Methods

This study focused on five questions:

(1)What is the overall foreign language emotional state of these international students in their beginning stage of language learning?

(2)What is the relationship between Chinese language enjoyment and anxiety?

(3)What are the impacts of gender，the number of foreign languages mastered and countries of origin of these students on their Chinese language enjoyment and anxiety?

(4)What is the predictability of Chinese language enjoyment and anxiety for students’learning outcome?

(5)What are the main sources of Chinese language enjoyment and anxiety of international students in their beginning stage of language learning?

The Chinese language emotion scale adopted in this study begins with a demographic section，collecting participants’information including age，gender，nationality，and academic major.The main body of the scale included three sections，with the first and second sections based on the foreign language enjoyment scale and the foreign language anxiety scale used by Dewaele and MacIntyre(2014)in their study.The first two sections contain 29 multiplechoice questions which asked about factual information with 21 questions about Chinese language enjoyment and eight questions about Chinese language anxiety.These questions were based on the five-point Likert Scale，and informants were asked to choose the statement which best describes a situation;each statement was assigned a point score between 1 and 5，with 1 for“strongly disagree”scoring 1 point，2 for“disagree”scoring 2 points，3 for“undecided”scoring 3 points，4 for“agree”scoring 4 points，and 5 for“strongly agree”scoring 5 points.The first 21 questions tested students’Chinese language enjoyment，involving four aspects of enjoyment(creativeness，proudness，interest and pleasure)，as well as the classroom atmosphere created by the teacher and fellow students.The next eight questions tested students’Chinese language anxiety，which，as Dewaele and MacIntyre(2014)did，were based on the eight questions in the“Foreign Language Classroom Anxiety Scale”developed by Horwitz et al.(1986).These eight questions tested physiological signs，stress and lack of confidence related to anxiety.Dewaele and MacIntyre(2014)believe that the shortened“Foreign Language Classroom Anxiety Scale”is as reliable and valid as the original scale.In this study，the internal consistency of the“Foreign Language Classroom Anxiety Scale”was 0.73，indicating that the Tai scale was highly reliable.The internal consistency of the 21 questions about foreign language enjoyment was 0.81，and that of the eight questions about foreign language anxiety was 0.68.This points to the fact that the questions about foreign language enjoyment were highly reliable and reliability of the scale was overall acceptable.Some questions were given reverse scores in the calculation of scores，with higher scores representing higher levels enjoyment/anxiety experienced by the informants.The scale used in this study had been adapted，for instance，the term“foreign language”was changed to“Chinese language”so the questions became more direct and relevant for this study.The third section of the scale included two open-ended questions，which asked informants to describe the most enjoyable and most anxious incidents in the classroom.

The study distributed 156 copies of questionnaire，and all of them were returned effective.Before the study commenced，students had been told of the project and a consent form was distributed among them.Students who agreed to participate voluntarily in the study signed the consent form before completing the questionnaire.The first four of the five questions mentioned before received a score on the basis of the answers to the factual questions about foreign language enjoyment and anxiety in the questionnaire，and an analysis using SPSS produced some tendentious conclusions.The fifth question was an analysis of the open ended questions in the last section of the questionnaire，in an attempt to find out factors that promoted enjoyment experienced by informants in learning Chinese.The data form was analysed，with two sets of abnormal value data patterns(beyond 3 mean and 3 standard deviation)excluded.Normal analysis result indicates that the Chinese language enjoyment data are not normally distributed，whereas the Chinese language anxiety data are normally distributed.