Title Page

ABSTRACT

Contents

CHAPTER I. INTRODUCTION 14

1.1. Background of the Study 14

1.2. Purposes and Significance of the Study 17

1.3. Definitions of Key Terms 20

1.4. Organisation of the Study 22

CHAPTER II. LITERATURE REVIEW 23

2.1. Introduction to testing in second language education 23

2.1.1. Problems in testing and negative impacts 23

2.1.2. Why do language teachers need sound knowledge of language testing? 27

2.1.3. Backwash and washback in testing 28

2.1.4. Validity in L2 testing 34

2.1.5. Reliability in L2 testing 38

2.1.6. Recent studies in oral testing 41

2.2. Oral testing in L2 education 45

2.2.1. Introduction to oral testing 45

2.2.2. Types of speaking in oral testing 47

2.2.3. Micro and Macro skills of speaking 48

2.2.4. Designing assessment tasks: imitative speaking and intensive speaking 49

2.2.5. Extensive speaking and types of assessments 52

2.2.6. Recent studies in oral testing 54

2.3. Task based assessment and testing in L2 education 60

2.3.1. Introduction to tasks in second language education 60

2.3.2. What is a task in terms of language education? 60

2.3.3. Important issues to consider when developing tasks in testing 61

2.3.4. Characteristics of assessments and tests in L2 education 62

2.3.5. What is task-based language assessment? 62

2.3.6. The challenges TBLA faces in L2 education 63

2.3.7. Potential problems in methods utilized in TBLA 64

2.3.8. Strengths in using tasks in testing 66

2.3.9. What are the required qualities to make tasks communicative? 69

2.3.10. Negative effects of tasks in testing 71

2.3.11. Recent studies in task-based tasks 72

2.4. Group Dynamics and possible effects on students' performance 77

2.4.1. Definition what is a group? 77

2.4.2. Group dynamics in education 78

2.4.3. Group cohesion and why is it important? 80

2.4.4. Group cohesiveness and group productivity 82

2.4.5. Being conscious of cohesiveness in the classroom 82

2.4.6. Helping foster group cohesion in your class 83

2.4.7. Important features within the group 84

2.4.8. Negative effects of cohesiveness 85

2.4.9. Nonverbal communication 87

2.4.10. Culture and applied nonverbal communication 87

2.4.11. Cultural Influences on Encoding Nonverbal Behaviours 89

2.4.12. Cultural Influences on Decoding Nonverbal Behaviours 89

2.4.13. Recent studies in Group Dynamics 91

CHAPTER III. METHODS 94

3.1. Research Questions 94

3.2. Participants 94

3.3. Course Overview 96

3.4. Instruments 97

3.4.1. Speaking tests (one to one evaluation interview style) 97

3.4.2. Multiple choice pen and paper test (Mid-term test) 98

3.4.3. Task-based testing (group oral speaking test) 98

3.4.4. Non-task-based test (group oral speaking test) 99

3.4.5. Group work dynamics 99

3.4.6. Questionnaires 100

3.5. Scoring 100

3.5.1. Rubrics 100

3.5.2. Raters 101

3.5.3. Rating 101

3.6. Procedures 104

3.7. Data Analysis 106

CHAPTER IV. RESULTS 107

4.1. Task-based testing and non-task-based testing results 107

4.2. Group Work Dynamics analysis 112

4.3. Students' attitudes and feelings towards the testing 118

4.3.1. General Perspectives and Attitudes 118

4.3.2. Chi-Square Analysis 119

4.3.3. Nerves and Attitude 119

4.3.4. Task Content and Attitude 119

4.3.5. Time Limits and Attitude 121

4.3.6. Group Preferences and Attitude 122

4.3.7. Creative Task, English Skills and Attitude 123

4.3.8. Prior Knowledge / Evaluation of topics and Attitude 124

4.3.9. Summary of chi-square analysis 125

CHAPTER V. DISCUSSION & CONCLUSION 125

5.1. Task-based testing compared to non-task-based testing 126

5.2. Task-based testing compared to non-task-based testing GWD 129

5.3. GWD negative communication 132

5.4. Questionnaire analysis 132

5.5. Conclusion 136

5.5.1. Summary of the Results 136

5.5.2. Pedagogical Implications 140

5.5.3. Limitations of the Study 143

5.5.4. Suggestions for Future Research 145

REFERENCES 149

APPENDICES 155

APPENDIX A. Task-based test used to test the student's oral abilities in the study 155

APPENDIX B. Non-task-based test used to test the student's oral abilities in the study 160

