Foreign Language Anxiety. A Case Study of Chinese University Students Learning English as a Foreign Language

Scientific Study, 2016

21 Pages



1. Introduction

2. Literature Review
2.1 Foreign Language Classroom Anxiety Scale
2.2 Factor analysis of FLCAS

3. Method
3.1 Participants
3.2 Instruments
3.3 Data collection and analysis

4. Results
4.1 The reliability and normality of FLCAS
4.2 Principal component analysis of the FLCAS
4.3 Confirmatory factor analysis of FLCAS

5. Discussion



Author Biography

Appendix I

Appendix II

Abstract: The subjects of the present study were 344 college students from Mainland China. They responded to a 5-point likert scale questionnaire adapted from Horwitz’s Foreign Language Classroom Anxiety Scale (FLCAS). Their responses to the questionnaire were submitted to exploratory factor analysis and confirmatory factor analysis (with the help of statistical package of SPSS and AMOS) for identifying anxiety dimensions. The results show that there existed four dimensions of anxiety among the subjects and the two dimensions of the FLCAS were closely examined to find the extent to which that the four-factor model adequately fit the data. The fit indices suggested that the instrument measuring the dimensions of anxiety specific to college students in China was construct-wise valid and reliable for future use.

Key words: Principal component analysis, Confirmatory factor analysis, FLCAS, Chinese Mainland Context.

1. Introduction

Foreign Language Anxiety (FLA), one of the most important affective variables on foreign language learning, has been the focus of many researchers. For example, Horwitz, suggested a general model of Foreign Language Classroom Anxiety, in foreign language classrooms where we can observe formal instruction of language pedagogy and stress among students trying to communicate in English (Horwitz, 2000, 2001; Matsuda & Gobel 2004). Following the lead by Horwitz, a large number of researches investigated the latent constructs of FLCAS in different contexts (Aida, 1994; Pascual F. et al., 2001; Matsuda,2004; Mak,2011; Park, 2014; Liu, 2015). While the latent constructs of FLCAS with exploratory component analysis have been investigated in different cultural contexts, the investigation of the constructs in China Mainland with both exploratory and confirmatory factor analysis has been lacking. The purposes of this study is to identify the constructs of the anxiety dimensions in Chinese Mailand context and to compare the results with ones form other cultural contexts.

2. Literature Review

2.1 Foreign Language Classroom Anxiety Scale

Anxiety, the subjective feeling of tension, apprehension, nervousness, and worry associated with an arousal of the automatic nervous system (Spielberger, 1972), is a topic that is closely related to human psychology. However, more and more researches concerning foreign language learning anxiety, consistently pointed out that anxiety is indeed an important variable to consider in speech production (Thomas Scovel, 1978; Tsai Yu Chen, 2000; Horwitz, Horwitz, & Cope, 1986). More specifically in classroom foreign language learning and teaching, it can be defined as “a distinct complex of self-perceptions, beliefs, feelings, and behaviors related to classroom language learning arising from the uniqueness of the language learning process” (Horwitz, Horwitz, & Cope 1991: p.31). For the last two decades, research about the negative effects of FLCA has flourished (Aida, 1994; Horwitz, 2000;Herron, 2006; Woodrow, 2006; Tallon, 2009; Liu, 2010), among which, the invention of anxiety scale by Horwitz, Horwitz, & Cope’s (1983) had a far-reaching influence on other studies in this area. In particular some researchers focused their study on exploring the effects of foreign language anxiety on second language acquisition; assessing the degree of FLCAS in classroom and investigating the validity of the scale (Zhang, 2000; Matsuda & Gobel, 2004; Elkhafaifi, 2005; Park, 2014; Mei, 2015;Woodrow 2016) .

