Page 140 - Getting of the fence

Chapter 6
coded data based on thematic or conceptual similarity” (Saldaña, 2013, p. 209). The analyses of this second cycle resulted in a final list of six themes. As a final step, the themes were mapped in order to be able to answer the research question. During both cycles, memo writing was used as an analytical technique to support the coding procedure. Because the collected data was in Dutch, the analysis was also conducted in Dutch.
Considering the subjectivity and contextual ramifications in qualitative research, we took various measures to maximise validity and reliability from a qualitative perspective (Miles, Huberman, & Saldaña, 2014; Yin, 2016). First of all, each of the participants received the interview manuscript for a member check (Lincoln & Guba, 1985). Second, the examples we provide in the results section were only translated when we came to a first full draft of this study in order to ensure staying as close to the initial utterances as possible. To increase the validity of the analyses, the initial coding was verified by an external researcher after the first coding cycle. Discrepancies between the analyses were discussed until full agreement was reached. This process of triangulation was repeated after the second coding cycle.
6.4 Results
6.4.1 Results research question 1: Time spent on the four approaches
Table 6.4 shows the average percentage of coded lesson time spent on each of the four approaches before and after the intervention. In both year 1 and year 2, most of the lesson time was spent on the Text approach (respectively 63% and 52%) and least of the lesson time was spent on the Language approach (respectively 6% and 10%). When analysing the change the lessons underwent, average time spent on the Text approach was reduced by 11% and time spent on the Context, Reader and Language approach was slightly increased (respectively 1%, 6%, and 4%). So, in comparing the average percentage of lesson time the eight teachers spent on each of the four approaches, we can conclude that after the intervention less time was spent on the Text approach and more time was spent on the Reader and Language approach (and to a very small extent on the Context approach).
138