APPENDIX C. List of positive group work dynamic (GWD) characteristics 161

APPENDIX D. List of negative group work dynamic (GWD) characteristics 162

APPENDIX E. Rating scale for speaking abilities on Evaluation Speaking Test 164

APPENDIX F. Rating scale for speaking abilities on Task-based test and Non-tasked based test 166

APPENDIX G. Questionnaire questions given to the students upon completion of the study to gather attitudes and feelings towards the two different types of testing 169

Kangwon National University PhD research permission confirmation 173

Table 1. The extended Levelt model illustrating task and performance 74

Table 2. Correlation analysis of group work dynamic (GWD) and the task motivation variables. 93

Table 3. Inter-Rater Reliability for the Task-Based Test and Non-Task-Based Tests. 102

Table 4. Inter-Rater Reliability for Group Work Dynamics (Verbal/Positive) 103

Table 5. Inter-Rater Reliability for Group Work Dynamics Negative Non-Task-based test (Verbal and Nonverbal) 104

Table 6. Descriptive statistics: Speaking evaluation and mid-term tests 108

Table 7. Results of descriptive statistics: task-based tests evaluation 109

Table 8. Results of descriptive statistics: non-task-based-test evaluation 109

Table 9. Results of class one paired t-test for TBT and NTBT 110

Table 10. Results of class two paired t-test for TBT and NTBT 110

Table 11. Results of class three paired t-test for TBT and NTBT 111

Table 12. Results of descriptive statistics TBT group work dynamics characteristics (verbal and non-verbal) 112

Table 13. Results of descriptive statistics NTBT group work dynamics characteristics (verbal and non-verbal) 113

Table 14. Class one: TBT and NTBT paired t-test for GWD positive verbal and non-verbal characteristics 113

Table 15. Class two: TBT and NTBT paired t-test for GWD positive verbal and non-verbal characteristics 114

Table 16. Class three: TBT and NTBT paired t-test for GWD positive verbal and non-verbal characteristics 115

Table 17. Results of descriptive statistics for task-based test negative GWD (Verbal and Non-verbal) 115

Table 18. Results of descriptive statistics for non-task-based test negative GWD (Verbal and Non-verbal) 116

Table 19. Class one: paired t-test for TBT and NTBT GWD negative non-verbal 117

Table 20. Class two: paired t-test for TBT and NTBT GWD negative non-verbal 117

Table 21. Class three: paired t-test for TBT and NTBT GWD negative non-verbal 118

Table 22. Questionnaire chi-square test: nerves and potential for video group selection 119

Table 23. Questionnaire chi-square test: grade expectation and nerves 119

Table 24. Questionnaire chi-square test: English ability and thinking about sentences 120

Table 25. Questionnaire chi-square test: grade expectation and making sentences 120

Table 26. Questionnaire chi-square test: Which test was best and a good way to speak English 120

Table 27. Questionnaire chi-square test: Which test was best and brainstonning/mind mapping 120

Table 28. Questionnaire chi-square test: Ability to show speaking skills and brainstorming 120

Table 29. Questionnaire chi-square test: Which test would you do again and brainstorming 120

Table 30. Questionnaire chi-square test: Fairness of Test and Brainstorming/Mind mapping 121

Table 31. Questionnaire chi-square test: Grade Expectation and Brainstorming 121

Table 32. Questionnaire chi-square test: Fairness of Test and Brainstorming/Mind mapping 121

Table 33. Questionnaire chi-square test: Which test was best and time limits 121

Table 34. Questionnaire chi-square test: Fairness of Tests and Time limits 122

Table 35. Questionnaire chi-square test: Difficulty of Test and Time Limits 122

Table 36. Questionnaire chi-square test: Grade Expectation & time limits 122

Table 37. Questionnaire chi-square test: Which Test was best and Group Learning 122

Table 38. Questionnaire chi-square test: Which test would want to do again and group type preference 123

Table 39. Questionnaire chi-square test: Difficulty of Tests and Nervous about Group types 123

Table 40. Questionnaire chi-square test: Grade Expectation and Tests with Friends 123

Table 41. Questionnaire chi-square test: Ability to show speaking skills and the task-based test was a good way to show... 123

Table 42. Questionnaire chi-square test: Fairness of Tests and the task-based test was a good way to speak English 124

Table 43. Questionnaire chi-square test: Grade Expectation and task-based test was a good way of speaking English 124

Table 44. Questionnaire chi-square test: Which test is better and conversation topic knowledge 124

Table 45. Questionnaire chi-square test: Which test is better and ease of topics 124

Table 46. Questionnaire chi-square test: Grade Expectation and Knowledge of Topics 124

Table 47. Questionnaire chi-square test: Grade Expectation and ease of topics 125

The study here was motivated by ten years of testing second language students. The goal was to develop and design a task-based test to elicit language in a slightly different manner to the more traditional topic centred oral test. Having administered ten years of oral testing, I was keen to develop a method that was inspired during teaching communication English classes at my university. I had previously administered a classroom task-based group project that was conducted during class time over three weeks ending in a group presentation. I was so impressed by the student's effort using their English during class time and the final presentation that I came up with the idea to condense the group project into a group oral test design, illustrated in this study. The aim was to get the students to interact the same way as they did in the class group project in the new group oral test. Hence the group oral task-based test was designed. I also wanted to test to see how this test design differed in students' performances compared to the traditional topic-based oral test. I had suspected that the task-based test may allow greater interaction, complexity of language, and creativity than what I had experienced when administering topic-based oral tests.

