Data Collection and Procedures

國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

37

the students with varying English proficiency levels. Except for students in Sea class, four students participated in the student interviews, with two belonged to the group of advanced level students.

3.3 Data Collection and Procedures

To ensure “credibility (p.301)” and “confirmability (p. 318)” of the study (Lincoln & Guba, 1985), the researcher employed various data collection methods, including teacher narratives, teacher semi-structured interviews, informal student semi-structured interviews, classroom observations, and copies of relevant course content materials to collect data and triangulate findings.

Teacher Narratives

Teachers construct their knowledge about teaching mainly from their

experiences (Johnson & Golombek, 2002). It is believed that the most appropriate and authentic way of exploring such experiences is through narrative as people think and make sense of their experiences via story telling (Bell, 2002; Clandinin & Connelly, 2000; Connelly & Clandinin, 2006; Johnson & Golombek, 2002). When teachers narrate their experiences, they invest the experiences with their interpretations

(Golombek & Johnson, 2004), rendering narrative the most suitable tool to understand teacher learning and teaching from teachers’ points of view. Over the years, more and more researchers have been using teacher narratives to explore teachers’ knowledge and teaching practices within the sociocultural contexts where they teach and learn to teach and how they make sense of their development process (Clandinin & Connelly, 2000; Golombek & Johnson, 2004; Johnson & Golombek, 2002).

The “legitimate knowledge” (Johnson & Golombek, 2002, p.4) that teachers produce out of their narratives of lived learning and teaching experiences (Golombek

& Johnson, 2004; Johnson & Golombek, 2002) is the research interests in the present

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

38

study. Thus, in this study, the teacher participant, Joy, was asked to recount her prior English learning, teacher training and teaching experiences. From the personal and contextualized narratives, the researcher hoped to learn, from Joy’s perspectives, about her progressive internal development process through which she continually constructed and modified her cognitions and practices concerning teaching large multilevel classes within the sociocultural contexts where she taught and learnt to teach. In other words, the researcher hoped to understand Joy’s internal cognitive development process of what and how she came to know about teaching in multilevel classes. The interview protocol for teacher narrative is included in Appendix A; it is designed based on Borg’s (2006) framework of language teacher cognition and Webster and Mertova’s (2007) and Carspecken’s (1996) guidelines.

Semi-structured Interviews

The researcher also conducted semi-structured interviews with Joy and ten of her students. Two types of semi-structured interviews were undertaken with Joy: one was to confirm, clarity and broaden the issues emerged from the teacher narrative; the other was to further investigate what had been noted through classroom observations.

As noted earlier, at the later stage of data collection, it was recognized that there was a need to hear voice from the students so to add additional perspectives on and ensure the trustworthiness of the data provided by Joy. Informal interviews were hence conducted with ten of Joy’s students that were with varying English proficiency levels.

One semi-structured teacher interview lasting about three hours was conducted to further probe into the significant events recounted in the teacher narrative and more contextual information about the teachers’ current teaching context. With the data elicited, the researcher developed a clearer understanding of Joy’s cognitive

development in teaching large multilevel English classes. The exact questions in the

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

39

semi-structured interview were subject to the narrative data.

In addition, since how a teacher utilizes her cognitions in classrooms is largely interpretive (Johnson & Golombek, 2002), to further investigate how Joy externalized her internal cognitions concerning teaching multilevel classes into actual classroom practices and to examine how the formation of material practices were facilitated and/or inhibited within the teaching context, the researcher undertook a

post-observation semi-structured interview every three or four days, depending on the teacher’s availability. In total, four post-observation semi-structured interviews were undertaken, each lasting one and a half to two and a half hours. The interview protocol for post-observation semi-structured teacher interviews is included in Appendix B; it is also constructed based on Carspecken’s (1996) guidelines.

At the later stage of data collection, it was found that data from the students should also be collected to shed additional light on the data provided by Joy. Informal individual interviews were hence undertaken with ten of Joy’s students. Each

interview took about ten to fifteen minutes to probe into how students of varying English abilities perceived their own English abilities, Joy’s teaching in large multilevel English classes and Joy’s interactions with students of diverse English abilities.

Near the end of data collection for the study, two follow-up interviews were conducted with Joy to clarify some issues that had not been answered quite clearly in the previous interviews.

Classroom Observations

As evident in the literature, teachers use their cognitions in teaching in a way that is constantly negotiated within the sociocultural contexts where they teach (Borg, 2003, 2006; Johnson & Golombek, 2002). Therefore, to investigate the interactive relationship between teacher cognitions and classroom practices within the teaching

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

40

context, the researcher conducted classroom observations of Joy’s teaching and interactions with the students. To avoid interruption (Carspecken, 1996), the researcher carried out classroom observations passively sitting in the back of the classroom. To guard against unnoticed biases towards the observation, the researcher followed Carspecken’s (1996) suggestions of “the use of a flexible observation schedule (p.48).” Despite so, three focuses could be identified: the procedures, content and manners of the teacher’s teaching, the responses of the students to their teacher’s teaching, and the interaction between the teacher and her students. The priorities of observation were the teacher, the students that interact with her, and the rest students. To ensure that the field notes produced are trustworthy in a certain degree, the researcher adopted Carspecken’s (1996) advice in constructing thick description. Besides, to assist the researcher in constructing field notes, the researcher also video-taped the observations of Joy’s teaching and interactions with her students after gaining her admission.

The researcher conducted two phases of classroom observation. The first phase of classroom observation was conducted before the first mid-term and covered the nine to ten periods that the teacher spent to finish teaching Lesson 2. In teaching Sun Class and Sky Class, it took ten class periods for Joy to finish teaching Lesson 2, but only nine for Sea Class. The second phase of class observation was conducted before the second mid-term and covered the seven to eight class periods that Joy used to finish teaching Lesson 5. She spent eight class periods to finish teaching Lesson 5 to Sky Class, while used seven class periods to finish teaching the same lesson to the other two classes. After the observation of several class periods, the researcher would conduct a post-observation interview with the teacher to delve into the questions emerged from the classroom observations.

Besides the data collection methods mentioned above, relevant course content

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

41

materials, such as lesson plans, supplementary teaching and learning materials, assignment sheets, and test and exam papers, were also collected to triangulate and elaborate the findings.

With the permission from the participants, the data collected through narratives, semi-structured interviews, and classroom observations, were audio and/or

video-taped and transcribed for further analysis (Carspecken, 1996; Nunan, 1992).

The researcher’s memos and research journals, where she recorded her own attitudes, assumptions, and any changes that occur were also put to analyze to maintain a critical perspective on the data (Hood, 2009).