Foreign language

Data Analysis Procedures

In this study, thematic analysis (Boyatzis, 1998) was adapted to analyze the qualitative data coming from the two distinct sources, the online discussions and semi-structured interviews.

The researcher first prepared hard copies of data gathered from reflection papers, the discussion sessions, and the interviews. Later, reflection papers and the discussion sessions were categorized separately for each individual and every single comment or question the participants directed to and received from each other were also listed in individuals‟ folders.For the interviews, the researcher first completed the verbatim transcription by writing every single detail that showed up in speech such as repetitions, incomplete utterances and affective expressions (i.e.,laughs).
Also, during the transcription stage, the researcher took notes about the preliminary themes that participants talked about. Then, printed copies of reflection papers,

discussion sessions, and transcribed interviews were compiled. In order to provide anonymity, the researcher gave codes to each participant (ex. Participant 1 [P # 1]) according to the alphabetical order of their names and then reorganized separate folders for each participant.

Mills, Durepos and Miebe (2010) state that “thematic analysis is a tactic for reducing and managing large volumes of data without losing the context, for getting close to or immersing oneself in the data, for organizing and summarizing, and for focusing the interpretation” (p.2).Therefore, upon compiling so many pages of raw data, the researcher decided to apply thematic analysis as the first stage. Conducting the analysis and categorizing the information require creating thematic codes which can be generated via three different ways: theory-driven, prior-research driven and data-driven which can also be called as inductive (Boyatzis, 1998). For the purposes of this study, inductive thematic analysis is more appropriate to complete the bottom- up feature of the analysis stage. As Ortaçtepe (2012) suggested “the inductive way results in codes that are driven by raw data.” (p. 46). Since constructing themes and creating codes require a neat work, Boyatzis‟ (1998)thematic analysiswas utilized to sequencethe analysis stage of the study. The steps of thematic analysis were defined as preparation, pattern recognition, creating a code, and comparing themes (as adapted by Ortactepe, 2012).
As the first step of thematic analysis, the researcher read the reflection papers and transcribed interviews several times and created the codes in these papers. Later on the researcher defined themes that naturally occurred over codes in order to categorize codes under broader concepts. Later the themes were color coded and organized under reflection types: reflection-on-action, reflection-in-action and reflection-for-action. It is important to state that Wilson‟s (2008)‟s temporality of

reflection was utilized to distinguish reflection types. Thus, past situations and experiences were categorized under reflection-on-action, participants‟ awareness on present conditions was listed under reflection-in-action, and future plans and intentions of teachers were categorized under reflection-for-action.

Figure 1. Data analysis procedure

For instance, participants‟ papers, comments and transcribed interviews were read and motivation, personal characteristics, autonomy and language learning experiences of participants were detected as codes. Looking at the codes from a broader perspective, the researcher realized that the issues in these codes revealed participants‟ learner identity;that way, the theme naturally occurred. Upon creating all codes and themes, researcher categorized the preliminary defined themes under reflection types reflection-on-action, reflection-in-action and reflection-for-action according to Wilson (2008)‟s temporality on reflection types.