Methods of data analysis

4.5.1 Thematic analysis of qualitative data

I collected qualitative data from a wide variety of sources. I used the following data for analysis:

(1) Interviews with teachers;

(2) Students’ answers to questionnaires (except students in other schools); (3) Audio and video recording of students’ discussions;

(4) Students’ documents.

I also took notes during the observations and used the questionnaires completed by students, as supplementary data to support the analysis of the interviews.

In order to analyze the above four types of qualitative data, I used the thematic analysis (Braun & Clarke, 2006). Before describing the process of the analysis, I will first define what type of approach to the thematic analyses I used, and for what reason. According to Braun and Clarke (ibid.: 79), “thematic analysis is a method for identifying, analyzing and reporting patterns (themes) within data.” Thematic analysis differs from content analysis and grounded theory, which are the major methods of qualitative analysis. While content analysis tends to calculate frequencies of categories and be followed by statistical analysis (Cohen et al., 2007: 473-491), thematic analysis tends not to count the frequencies (Braun & Clarke, 2006: 98). Also, unlike grounded theory, which involves theoretical sampling (Cohen et al., 2007: 491-500), thematic analysis does not involve such a continuous sampling process. In this study, the purpose of analyzing the qualitative data was to seek to understand the overall picture of each set of data by interpreting the relations of themes in each data set; thus the approach, which involves discussing the data based on the frequencies of emerged categories, did not fit into the purpose. As regards grounded theory, it is related to the whole process of research, including collecting data, and also its aim is to build a theory. Thus, I employed thematic analysis, which “is not

action research project.

According to Braun and Clarke (2006), thematic analysis has two main approaches. One is an inductive approach while the other is a theoretical approach. In an inductive approach, coding is not driven by the researcher’s theoretical or analytical interest. Data are thus not fit into a pre-existing coding frame. On the other hand, a theoretical approach is driven by the researcher’s theoretical or analytical interest. In this sense, codes for data can be developed in relation to the researcher’s research question. In this study, I chose an inductive approach, because it was an exploratory study by nature. As I discussed in Chapter 3, this study initially aimed to develop an appropriate pedagogy for treating culture in textbooks by exploring teachers’ views of culture teaching (RQ1) and critical reading (RQ2), and students’ responses to critical reading (RQ3). This study also involved a number of action and reflection to identify the area of focus, and thus it was impossible to set categories in advance. In this regard, this study employed an inductive approach for analysis.

Braun and Clarke (2006) also point out that, as well as deciding which approach to use, identifying themes is also an important decision. Themes can be identified either at a semantic level or at a latent level. The analysis at the former level involves looking for themes at a surface or explicit level, and theorizing significant patterns and their broader meanings. Thus, what a participant has actually said or written is only analyzed, and meanings beyond his or her account are not considered at this level. In contrast, an analysis at a latent level examines underlying meanings embedded in the data. Since the features that give data particular meanings are identified, themes in the data are interpretatively developed. Deciding the level of analysis, i.e., either a semantic level, or a latent level, highly depends on the paradigm of a research.

Thematic analysis can be conducted in a realist/essentialist or constructionist paradigm. Braun and Clarke (2006) use the terms, “realist/essentialist” and

“constructionist,” but here I use the terms, “positivist” and “constructivist,” which were also used in Section 4.1.2. Within a positivist paradigm, experiences, meanings and experiences are reported as objective realities; on the other hand, within a constructivist paradigm, events, realities, meanings and experiences are considered as the meanings constructed in social contexts. As discussed in Section 4.1.2, the research paradigms which this study adopts are constructivist and critical paradigms. This means that the process for categorizing is interpretative. I thus analyzed data at a latent level.

4.5.2 The process of analysis

To analyze the qualitative data, I followed the six phases as suggested by Braun and Clarke (2006) (Table 4.5.2).

Table 4.5.2: Six phases of a thematic-analysis (Braun & Clarke, 2006:87)

Phase Description of the process

1. Familiarizing yourself with your data:

Transcribing data (if necessary), reading and re- reading the data, noting down initial ideas.

2. Generating initial codes: Coding interesting features of the data in a systematic fashion across the entire data set, collating data relevant to each code.

3. Searching for themes: Collating codes into potential themes, gathering all data relevant to each potential theme.

4. Reviewing themes: Checking if the themes work in relation to the coded extracts (Level 1) and the entire data set (Level 2), generating a thematic ‘map’ of the analysis.

5. Defining and naming themes: Ongoing analysis to refine the specifics of each theme, and the overall story the analysis tells, generating clear definitions and names for each theme.

6. Producing the report: The final opportunity for analysis. Selection of vivid, compelling extract examples, final analysis of selected extracts, relating back of the analysis to the research question and literature, producing a scholarly report of the analysis.

(1) Transcriptions (for spoken data only)

recordings, the analysis of those data began with transcription. Transcribing is a meaningful act rather than a mechanical process. Transcribing involves familiarizing yourself with your data (Riessman, 1993 cited in Braun & Clarke, 2006.). Although data can be transcribed by someone else, transcribing the data yourself makes you to pay greater attention to the data; it also helps you to retrieve inaudible parts from memory (Rubin & Rubin, 2005: 205). Furthermore, if someone transcribes spoken data for you, you need to spend more time for familiarizing yourself with the data and checking the transcripts for accuracy (Braun & Clarke, 2006: 88). Transcribing also involves interpreting the data as part of analysis (Bird, 2005: 227 cited in Braun & Clarke, 2006). Thus, it is important to keep in mind that the analysis of spoken data begins during the interview process, in which interviewers interpret and clarify certain statements or make connections with earlier statements; moreover, the spoken data does not exist as a pure source of information (Richards, 2003: 80-81).

Before starting to transcribe data, the level of precision of transcripts needs to be decided. How precise transcriptions should be depends on the type of analysis. Conversation, discourse, and narrative analysis would require you to use or construct systems of detailed transcription. On the other hand, thematic analysis does not require you to describe data as detailed as those analyses; nevertheless it should retain all the verbal utterances which are true to the original data at a minimum (Braun & Clarke, 2006: 88). Richards (2003: 81) shows the seven basic features of transcription necessary for the qualitative analysis of interview data: pauses, overlap, emphasis, fillers intonation, problematic features and non-verbal features. I included these features in my transcriptions because I thought that they might be helpful to later reflect on, though thematic analysis focuses on the content of the data rather than the interaction between the interviewer and interviewee. Apart from the scheme of transcription, Richards (ibid.: 82) suggests formatting “a reliable line numbering system, easy transfer of the main text

to the final paper, report of thesis, and space for comments or notes.” Considering the convenience of reference and coding, I followed this suggestion. The last things to note is naming those who were involved in the process. I used “IR” for the interviewer as recommended by Richards (ibid.: 82-83), and pseudonyms for the interviewees and those who were mentioned in the interviews for their privacy. The following text box (Box 4.5.2) is the transcription convention.

Box 4.5.2: Transcription convention

The language used for transcription is Japanese, because transcripts translated in English do not represent exactly the same meaning as what was spoken in Japanese. Extract 4.5.2 is an example of transcripts based on my translations.

Extract 4.5.2: Example of transcript

70 IR: わかりました．えーでは[２つ目の