Classroom observation was applied to confirm the reflective writing data and the interview data. According to Marshall and Rossman (1989, p. 79), an observation was “the systematic description of events, behaviours and artifacts in the social setting chosen for study”. It encouraged the researcher to be close to the participants’ contexts, exposed to “large amount of general data” (Borg 2006) and focused on the issues under investigation (Gass and Mackey 2007) so that she could interpret and described them as a holistic picture. Generally, observation was categorised into four types: complete participant, participant as observer, nonparticipant/observer as participant and complete observer (Creswell 2013).
This study used non-participant observation to examine the extent to which the participants employed critical thinking pedagogies in their teaching and the degree to which what they mentioned in their reflective writing and interviews appeared in actual classroom settings. The method was selected due to five justifications. First, it permitted her to collect relevant data as well as seemingly taken-for-granted data constructive to answer the third research question. Second, as a non-participant observer, the researcher had her own private place to keep track of the objects or aspects of observation. Therefore, she could listen attentively, record and reflect on what happened in the classes. Third, as an outsider observing at a distance, an authentic classroom atmosphere could be maintained to some extent. Fourth, as the researcher did not interfere with the participants’ teaching, the power dynamic between the researcher and them was consequently balanced; they still had authority in the classroom (Marriam 1998). Fifth, with the researcher’s minimal interferences and with the teachers’ usual classroom power, students would feel comfortable residing in their regular conditions.
Although prolonged observation was encouraged for ethnographic research (Bernard 1994; Creswell 2013), “There is no ideal amount of time to spend observing” (Merriam 1998, p. 98). This study used a one – time observation for each participant because of the researcher’s time constraint and the participants’ availability. However, the observation method in this study was triangulated to confirm the data from other two methods. Sufficient data, hence credibility, could be expected. Using 12 teachers as the research participants, this study was assumed to increase a variety of perspectives on the topics under investigation. The 5 sites selected were also anticipated to expand the degree of transferability for the readers interested in replicating this study or implementing the study’s findings in their similar contexts.
In observation classes, an audio-recorder was chosen instead of a video recorder because in both pilot studies the former was proved effective in capturing the data with less intrusion in class. Moreover, “if the recording device is unobtrusive, they are less likely to be distracted by it” (Copland and Creese 2015, p. 47). Aside from using an audio-recorder, field notes were written not only for the sake of data analysis, but also for the researcher to “return to field notes to conjure up the experience again, reflect on it, and seek a deeper understanding of the actions and practices that occurred during that time and place” (Copland and Creese 2015, p. 43).
The researcher took field notes using an observation protocol. This protocol included observation date, time and place (Schensul et al. 1999), a sketch of the classroom and descriptive session (see Appendix). Although Creswell (2013) suggested that the protocol should contain both descriptive and reflective sessions, Copland and Creese (2015) asserted that “field notes cannot avoid being evaluative: a language of description does not allow us to be neutral. Whatever we write down positions us in relation to what we observe in one way or other” (p.42). Moreover, as evidenced in the second pilot study where an observation protocol used contained both sessions: descriptive and reflective, it rather caused unnecessary workload to differentiate between the two when writing down what to observe. So, the researcher decided to include only the descriptive section in the protocol used in the main study. Nevertheless, she listed topics to be observed as an observation guide attached to the protocol (see Appendix) so as to remind herself of what needed to be included when recording. The guide was established as a result of the second pilot study as she realised that without it and with unexpected distractions in each class, some important observational issues were missed. When taking notes on the protocol, she used both English and Thai so that the process could run fluently without any language barrier.