Task-based testing in EFL has been investigated previously regarding oral testing, however, previous studies were not concerned with extensive L2 production and interactive communication whilst completing the tasks. This study attempts to investigate the L2 production in groups of EFL students whilst completing two types of testing a task-based test (TBT) designed oral test and a non-task-based test (NTBT, conversation topic-test). The students' groups during testing were also evaluated for their group work dynamics to further understand the importance of group work dynamics in performance and to understand any differences created by the testing design. The study was conducted during the course of one academic semester in a Korean university. The investigation was centred on answering three key questions: a) whether the different oral speaking test designs of task-based-testing or non-task-based testing (conversation topic test) create any differences in L2 production in EFL learners; b) whether group work dynamics is affected by the two different oral tests used in the study; and c) what are the students' feelings and attitudes towards the two main types of oral tests completed during the university course. The research questions were examined against three separate basic communication English classes at a Korean university where the students' levels were approximately low-intermediate to upper intermediate ability of speaking English.

Moreover, the study used a total of 71 undergraduate Korean university students taking their basic communication English course, all students signed a consent form to be part of the study. The three classes included two English department classes one consisted of 16 students', the second consisted of 26 students', and the third class contained a mixture of degree students' 29 in total. At the beginning of the course, all the students' received an evaluation oral speaking test, the course then followed a standard EFL textbook. The students were asked to sit in groups of 3-5 people during class time and advised they could freely change. During week 8 a multiple-choice test was administered based on the earlier classes. In week 9 an in-class practice test of the task-based oral test was conducted, at this time the students were advised after the practice test, they must stay in the same groups for the remainder of the testing. Week 10 of the course they were instructed to video themselves completing the task-based test outside of class and email the recorded test within a one-week deadline. In week 11 of the course, the practice test of the non-task-based test (conversation topic test) was conducted, and then in week 12, it was set as a test to complete and email within the same one-week deadline. The two oral tests conducted in weeks 10 and 12 were reviewed and analysed according to a set criterion (see appendices). In the final class of the course, the students were given a questionnaire to complete concerning their feelings and attitudes towards the testing they went through. The results obtained underwent a quantitative data analyses, SPSS 19.0 descriptive statistics and paired t-test analysis were applied to the TBT, NTBT and group work dynamics of the students' groups in the TBT and NTBT. The questionnaires were analysed by applying the chi-square test of independence to determine the significant difference between independent categorical variables and other categorical variables in the study.

Subsequently, the major findings from the study discovered that there was a significance in the task-based test results in two areas of students' performances in complexity, and creativity. In contrast, the non-task-based tests showed a significance in the students' performance in fluency. The interesting discoveries reveal that during group oral testing there is a potential to somewhat control what we expect to see from the student's performances. Additionally, the group work dynamics analysis showed that the task-based test illustrated a significance in the students' behaviour in leadership qualities. Whereas the non-task-based test displayed a significance in loud laughter among the groups taking the test. Interestingly, the group work dynamics showed an interesting glimpse into the life of the group during testing conditions, an area still with limited understanding. Also, the study found from the questionnaire analysis that 67% of the students indicated that they preferred the task-based test. Indeed, there was a general level of preference for this test with 97% considering it a good way to speak English. The results found in this study further underline the need for further research in oral testing and group dynamics in the field of EFL.

In addition, the implications of the results discovered in this study show some interesting intricacies in group oral testing. If we take the results discovered here and continue further investigations in group oral testing, we will be in a better position to provide greater insights into students' English communication abilities to potential employers or education institutions. With this greater understanding of testing and student evaluations, we can further advance the professionalism of testing and teaching in EFL. However, the study has limitations with specifically the group dynamics analysis was a simplified version of a ground-breaking method used in language education. Therefore, the method of data capture and analysis still requires further refining to better understand the importance that group dynamics plays in language education. Also, the group oral testing despite the positive findings still requires further investigations into different types of tasks to elicit language production with comparisons to this and other studies on tasks as a means of encouraging language communication.