As is common in qualitative research, data analysis was an ongoing process throughout the whole process of research. I was aware that a wide scope of my research needed to be narrowed down. I made attempts, throughout the data collection and transcription stages, to refine my research objectives and guiding questions. To accomplish this goal, I re-read my researcher journals, made notes of interesting perspectives after each interview and kept reading relevant studies to my research in order to enhance my understanding and interpretation of data. I believe this was a critical step in data analysis even though I did not yet engage in data analysis in an “official” way. In other words, I did analyze data in an unofficial manner while collecting data. Because of an emerging focus I sometimes had to modify interview questions or shift focus in classroom observation so that I could incorporate the modified focus in subsequent interviews and observation.
I used Microsoft Excel Workbook to organize codes and also to import the relevant parts of observation notes, interview quotes and information in written documents into under each theme column. The use of Workbook helped to keep track of codes and to import all quotes and relevant information under each quote when necessary. Interviews and class observation were major sources of data in the study. After completing interview transcription and observation notes, I began to search for categories and themes that would answer each of the two research
questions. Most codes (themes) emerged while analyzing interview and observation data. I employed the constant-comparative method, one of the most common data analysis methods, in dealing with a large amount of qualitative data (Dillon, 2012; Lincoln & Guba, 1985). I started with open coding. In this first coding process, numerous concepts and categories were identified. In the next step, all these concepts and categories initially identified in each data source were constantly compared to identify distinct concepts and categories across data (axial coding).
It should be noted that although I was interested in describing focal participants’
perspectives and experiences related to L2 writing in their own terms, my particular focus was on identifying themes that would be the most useful to L2 writing teachers, administrators and policymakers. Therefore, the identified themes include not only the ones my participants
believed to be significant in their literate and teaching lives (emic perspectives), but also the ones I thought to be of importance to stakeholders in L2 writing (etic perspectives) (Denzin & Lincoln, 2003; Leki, 2007).
To answer the first research question, conceptualization of L2 writing by an administrator and teachers in two contextually different programs, two large categories (i.e., pedagogical approaches and contextual factors) were first created because the question aimed to identify pedagogical conceptions and contextual factors that led to these conceptions. Under each of the large categories, tried to identify emerging themes or sub-categories. Under each theme, sub- themes were also identified. Each of the participants’ interviews and observation notes were coded according to this hierarchical structure of large categories, themes and sub-themes. The following is a selection of themes and sub-themes under each of the two large categories.
I. Pedagogical Approaches
Basic Beliefs about writing
Writerly voice Academic genre Professional genre Assignments Types Rationale Feedback practice Challenges Diverse needs Material conditions
Institutional support for the program Employment contract
…
II. Contextual Factors
Teacher training background Expertise in teaching L2 writing
Theory of pedagogy
Material selection (e.g., readings, paper samples) Teaching experience
Experience of teaching L2
Experience of teaching L2 writing Understanding of student needs
Current needs
Personal, academic, professional Future needs
Personal, academic, professional Knowledge of institution
Writing demands across the curriculum Policy related to language teaching …
These themes and sub-themes were obtained through an inductive, reiterative data analysis which involved multiple readings of data, constant and evolving interpretation, and understanding relationship between the themes (Denzin & Lincoln, 2003).
To address the second research question on students’ perceptions of need for L2 writing in two contextually different programs, a similar process of data analysis described above was taken. I came up with the two large categories (i.e., student needs and contextual factors). To identify themes that are relevant to student needs and contributing contextual factors, I mainly drew on student interview data and school policy documents. The following is a selection of themes and sub-themes related to research question 2.
I. Student Needs for L2 Writing
Beginning of the semester Current needs Academic Professional Personal Writing-to-learn language Future needs Academic Professional End of the semester
…
II. Contextual factors
Regular school setting After-school program
Cram school for tests
Communication oriented program Study abroad
Writing Experience (L1/L2)
First language writing experience School-sponsored
Self-sponsored
Second language writing experience School-sponsored
Self-sponsored …
Multiple sources of data in the study enabled me to draw on data triangulation, i.e., the use of different sources of data to examine a phenomenon (Denzin, 1978), particular at the stage of data interpretation. Under each theme I juxtaposed different data sources to interpret
participants’ perspectives. For instance, class observation notes and policy documents helped me to gain a more in-depth understanding of what a particular teacher shared in her interview. I believe multiple data sources contributed to an increase in the validity of the study (Davis, 1995).