Forming the preliminary classes from the material

Ruusuvuori and Nikander (2017, p.368) state that the analysis method determines the way to do transcription. In this research, it is relevant to study how experts speak about gender and equality through themes that will be found from data. Therefore, word for word transcription was enough to analyze and interpret material in approvable level. It was crucial to know what the informants say, rather than how do they say something. The rhythm of speaking was marked with punctuation marks and full stops. Conducted interviews were

relatively intense, and in discussion with the informants, we managed to keep the focus on the topic of interest. Therefore, there was no need to notably leave big parts of the text out.

The thematical analysis started already during the transcription of interviews. Eskola and Suoranta (1998, p.126) describe that in thematic analysis certain topics rise from data and enlighten the research problem. When I started to read through texts in the very first phase of classification, my mind was already processing and analyzing interviews. Through analyzing themes related to experts’ ways to speak about equality and gender, and even in the phase of classification it was possible to see how perceptions and definitions of gender and gender equality vary.

After the first active reading researcher starts to make a classification. The researcher moves through the classification of the material to theoretical pondering and analysis and finalizes the process by reducing the material. (Hirsjärvi and Hurme, 2015, pp.143-144.) In this study, the rotation of classification, combination of theory and analysis of themes that were found by classifying material, was repeated three times. Firstly, I created the first classification project in Atlas.TI program, which is a tool for qualitative analysis. I read through all texts systematically marking similar ways to speak of gender, to find preliminary classes.

Simultaneously I read previous literature to be able to link the classes to suitable concepts.

Classes are conceptual tools, that help with finding a suitable theoretical framework. In addition to the conceptual nature of classes, they also must have an empirical relation to the phenomenon that is studied and be aligned with the data in practice (Hirsjärvi and Hurme, 2015, p.147.) For me, forming classes was necessary when I wanted to name the key characteristics of the material.

Relations of classes, material, and theoretical concepts are constantly compared. I let the text content lead and bring out possible concepts. Therefore, this analysis slightly applies grounded theory. Grounded theory is described to be a methodology to create a theory, which is embedded in the data. (Hirsjärvi and Hurme, 2015, pp.164-168.) While searching for classes, I must constantly draw attention to how classes are related to each other’s and simultaneously be part of the relevant theories. After the first classification of the material, I exported the list of classes from Atlas.TI program to start working on them.

Secondly, I reorganized and reduced the list of classes and read literature related to the conceptual classes, and Finnish policy papers regarding gender equality and focus on women and girls in development policies. Regarding Hirsjärvi and Hurme (2015, pp.147-149) the process of classification happens always with the research problem, data, and theoretical understanding of the researcher. Therefore, I aimed to keep the research question in mind all the time, especially while reducing classes.

The result was a structure of classes in three levels; under three main themes, there were subgroups on secondary and tertiary levels. Informants spoke about their perceptions of gender equality and its implementation through three main classes that are: Significance of gender equality, implementation of gender objectives, and cooperation between Finland and Namibia. The main classes are partly overlapping but still have certain differences. The significance of gender equality is merely abstract and performs at a theoretical level and reflects informant’s perceptions considering general policies. Whereas the implementation of gender objective brings out how experts tell about their acts to obtain the goals, that gender mainstreaming possesses. Cooperation between Finland and Namibia overlaps with other classes the most often and explains the informant’s interpretation of the relation of Finland and Namibia within gender equality. After this phase, formed codes were tested in practice and brought to the texts. I started to apply new tertiary level codes into texts in Atlas.TI, aiming to reduce and condense the material. First, I worried that would all codes fit the text, and would primary and secondary codes have equivalence with the significance of the text.

As Hirsjärvi and Hurme (2015, pp.143-144) present, research material was broken down into smaller entities and classified in analysis, whereas in synthesis I aimed to construct a general view and present the phenomenon of interest in the new perspective. In the third phase of classification, my challenge was to decide which of the classes were the most crucial ones, and for that, I had to read previous literature again to create cohesion and combine classes.

I exported all citations from Atlas.TI and I started making synthesis and combining the material with previous literature.

For further analysis, I chose tertiary level classes by their recurrence and tendency to shape discussions. After that, I reconstructed four new main themes based on selected classes. Four new main themes were Finnish country promotion, communication and impact, development policy priorities and changes, a realization of equality between Finland and Namibia and

Equal rights and opportunities of women. These four main themes with one level of lower classes would form preliminary chapters of the analysis rapport.