2.2 Factor analysis of FLCAS

The exploratory factor analysis is an interdependence technique whose primary purpose is to define the underlying structure among variables and is commonly used in data reduction to identify a small number of factors that explain most of the variance observed in a much bigger number of variables (Joseph F. Hair Jr. 2014). Since Aidas (1994) starting factor analysis, some researchers have been keen on identifying the underlying components of FLCAS with the application of exploratory factor analysis.

When Horwitz, Horwitz and Cope developed the FLCAS(1986), they thought foreign language anxiety was composed of three distinct performance anxiety entities: Communication Apprehension, Test Anxiety and Fear of Negative Evaluation. Horwitz’s analysis of anxiety entities of FLCAS yieldded a 0.93 Cronbach’ alpha coefficient and a high reliability (r=.83, p<0.1). Aida(1994) examined anxiety dimemsions among 96 university students to see whether the structure identified in her research coincides with the three-factor structure reported in Horwitz and Cope (1986). Aida used the principal component analysis with varimax rotation to extract four significant latent factor . The four factors extracted from her approach were Speech Anxiety, Fear of Failing the Class, Comfortableness with the Foreign Language and Negative Attitudes, among which Speech Anxiety and Fear of Failing the Class appeared as the most important components, accounting for 37.9% and 6.3% of the total variance respectively, which partially support Hortwiz’s construct.

Pascual F. et al. (2001) replicated Aida’s study of anxiety dimensions with 198 post-beginner students of English at an Italian official language school. This study had a cronbach’s alpha 0.89 and a high rest-retest reliability (r=.9041 p<.000). In the research, Principal component analysis with varimax rotation produced four significant latent factors. However, the four factors found in their research was somewhat different from the four factors extracted by Aida, in that some of their items loaded on different factors identified by Aida.

Matsuda (2004) investigated the anxiety dimensions using FLCAS in Japan. The principal component analysis with varimax rotation was conducted to the data from 252 Japanese university students majoring in English, yielding seven latent factors with eigenvalues greater than one. However, the last five factors were discarded because only a few items loaded significantly on them. The first factor accounted for 31.1% of total variance, labeled as General English Classroom Performance Anxiety; while the second factor accounted for 6.1% of total variance, labeled as Low self-Confidence in Speaking English.

Mak also ran factor analysis with the application of FLCAS in the context of Hong Kong (2011). For his 313 subjects in Hong Kong, he adapted the five-point scale to four-point scale to avoid the middle point and extracted five latent factors, namely, Speech Anxiety with Native Speakers, Negative Attitudes towards the English Class, Negative self-evaluation, Fear of Failing the Class and Consequences of Personal Failure.

Park (2014) conducted a research on factor analysis with the instrument of FLCAS inthe Korean context. Two latent factors was extracted with the help of Maximum likelihood exploratory factor analysis with direct oblimin rotation from 217 Korean subjects who take English conversation courses required by the university in Korea.The first factor was labeled as Communication Apprehension and Understanding and the second factor was labeled as Communication Apprehension and Confidence. His confirmatory factor analysis indicated a good fit between the proposed model and the implied model

As stated above, most of the studies adopted exploratory factor analysis, which can be used to extract latent factors that share common variance among different indicator items. However, only a few research studies use confirmatory factor analysis to examine whether their data fitted a priori hypothetical models (Cao, 2011). Researchers also attempted to match each item with hypothetical one-factor, two-factor, three-factor, four-factor models using various fit indexes such as chi-square, the root mean square error of approximation (RMSEA), incremental fit index(IFI), and comparative fit index(CFI), and found that four-factor model of the FLCAS fit the data better than others (Park, 2012).

Thompson (2004) pointed out that understanding the factors derived from the FLCAS items through factor analysis was crucial because it could provide evidence for the construct validity of a scale. The factor could be used for subsequent analyses to further investigate the potential effects of anxiety on L2 acquisition. Thompson (2004) further stated that a conclusion can be drawn that different latent components of foreign language anxiety can be extracted from different cultural backgrounds and Horwitz also pointed out the necessity of considering the specific learner populations and learning contexts where foreign language anxiety is being examined since the components of the FLA is likely to vary in different learner populations, especially with respect to cultural and proficiency difference. It is particularly important to consider that FLA has different triggers and manifestations in different culture(Horwitz 2016). The implication from these studies on anxiety dimensions indicated that the dimensions may vary depending upon learning situations and cultural contexts.

