The data was analyzed by using the method of qualitative content analysis, in which the transcriptions of the interviews are analyzed to identify and categorize commonly
emerging themes. According to conventional content analysis, the data, such as interviews, is systematically classified and coded to identify themes and patterns (Kohlbacher, 2006;
Kondracki & Wellman, 2002). Content analysis aims to recognize similarities and differences in the data and to form a summarized picture of the research topic, which can be connected to a broader research context within the field (Tuomi & Sarajärvi, 2009). Qualitative content analysis is used to interpret and describe data “to provide knowledge and understanding of the phenomenon under study” (Downe-Wamboldt, 1992, p. 314).
The analysis in this study uses the inductive and data-based approach as the themes and categorizations were not built based on previous theories but recognized from the data to allow the ideas to naturally emerge without the author to test certain hypotheses or to define
categories beforehand, as proposed by Kondracki and Wellman (2002). However, in addition to inductive content analysis where categories are recognized directly from the data and not from previous theories, the results of the analysis are later discussed in the light of previous theories, and the ideas presented in the literature review are connected to the data. Hence, the analysis also theory-bound and abductive, which is an approach that connects findings from the data with previous theories to explain and strengthen the findings (Tuomi & Sarajärvi, 2009).
The qualitative inductive content analysis requires organizing the data, which is done through the following procedures: open coding, creating categories, and abstraction (Elo &
Kyngäs, 2008). The analysis process begun by defining the unit of analysis. For the purposes of this study, focusing on single words or sentences was seen as too limiting. For this reason, phrases, or passages of multiple sentences describing the same theme, were chosen as the appropriate unit of analysis. The analysis proceeded by the author reading each interview transcript carefully to form a comprehensive perception of the interviews and to familiarize herself with the topic. Next, the interviews were again read, now word by word, to recognize parts that would express opinions or feelings. After that, phrases (codes) that represented the opinions and thoughts of the interviewees were highlighted in the text. The data was analyzed to search for interviewees’ thoughts and experiences of interunit communication, which would answer the research aim of this study. Notes about the findings were created and relevant remarks made by the interviewees were noted and extracted from the transcriptions.
Based on the notes and the highlighted parts of texts that represented the emerging themes of the interviewees’ thoughts, an initial coding scheme was developed. Through recognizing similarities and differences in the data, codes were then organized into meaningful clusters and to further describe and organize these clusters, tentative sub categories were developed. When working through the interviews with existing codes and
categories, code clusters and sub categories were revised and combined into main categories, if necessary, to form the most meaningful and comprehensive categories. This process of abstraction, in which categories are developed step by step, aims to finally form a generic description of the research theme (Elo & Kyngäs, 2008). The data was examined and
abstraction continued until workable. The final categories are presented in Chapter 5 in Table 1. After analyzing the data, the results were considered and reflected in the light of previous research and theories, and conclusions of the analysis were made to answer to the research aim of this study.
It is acknowledged that the method contains an implication of reliability and validity issues. As this type of qualitative content analysis is mostly based on the researcher’s categorization of the data, author’s subjective constructions may have an impact on the interpretation of the interviewees’ answers, as suggested by Silverman (2005). Language is ambiguous and certain expressions may contain multiple meanings, and one person might interpret some content differently than another individual. Nevertheless, the interpretative approach to the data offered an in-depth and a holistic understanding to the research theme, and enabled a comprehensive examination of the informants’ emotions and thoughts.
5 Results
In the data presentation, the research data has been divided into three main categories and the results are presented according to these categories. The main categories include subcategories, which are presented within each main category. Next, each category is described and discussed in detail, and presented with extractions from the original data to represent the perceptions of the informants. The final categories were develop based on the relevancy of information they provided related to the research questions of this thesis.
Additionally, the categories include information that was seen as relevant for the case company’s needs and current situation, although not directly related to the research questions.
The categories are presented below in Table 1.
Table 1
Main category Subcategory
1. Organizational structure and interunit relationship
1. Local autonomy versus global control 2. Dispersed units in interorganizational network
2. Interunit knowledge management 1. Knowledge-flows
2. Communicational policies and practices 3. Expatriate as bridge builder
3. Language diversity 1. Common corporate language and language management
2. Language competency
3. Language as value of internationalization 4. Language as gatekeeper of interunit communication and grouping
Details for example about the interviewees position are not mentioned when presenting examples of the data to ensure the anonymity of the interviewees. The interviewees are referred to either Dutch or Finnish interviewees based on their working location, as some of the examples can be only understood when referring to the country unit they work in.
Next paragraphs present and discuss the results in accordance with the above-mentioned the main- and sub categories.