Chapter 3 Methodology
3.4 Data analysis
3.4.2 Content Analysis versus Grounded Theory
Cohen, Manion and Morrison‟s (2007) discussion of qualitative data analysis provides a comparison between Content Analysis and Grounded Theory, which helped me to decide which to apply to my data. They define Content Analysis as follows:
“Put simply, content analysis involves coding, categorizing (creating meaningful categories into which the units of analysis – words, phrases, sentences etc. – can be placed), comparing (categories and making links between them), and concluding – drawing theoretical conclusions from the text.” (Cohen, Manion and Morrison, 2007:476)
Since this form of analysis is systematic and verifiable (through reanalysis and replication), this description of its features seemed to make it suitable for the purposes I had in mind. However, Cohen, Manion and Morrison (2007) also argue that Content Analysis has strong positivist overtones, since it is often used to detect the relative frequency and importance of certain topics, and uses statistical techniques to do this. For example, their Step 9 (“conducting the data analysis”) describes how “once the data have been coded and categorized the researcher can count the frequency of each code or word in the text, and the number of words in each category” (Cohen, Manion and Morrison, 2007: 481). This approach implies that frequency is an indicator of importance, and as these authors point out, this may not actually be the intended meaning of the interviewees: “Frequency does not equal importance and not saying something (withholding comment) may be as important as saying something” (Cohen, Manion and Morrison, 2007: 481).
I strongly agree with this contention, and regard the assumption that the frequency of interviewees‟ comments on a theme should be considered as an indicator of importance as highly problematic. For example, I was aware during my interviews that certain students refrained from making comments which
could be taken as direct criticism of their tutors or other students. Whatever the motivation behind their choices of what to include and exclude from their comments, treating the interview transcripts as complete and quantifiable sets of evidence would clearly be contrary to the theoretical perspective of this thesis, and lead to an ineffective means of analysis since certain aspects of my participants‟ experience would need to be inferred by “reading between the lines” – a process which seems contrary to the positivist thinking underlying Content Analysis.
In sum, whilst Content Analysis offered valuable advantages of being systematic and verifiable, Cohen, Manion and Morrison‟s (2007) view that it is used to draw conclusions from qualitative data by counting, patterning and clustering convinced me that it would be inappropriate for my aim of exploring students‟ accounts of their experiences. Furthermore, Content Analysis seemed to require applying a pre-determined set of codes and categories to the data in order to test pre-existing theory, whereas I was more interested in exploring the data with a view to building new conceptual categories. Since my approach in this project was exploratory, I needed a more open-ended and flexible analytical method which would allow my investigation to move into a number of different directions. I therefore decided to use an analytical approach influenced by Grounded Theory.
According to Cohen, Manion and Morrison (2007), in seeking to build theory which is grounded in the data, Grounded Theory, like Content Analysis, relies on systematic data collection and analysis. However, unlike Content Analysis, which tends to reduce the complexity of the data by applying codes and categories, Grounded Theory seems to pay more attention to the complexity of context: “It takes account of apparent inconsistencies, contradictions, discontinuities and relatedness in actions” (Cohen, Manion and Morrison 2007: 491). Flick (1998: 41), cited by Cohen, Manion and Morrison (2007: 491), sees the aim of Grounded Theory as: “not to reduce complexity by breaking it down into variables, but rather to increase complexity by including context”. In practice, this aim of increasing complexity is achieved by first coding comments and then proceeding to identify categories which these comments appear to suggest. Categories are then either supported or undermined in an iterative
process of “constant comparison” with the data (Glaser and Strauss, 1967) which allows the researcher to check and elaborate on their specific characteristics.
An analytical procedure which respected the complexity of context seemed entirely appropriate for this thesis since I was consciously attempting to work counter to the typologies evident in some of the earlier work on Chinese learners. However, the main advantage of Grounded Theory for my data analysis was its avoidance of pre-determined theory: “Grounded theory starts with data, which are then analysed and reviewed to enable the theory to be generated from them; it is rooted in the data and little else” (Cohen, Manion and Morrison 2007: 492). Although, as I have made clear in the introduction, my choice of research topic is rooted in my professional experience and reading, this experience did not result in any choice of pre-determined theory before I embarked on my research. As the literature review demonstrates, I certainly developed an attitude of suspicion toward particular research approaches (positivist, statistical, “large culture” approaches), but this suspicion prompted me to adopt an exploratory approach through which I hoped to obtain a rather more complex understanding of students‟ experience of a particular set of pedagogical styles, rather than to test specific theories. It therefore seemed clear to me that Grounded Theory was an appropriate approach on which to base my analytical method.