prone to have
irrational ideas, are less able to control their
impulses, and cope less successfully than other
learners.
Not surprisingly, such performance im-
pairment results in lower levels of achievement.
Thus, a number of researchers have found
negative relationships between language anxiety
and various measures of achievement, though,
in general, this correlation was moderate
(Aida, 1994; Argaman&Abu–Rabia, 2002; Cheng,
Horwitz, & Schallert, 1999; MacIntyre & Gardner,
1991; Onwuegbuzie, Bailey, & Dailey, 1999;
Phillips, 1992). Indeed, Dewaele (2002) and
Sparks and Ganschow (1991) had difficulty
confirming the role of FLA in foreign language
learning. However, Aida’s (1994) study of 94
English-speaking students studying Japanese at
918 The Modern Language Journal 97 (2013)
a U.S. university found that highly anxious
students received significantly lower grades than
students with low anxiety.
Linking language performance, FLA, and EI,
the following nexus thus becomes plausible:
High EI individuals believe that they can regulate
their emotional reactions over time, manage
stress, and be assertive. They are also likely to be
confident in their ability to communicate effec-
tively in a foreign language, a quality that makes
them less likely to experience FLA (Dewaele,
Petrides, & Furnham, 2008), which, using Krash-
en’s terminology, was previously referred to as
having a lowered affective filter. This common-
sensical assumption seems to have been con-
firmed in recent research focused on examining
the connections between EI and FLA. For
example, Dewaele et al. (2008), studied the
effects of EI and a set of sociobiographical
variables on communicative anxiety and FLA
among 464 multilingual adults. They observed
that participants’ EI and FLA scores were
inversely correlated and that diverse sociobio-
graphical variables, such as age of onset of
acquisition, context of acquisition, and frequency
of use, also linked with FLA.
Theoretical and empirical insights such as
these become particularly consequential for
educational systems with considerable numbers
of foreign language learners, such as China, the
country that has the largest number of EFL
students in the world. Most of these learners have
little opportunity to speak English outside the
classroom and often fear using the language to
communicate in public or naturalistic settings,
even when they are high achievers. There is, then,
an urgent need to address their emotional needs
and feelings during foreign language learning,
for the obvious reason that knowledge on the part
of teachers and administrators, even on the part
of the students themselves, of how to reduce
anxiety, might enhance learners’ ability to
become successful language users and language
learners. As Wang (2005) observed, “success in
language learning is largely connected with how
teachers take care of students’ emotional needs.
When students’ emotional needs are met, then
students in turn concentrate on learning the
language” (p. 2).
Situated in the Chinese EFL context, the
present study aimed to examine students’ EI
and FLA at the tertiary level with the goal of
contributing to a more differentiated awareness
and understanding of the relationship among
three main types of learner variables: emotional
intelligence, language anxiety, and English profi-
ciency. The study was guided by the following
research questions:
RESEARCH METHOD
Participants
Participants in the study were 510 non-English
major first year students at three universities in
Hangzhou, People’s Republic of China, 355
females and 155 males. We chose nonmajors in
order to facilitate comparability among the
learners. Students’ ages ranged from 16 to 21,
with a mean age of 18.35 (SD¼ 1.75). All had
passed the college entrance examination and
registered for four-year full-time undergraduate
study. They were required to take an English
course for 3.5 hours per week for two academic
years. At the end of the first academic year, all
students took the College English Test 4 (CET4),
the only national and official college English test
in China.
Procedures
Participants were recruited by homeroom
teachers and informed about the study’s focus
on “their emotional intelligence and foreign
language anxiety.” In September 2011, at the
beginning of the academic year, participants
completed the Chinese translated version of the
Trait Emotional Intelligence Questionnaire–
Short Form (TEIQue–SF) and the Foreign
Language Classroom Anxiety Scale (FLCAS)
during a twenty minute homeroom period.
They were informed that their scores on the
CET 4 would be collected at their respective
universities. In June 2012, after two semesters of
study, participants’ EI and FLA scores were
matched with their English record. Also, students
self-rated their English proficiency, using a 5-
point Likert scale to respond to the question:
“What is your overall English proficiency?”
RQ1. What are the general tendencies of
students’ EI and FLA?
RQ2. What are the associations among students’
EI, FLA, English achievement, and self-rated
English proficiency?
RQ3. Does FLA play a mediating role in the
relationship between EI and English
achievement? If so, what is the extent and
significance of such mediation?
RQ4. Does FLA play a mediating role in the
relationship between EI and self-rated
English proficiency? If so, what is the
extent and significance of that mediation?
Kaiqi Shao et al. 919
Instruments
The study used two well-established assessment
instruments, the Trait Emotional Intelligence Ques-
tionnaire–Short Form (TEIQue–SF) and the Foreign
Language Classroom Anxiety Scale (FLCAS).
The Trait Emotional Intelligence Questionnaire
(TEIQue) is a measurement instrument based
exclusively on trait EI theory. Its measures are
based on an international scientific research
program that aims to integrate various nontradi-
tional intelligences (e.g., social intelligence,
multiple intelligences) into mainstream models
of personality and differential psychology (Pet-
rides & Furnham, 2001, 2003). The TEIQue–SF is
the short form of the TEIQue, comprised of 30
items with two items each from the 15 subscales
(see the Appendix, items 1–30). The test yields
scores on four primary factors: well-being, self-
control, emotionality, and sociability, in addition
to global trait EI. It is coded on a 7-point Likert
scale continuum from “Completely Disagree (1)”
to “Completely Agree (7).”Within a possible score
range from 30 to 210, a total score of more than
150 indicates a well-developed EI; a total score of
120 to 150 suggests a moderately developed EI;
and a total score of less than 120 reveals an
underdeveloped EI. The internal consistency of
the test is usually above .80 and has not dropped
below .70 in any studies. The reliability (Cron-
bach’s alpha) of the test in this study was found to
be as high as .86. The validity of the test has
also been confirmed by several researchers
(Freudenthaler et al., 2008; Mavroveli et al.,
2008; Mikolajczak et al., 2007; Petrides, Pe´rez–
Gonza´lez, & Furnham, 2007).