3. Method

3.1 Participants

The participants of the present study consisted of 343 first-year students (N=343; Male=133; Female=210) from two universities in China Mainland, their age ranging from 17 to 21, all of whom have studied English for at least 6 years. They were from different majors, including clinics, law, computer, statistics, psychology, administration, medicine, biology etc. All the subjects had already taken a comprehensive English course required by university in China Mainland.

3.2 Instruments

FLCAS consists of 33 Likert-scale items rated on a 5-point scale, with five alternatives ranging from 1: strongly disagree to 5: strongly agree (see Appendix 1). The Chinese version of the FLCAS (see Appendix 2), which was translated with the help of three English translation professors, was used to minimize subjects’ errors in comprehending the original version. In this translation, the terms “foreign language” or “language” used in the original version were replaced with “English” to help Chinese students to understand the scale better.

3.3 Data collection and analysis

FLCAS questionnaires were handed out in classes and an explanation was given on the nature of this study. Subjects are specifically asked to respond to the items carefully because high scores in some items mean high anxiety while high scores in other items mean low anxiety.

For data analysis, the normality of the data was first checked using univariate kurtosis and skewness and then, the reliabililty was determined using Cronbach’s alpha. Subsequently, principal component analysis with varimax rotation was performed to identify latent components with the application of FLCAS. In addition confirmatory factor analysis was performed to examine whether the components identified in the principal component analysis fit the data adequately. The SPSS and AMOS statistical package were utilized to conduct exploratory and confirmatory factor analysis respectively.

4. Results

4.1 The reliability and normality of FLCAS

Internal consistency was computed for the Chinese version of FLCAS. The overall Cronbach’s Alpha coefficient for all the 33 items of FLCAS was 0.794. The values of kurtosis and skewnesss of the 33-item FLCAS were within the range of -2 to +2, indicating the items did not violate the normality assumption.

The Kaiser-Mayer-Olkin measure of sampling adequacy was 0.899, while the approx. Chi-Square value of Bartlett’s Test of Sphericity was 367.830 (df=528, p<0.001) (see Table 1 for KMO and Bartlett’s Test), both indicating that teh principal component factor analysis was appropriate for the data set in this study.

Table 1: KMO and Bartlett's Test

illustration not visible in this excerpt

4.2 Principal component analysis of the FLCAS

Principal component analysis with varimax rotation produced 9 latent factors with eigenvalues greater than 1, The variance accounted for 59.59% of total variance, however, retaining all the 9 factors would create a model too complex in comparison to the number of factors identified in the original study. Therefore, a principal component analysis with fixed number of factors of 4 was conducted (Park ,2012,found that four-factor model of the FLCAS fit the data better than others ) to extract four significant factors, accounting for 41.90% of total variance. Items 15, 16 and 30 are deleted because they did not load significantly on any factor and had low communalities (see Table 2 for Rotated Component Matrix and communalities) .


Excerpt out of 21 pages


Foreign Language Anxiety. A Case Study of Chinese University Students Learning English as a Foreign Language
Catalog Number
ISBN (eBook)
ISBN (Book)
File size
1104 KB
foreign, language, anxiety, case, study, chinese, university, students, learning, english
Quote paper
Yin Xiaoteng (Author), 2016, Foreign Language Anxiety. A Case Study of Chinese University Students Learning English as a Foreign Language, Munich, GRIN Verlag,


  • No comments yet.
Read the ebook
Title: Foreign Language Anxiety. A Case Study of Chinese University Students Learning English as a Foreign Language

Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free