3.3 Data Collection
3.3.1 Motivational Factors Questionnaire
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
offices. Communicative opportunities often occurred when the students encountered the teachers in the school. Baker and MacIntyre (2000) found that immersion programs offer more opportunities for students to use the L2; therefore, the opportunities enhance the students’ language abilities. Although the school in the present study was not an immersion school, students in this school could still encounter foreign teachers, and communicate with them easily. It would be possible to say that this bilingual school offered a semi-immersion environment for the students within. Finally, private bilingual schools had become quite popular in Taiwan. Understanding this group of students was significant for English educators, and the findings could be useful for learners who were in similar contexts.
3.3 Data Collection
In this study, data were collected through several ways, including Motivational Factors Questionnaires (MFQ), class observations, student journals, a semi-structured interview with the participant’s ESL teacher, and two semi-structured interviews with the participant. The various data collection instruments used in this study were revised and adapted from Yashima et al.’s and Macintyre et al.’s studies (Yashima et al., 2018;
Macintyre et al., 2011).
3.3.1 Motivational Factors Questionnaire
The first questionnaire that the researcher used was the MFQ. The questions in the MFQ were categorized into five sections that correspond to five sub-factors of the two enduring factors in Layer V of the Heuristic Model. Table 3.1 shows the sub-factors and factors that the researcher aimed to explore through the MFQ in the present study.
‧
Table 3.1: Layer, Name of the Factors, Name of the Sub-factors, Number of Questions, and Total Number of Questions on the MFQ
Layer Name of Factors in the Heuristic
Model
Name of Sub-factors in the Heuristic Model
No. of Questions
V Intergroup attitudes Integrativeness 7
Fear of Assimilation 7
Motivation to learn the L2 7 V Intergroup climate Perceptual and Affective
Correlates
6 Structural Characteristics of the
Community
8
Total Number of Questions 35
From Table 3.1, it shows that the researcher used the MFQ to investigate two enduring factors, intergroup attitudes, and intergroup climate through investigating the five sub-factors, integrativeness, fear of assimilation, motivation to learn the L2, perceptual and affective correlates, and structural characteristics of his community.
In the MFQ, seven items from Yashima’s factor analysis of Japanese learners’
orientations (Yashima, 2002) were modified and used. Six items from MacIntyre, Baker, Clement and Conrod’s (2001) social support index were also modified and used. Twenty-two items from Ryan’s Motivational Factors Questionnaire (Ryan, S., 2009) were also modified and used. The participant rated all items on a six-point scale. The reliability and validity of the questionnaire are explained below.
3.3.1.1 Reliability
The thirty-five questions on the MFQ were categorized into five sections measuring the participant’s integrativeness, fear of assimilation, motivation to learn the L2, perceptual and affective correlates, and structural characteristics of his community. To assessed the reliability of the MFQ, the researcher chose the forms of internal consistency and test-retest to assess the reliability of the MFQ. In terms of internal consistency, the
‧
thirty-five questions on the MFQ were categorized into five sections. Internal consistency reliability is measured by calculating Cronbach’s alpha (Litwin, M. S., 1995). According to Litwin (1995), Cronbach’s alpha reflects how well the items complement each other in the measurement of different aspects of the same variable. According to Fink (1995), when the size of the correlation is zero to zero point two five, the items have little or no relationship. When the size is zero point two six to zero point five, the items have fair degree of relationship. When the size is zero point five one to zero point seven five, the items have moderate to good relationship. And when the size is over zero point seven five, the items have very good to excellent relationship. The data show that the internal
consistency of the items in the sections of fear of assimilation, and structural
characteristics of the community are 0.692, and 0.63, indicating that the items in these two sections have moderate to good relationships. And the internal consistency of the items in the sections of integrativeness, motivation to learn the L2, and perceptual and affective correlates are 0.905, 0.941 and 0.791, indicating that the items in these three sections have very good to excellent relationships. Table 3.2 shows the internal consistency of each section.
Table 3.2: Internal Consistency of the MFQ
Layer Name of Factors Name of Sub-factors Cronbach’s alpha V Intergroup
attitudes
Integrativeness 0.905
Fear of Assimilation 0.692 Motivation to learn the L2 0.941 V Intergroup
climate
Perceptual and Affective Correlates 0.791 Structural Characteristics of the
Community
0.630
In terms of the test-rest reliability, the researcher also applied the test-retest method to assess the reliability of the five sections of the MFQ. Test-retest reliability is measured by having the same set of respondents complete a survey at two different points in time to see how stable the responses are (Litwin, 1995). Correlation coefficients, or r
‧
values, are then calculated. According to Litwin (1995), r values are considered good if they equal or exceed zero point seven. All the r values of the MFQ in the present study exceed zero point seven, indicating that the test-retest reliability of the MFQ is good.
Table 3.3 shows the test-retest reliability of each section.
Table 3.3: Test-Retest Reliability of the MFQ
Layer Name of Factors Name of Sub-factors Correlation V Intergroup
attitudes
Integrativeness 0.801
Fear of Assimilation 0.873 Motivation to learn the L2 0.837 V Intergroup
climate
Perceptual and Affective Correlates 0.851 Structural Characteristics of the
Community
0.972
3.3.1.2 Validity
The researcher applied face validity and content validity methods to assess the validity of the MFQ. In terms of face validity, the researcher asked two untrained
individuals to review the items in the five sections of the MFQ. According Litwin (1995), face validity is based on the review by untrained judges. One of the judges had a master degree, and had worked as a high school English teacher for more than fifteen years. The other judge had a bachelor degree, and had worked in the field of commerce for more than ten years. The two untrained judges thought the thirty-five items in the six sections of the MFQ were all suitable for measuring the participant’s integrativeness, fear of assimilation, motivation to learn the L2, perceptual and affective correlates, and structural
characteristics of his community.
In terms of the content validity of the MFQ, the researcher asked one professional who worked in the bilingual school to review the items in the five sections of the MFQ.
According Litwin (1995), content validity is a subjective measure of how appropriate the items seem to reviewers who have some knowledge of the subject matter. The
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
professional had a master degree from the Department of Applied Foreign Languages of National Taiwan University of Science and Technology. He provided some useful suggestions to the MFQ. According to him, the questions were few, but also suitable for measuring the enduring factors that the researcher aimed to explore.
The researcher administered the MFQ twice during the study. The first time that the researcher administered the MFQ was on May 31st, 2019, and the second time was on August 28th, 2019. The reason for administering the MFQ twice was because the
researcher wanted the participant’s answers to be certain.