ite a slight decrease in the assessment indices of the Validation dataset, these model-fit estimates confirm the correspondence of the data to the CFA model. All this, in turn, verifies the construct validity of the final version of the inventory for its intended purpose. (pp. 221-222)
Table 3.2 presents the details.
Absolute and Incremental Fit Indices for CFA model in ELTRI (Akbari, Behzadpoor, & Dadvand, 2010)
Absolute Fit Indices
Incremental Fit Indices
The questionnaire can be found in Appendix B.
The scoring system was simply adding up all the likart scales and the highest score indicates the participant with the highest self reflection score.
Language schools in Iran, compared to schools and universities, have been more successful in teaching English. A number of reasons were stated for the comparative success of the language schools. For one thing, the instruction duration in language schools is longer. What is more, they have more organized pre-service and in-service teacher-training programs and quality control systems. Additionally, course books and teaching methods in language schools are more up to date. Finally, the students attending language schools are more motivated compared to the students learning English during formal education (Talebinezhad & Sadegi Benis, 2005). For these reasons, private language schools were chosen as the setting of this study.
The study was conducted in a private language school called Safir Language Academy on its Tehran branches. The school was established in 1999, and at the time of conducting the research, it had 20 branches in Tehran and 34 branches in other cities of Iran. The school employed 900 (male and female) English teachers, and over 30,000 students were attending it. Out of the 20 branches in Tehran, one was the center for teaching French and two for the preparation for IELTS examination, and the remaining 17 branches exclusively offered English courses. While IELTS/French branch was coeducational, six branches were boys-only and 11 branches were girls-only centers. Based on the directives issued by the Ministry of Education in Iran, only male teachers can teach in boys-only branches, while in girls-only branches both married male teachers over 35 years of age and female teachers can teach.
Instruction in Safir is term-based, each term lasting for two months. Students are offered a choice of attending a regular course (six hours a week), an intensive course (12 hours a week), or a very intensive course (30 hours a week) (Safir, 2013).
The classrooms in Safir are all equipped with DVD-players, LCD sets, and audio systems. Most of the classrooms have built-in cameras and microphones, which facilitates supervisory classroom observations. Supervisors can observe live classes through the monitor in the management room, or they can observe the classes later, as all video/audio data of all classrooms are automatically saved to a computer hard disc.
Each classroom is large enough to seat 17 students comfortably. The maximum number of students in every classroom in Safir is 17 for the elementary to pre-intermediate levels, and 12 for intermediate and higher levels. Seats in all classrooms are arranged in a U-shape, so that while everyone can face the teacher, interaction among students is facilitated and the teacher can easily move around the classroom (Safir, 2013)
In Safir, students are admitted in two age groups: juniors (9 to 14 years old) and seniors (15 years old and above). While there is no maximum age limit for studying in Safir, about 85 percent of the students are 15 to 40 years old. Female students with about 80 percent of all student population are in majority (Safir, 2013). The teacher recruitment system in Safir is so that only applicants who are fluent in spoken and written English are employed. Prior to employment, they go through a 30-hour teacher-training course.
As the case is for the students, the majority of Safir English teachers are female (Safir, 2013).
In order to investigate the null hypotheses of the study and to investigate any significant relationship or difference between teacher’s reflection and teacher-efficacy the following steps were carried out.
Initially out of 900 EFL adult teachers between 20 to 35 years of experience, the participants were sampled into experienced and novice groups based on their hours of teaching experience. In this research, teachers who taught senior classes were selected. Out of 900 participants, only 719 participants answered the questionnaires. As the result, nine papers were discarded because they did not finish either or both the self-efficacy and or teachers’ reflection questionnaires. Another three cases were dropped from the main study because their values were higher than the standard and ticked the highest point. Thus the main study included 707 cases.
383 experienced female teachers varying from intermediate to advanced level teaching experience were selected. 324 novice female teachers who taught at elementary levels were selected. English Language Teacher’s reflection Inventory (ELTRI) and Teachers’ Sense of Efficacy Scale (TSES) which measured the dependent variables of the study (teacher overall reflectivity and teacher overall self-efficacy), were distributed to measure their reflection and self-efficacy levels. Due to the large number of participants and lack of practicality of explaining the instructions individually to each participant, 21 branch managers of the branches were invited to a meeting and the procedures , instructions, and the aim of the project was described to them in details and the probable upcoming issues were discussed . Next the researcher distributed the 2 questionnaires to them and asked them to go over the questions. As the 21 branch managers were all among the experienced teachers, naturally they were the ones who were asked to answer the questions as well. The next step, inside the branches a meeting was held to describe the instructions and the questionnaires were distributed and the participants were requested to take the inventories home, answer the questions and give them back to the branch managers. The filled questionnaires were gathered in a week from all 21 branches and the data was inserted into excel sheets one by one in details. Finally the related analyses were done to investigate the possibility of a significant relationship between teacher-reflection and teacher-self-efficacy and their components. The differences between the two variables were analyzed to guarantee the effectiveness of the research.
As the researcher aimed to test a relationship between teacher-reflection and teacher self-efficacy the design of the study was a descriptive correlational ex post facto research study.
The variables of the study were as follows:
Predicted (dependent) Variables: Teacher-Reflection and Teacher-Efficacy
Predictor (independent) Variables: Being Novice and being Experienced EFL Teachers
Control variables: gender and age
3.6 Statistical Analysis
After collecting the data the correlation and regression statistical analyses were utilized to analyze the data in this study.
Analysis of Data
This chapter presents the detailed statistical analyses conducted throughout the research and tests the thirteen hypotheses of the study based on the obtained data. To begin with, Univariate and Multivariate Normality of the data were measured. Then comes the data analysis related to testing the hypotheses with different tables and figures, followed by, predictability of EFL Teachers’ Reflection based on experience which was measured through regression. At last, the reliability of both questionnaires ‘Teacher Sense of Efficacy Scale (TSES)’ and ‘English Language Teaching Reflection Inventory (ELTRI)’ are evaluated.
4.2 Testing Assumptions
The research questions posed in this study were analyzed through the parametric statistical analyses of Pearson product correlation and independent samples t-test. That is why the researcher had to confirm that the dependent variables (self-efficacy and teaching reflection) were measured on an interval scale, the data enjoyed normal distribution, and variances were homogeneous in conducting the independent samples t-test.
As displayed in Table 4.2 the values of skewness and kurtosis were within the ranges of +/- 1.96 (Filed 2009).
Testing Normality Assumption; Self-Efficacy and Teaching Reflection
It should be noted that the assumptions of linearity and homoscedasticity will be discussed when reporting the Pearson product correlations results.
4.3 Analysis of Outliers
Out of the 719 subjects, nine papers were discarded because they did not finish either or both the self-efficacy and or teachers’ reflection questionnaires. Another three cases were dropped from the main study because their values of Mahalanobis Distances were higher than 13.81 (critical value of chi-square at .001 level for two dependent variables). Thus the main study included 707 cases.
4.4 Testing Assumptions
The research questions posed in this study were analyzed through the Pearson product correlation and linear regression analyses which are based on three main assumptions, i.e. normality,