The decision to gather data through interviews consequently affected the choice of method for data analysis, and thus, this process was conducted by using the content analysis method. In qualitative research, content analysis usually focuses on different forms of textual data, i.e. documents, such as transcribed interviews, written reports or journals, with the aim of verbally describing their content by categorizing or summarizing the chosen data (Tuomi and Sarajärvi 2009: 103-106). The result of effective content analysis is a coherent and comprehensive description of the studied material, which may eventually be linked to existing theories and earlier studies on the topic (Tuomi and Sarajärvi 2009: 103-108). More precisely, Tuomi and Sarajärvi (2009: 99) define three varieties of qualitative content analysis: a theory-based (teorialähtöinen), a theory-driven (teoria-ohjaava) and a data-based (aineistolähtöinen) approach. In the theory-based approach, the process of data analysis is heavily guided by previous scientific theories or models in the field (Tuomi and Sarajärvi 2009: 97). For instance, the theory-based approach may focus on the testing and application of existing models in new scientific contexts, a design preferred in studies of natural sciences (Tuomi and Sarajärvi 2009:
97). In the second option, in the theory-driven approach, existing knowledge and theoretical background may be utilized to guide the analysis process, so that the findings are eventually linked to some existing theory. However, the analysis may also introduce wider perspectives and new ideas on the topic, as long as the connection to predetermined theoretical background is maintained (Tuomi and Sarajärvi 2009: 96-97).
In the data-based approach, on the other hand, the significance of data is prioritized, allowing existing theories to be adjusted to the data and the findings deduced from it (Tuomi and Sarajärvi 2009: 95).
This approach was chosen as the method of the present study, since it let the research questions function as the guideline for data analysis, rather than obliging the analytical focus to be predetermined by theory. Effectively, data-based content analysis moves from empirical data towards conceptual understanding of the topic (Tuomi and Sarajärvi 2009: 112), which seemed an appropriate approach for the study at hand. It was preferred to let the analysis be more data-based and guided by the research questions, than make it strictly rely on theoretical framework. Another reason for this choice was the complexity of existing terminology surrounding the study unit at hand. Even though as the researcher I was familiar with the basic concept of the team period and its working methods already before the process of data collection, the essential theoretical framework was difficult to determine beforehand. Depending on the point of view, the team period may be regarded as an educational experiment incorporating multiple different approaches, such as phenomenon-based learning (PhBL), content and language integrated learning (CLIL), collaborative learning or
49
integrated curriculum, to mention a few. Due to this overlap of educational approaches, it was preferred to prioritize the data and its implications on the topic of EFL studies in the team period first.
Afterwards, these findings could be used to determine and be linked to the essential theoretical concepts.
As Tuomi and Sarajärvi argue (2009: 95-96), the goal of the data-based approach is to form a theoretical understanding of the data, in this case, based on the interviews and the findings deduced from them. The researcher has to organize the collected data into a concise and explicit form, so that the dispersed pieces of information gathered from several participants may be presented in a more meaningful and coherent manner (Tuomi and Sarajärvi 2009: 108). Thus, in qualitative research, it is essential that the data is not presented merely as a list of isolated statements or as ‘results’, but instead, the researcher should succeed in drawing meaningful conclusions on the basis of the data (Tuomi and Sarajärvi 2009: 103). In order to meet this objective, the data must first be deconstructed and conceptualized, and only later reconstructed into a logical entity in the form of categories or themes, thus introducing the most essential phenomena concluded from it (Tuomi and Sarajärvi 2009: 108).
Consequently, the analysis process of the present study consisted of three stages, according to the design presented by Miles and Huberman (cited in Tuomi and Sarajärvi 2009: 108): data reduction, data clustering and data abstraction. The analysis process began by listening and transcribing the interview recordings. Then, in the data reduction phase, the essential extracts related to the research questions were color-coded and reduced into simplified versions. Next, by collecting and assembling the extracts describing similar topics, they were clustered into sub-themes. Evidently, this data clustering process contributed to the formation of the present study’s structure. Afterwards, on the basis of the sub-themes, the clustering process was repeated to place them into main themes (Tuomi and Sarajärvi 2018: 124-125). The main themes and sub-themes of the study are presented in the following table:
Table 2. Clustering of sub-themes and main themes
Sub-themes Main theme
- The planning process
- EFL teachers’ approaches to integration - The role of formal EFL lessons
- The use of English language in the projects
Integration of EFL and the English language
50 - Affordances for EFL teachers’ professional development
- Promotion of spoken language skills and confidence in speaking
- EFL and English language in contact with other subjects
Affordances of EFL integration in the team period
- Time restrictions and lack of diversity of EFL studies - Defining the role of EFL in the team period
- Disproportionate influence of weekly projects on the EFL course grade
Challenges of EFL integration in the team period
Ultimately, the analysis yielded ten sub-themes, grouped under the three main themes presented in the table. The following chapter of this thesis presents and examines the above themes in more detail.
Lastly, the results of data abstraction process – the final stage of analysis – are discussed and concluded in the last chapter of the thesis, as the information of the present study is used to form theoretical descriptions on the topic (Tuomi and Sarajärvi 2018: 127).