The Foreign Language Classroom Anxiety Scale
(FLCAS), the second instrument used in the study,
was originally developed by Horwitz and her
colleagues (1986) to meet the needs of the
research community for a consistent way of
measuring foreign language classroom anxiety.
Based on students’ self-report, clinical experi-
ence, a review of related instruments, and the
researchers’ personal experience as language
teachers, it differentiates three main sources of
language anxiety. Communication apprehension is a
type of shyness characterized by fear or anxiety
when communicating with people in the foreign
language. Test anxiety refers to a type of perfor-
mance anxiety that stems from a fear of failure in
examinations. The third component, fear of
negative evaluation, is defined as “apprehension
about others’ evaluation, avoidance of evaluative
situations, and the expectation that others would
evaluate oneself negatively” (Horwitz et al., 1986,
pp. 127–128). Since its first presentation, the
instrument has gained widespread popularity in
researching the role of anxiety in various
language learning contexts and thus became
the instrument of choice for our study.
The FLCAS contains 33 items (see Appendix,
items 31–63) whose responses are recorded along
a 5-point Likert scale, from “strongly disagree (1)”
to “strongly agree (5).” Within the total possible
score range from 33 to 165, the following levels are
generally identified: Scores above 132 signify high
anxiety; scores between 99 and 132 denote a
middle level of anxiety; and scores below 99 imply
little or no anxiety. The test has shown very high
internal consistency (above .90) in several studies
(Aida, 1994; Horwitz, 1986; Liu & Jackson, 2008;
Rodriguez & Abreu, 2003). In this study, the
reliability of the scale as measured by Cronbach’s
alpha was .92. The test–retest reliability of the
questionnaire was .83 inHorwitz (1986) and .80 in
Aida’s (1994) study. Its validity, too, has been
supported or partially supported by research (e.g.,
Aida, 1994; Argaman & Abu–Rabia, 2002, 2002;
Horwitz, 1986; Liu & Jackson, 2008; MacIntyre &
Gardner, 1989; Rodriguez & Abreu, 2003).
Because half of the items (items 2, 4, 5, 7, 8, 10,
12, 13, 14, 16, 18, 22, 25, 26, and 28) on the
TEIQue–SF and eight items (items 2, 5, 8, 11, 14,
18, 22 and 28) on the FLCAS are negatively
worded, responses to these items were reversed
and recoded. As a result, a high score on these two
scales represents high trait EI or high language
anxiety.
RESULTS
General Tendencies of EI and FLA
To establish general tendencies of students’ EI
and FLA, descriptive statistics such as the total
scores, standard deviations, mean,median, mode,
minimum, andmaximumwere calculated.Table 1
presents the descriptive statistics of the above
parameters on the TEIQue–SF and FLCAS.
As displayed in Table 1, the TEIQue–SF had a
mean of 140.78, a median of 142, and mode of
135. With all three parameters falling into a
middle range, one can conclude that more than
half of the participants believed that they were
aware of the emotional states of others as well as
their own and capable of regulating these
emotions according to environmental changes.
For the FLCAS, the mean (92.03) and median
(95) fell below the low-middle threshold score of
99; however the mode (104) was higher than the
average score. Additional frequency analysis
920 The Modern Language Journal 97 (2013)
showed that at least one third of the students
experienced moderate to high levels of anxiety in
English classrooms, a result that is consistent with
the finding in Liu and Jackson’s (2008) study.
Table 1 also shows that students varied greatly
on their EI and FLA scores, as indicated by the
respective standard deviations (20.75; 20.42).
Some students (e.g., with a score of 189) reported
a very high level of emotional competency
whereas others (e.g., with a score of 83) clearly
exhibited a serious deficit in their emotional
command. Likewise, some students (e.g., with a
score of 36) felt extremely confident in classroom
English learning while others (e.g., with a score of
161) were overwhelmed by their anxiety in
English class.
Associations of EI and FLA With Students’ English
Achievement and Self-Rated English Proficiency
To assess the relationships among students’ EI,
FLA, English achievement, and self-rated English
proficiency, a Pearson Product correlation analy-
sis was performed. Table 2 presents correlations
among all these variables.
As shown in Table 2, students’ scores on EI and
FLA (r ¼ �0.681, p < 0.01) were negatively and
significantly correlated with each other. Students
who scored higher on EI tended to score lower on
FLA. This indicates that students who were better
at controlling emotions in different situations
were less likely to develop language anxiety in
English classrooms, a result that supports the
findings in Chuan–Ta’s (2003) study, which
found that students’ FLA scores correlated
significantly with each of the EI skills, EI total
scores, and some EI problem areas. Table 2 also
shows that students’ English achievement was
positively and significantly related to their EI
(r ¼ 0.327, p < 0.01) but negatively and signifi-
cantly related to their FLA (r ¼ �0.320, p < 0.01).
As illustrated by the regression lines in Figure 1,
which visualize this information, students’ En-
glish score scattered around the upward slant of
EI and the downward slant of FLA.
This implies that students who possessed a
higher level of EI or experienced a lower level of
FLA weremore likely to perform better in English
examinations. Both the positive effect of EI on
English success (Pishghadam, 2009; Shao, Yu,
& Ji, 2012) and the negativ