A Diligent Student’s Distress in English-Medium Instruction Classes: A Longitudinal Study45The study draws on multiple data sources. One of the primary data sources is Satomi’s weekly reflective journals for 16 weeks during the semester, which she sent every week via smartphone （23 written reports in her L1）. Satomi was asked to write about her reflections and emotions freely and send the text immediately on the class day, or at least within a few days, because retrospective reports are more accurate when produced as soon as possible after the class （Heigham & Croker, 2009）. This is equivalent to learner diaries, which “record the experiences of language learning from the learner’s perspective” （Barkhuizen et al., 2014, p. 35）, offering a rich picture of the learner’s social and cognitive dimensions （e.g., Bailey, 1983）. Another primary source was the two recorded interviews conducted on Zoom （total 105 minutes） at the beginning of and after the semester, which elicited more meta reflections. This method gathered narratives, in which she could better articulate some topics retrospectively with time （Bamberg & Demuth, 2016）. These narratives present how participants （re）construct their past events （Duff, 2012）, which enables structural analysis of the narrative approach （Riessman, 2008）. Lastly, notes from a total of 11.5 hours of classroom observation were added to the dataset for triangulation to understand the “contexts of construction” of the narratives （Barkhuizen, 2010, p. 282）. In attending observations, I also read all the class reading materials.For the field notes on classroom observation, a memoing space was provided in the margin to write my reflections as a preliminary analysis. During this process, questions emerged, and these were checked in accordance with Satomi’s weekly journals. All the weekly journals were transferred to spreadsheets for thematic analysis （Braun & Clarke, 2006）. The interview narrative data were transcribed verbatim and coded using Atlas.ti. I conducted thematic analysis （Braun & Clarke, 2006） and structural analysis following Riessman （2008）. Thematic analysis, also known as content analysis, analyzes what has been said in the narratives, whereas structural analysis explores how a speaker strategically conveys meaning. Because a narrative structural organization is a linguistic choice of the speaker, analyzing narrative structure reveals how meaning is expressed with a particular arrangement of structural elements. This analysis supplemented the shortcomings of the thematic analysis. The structural analytical method in this paper followed Riessman’s adoption of Labov’s （1972） structural coding （abstract, orientation, complicating action, evaluation, and resolution） and other coding techniques. For analytical transparency, I left my codes in the data Excerpts. Thus, the multiple data sources and structural analysis aim to overcome the weakness of thematic analysis of personal narratives in seeking “subject reality, life reality, and text reality” （Pavlenko, 2007, p. 163）.My identity at the site of this research was as an outsider-observer. I was not a student or a teacher in the course, but participated in the classroom observation as a graduate student who belonged to The DataThe AnalysisResearcher Positioning
元のページ ../index.html#51