Methodology: The Power-Knowledge involved in the Research Process
4.4 Methodological Analysis
4.4.1 a Pupil's Written and Verbal Data
The language analysis primarily focused on the written (survey and worksheets) and verbal (interviews and group discussions) data gathered from pupils. To ensure that the data could be methodically examined, a colour-coding system was used for the survey and worksheets, where pupils were divided according to gender (for easier categorisation), with additional colours for those pupils who had used certain words to describe terrorism in the starter activity (Simons 2009:121). All recorded verbal data was transcribed into a computer and a similar colour coding system was used to aid analysis. For the teacher interviews, additional codes were used to highlight those comments relating to pupil views or possible school influences, such as lessons or assemblies.
To answer the research questions, I used Foucault's suggestion that perceptions are expressed by specific words (Foucault 2002:91) and those unspoken habits of thought (Foucault 2002:297) seen in sentences and the spontaneous language used (Foucault 2002:158). Therefore the analysis began with a specific examination of the words associated with terrorism, followed by an investigation into the descriptive language used and the historical examples, or recollected ideas, divulged by the pupils.
I began by examining the range of words used during the starter activities from both session 1 and session 2 in each case study. From this, I chose certain words that I thought would help answer the research question: in particular, “religion”, although similar words such as “religious”, “beliefs” and so on, also required some consideration.
I then discovered the frequency by which pupils used “religion” during the starter activity, because the pupils had to write words they most associated with terrorism, therefore if they wrote “religion”, then they already perceived it as somehow associated with terrorism prior to any direct input from the researcher (although the teacher interviews highlighted some of the additional considerations that needed to be taken into account for each case study). I repeated this exercise for the survey responses, paying particular attention to Q13, where “religion” was included as a motivation for terrorism (see Murphy 2010 for a similar approach to conducting a linguistical analysis associated with lexical meaning).
However, discovering the frequency by which “religion” was used as a single word only demonstrated how many of the pupils perceived it as something associated with terrorism: to better understand the scope of these connections, it was also necessary to examine where “religion” was used in sentences during the group discussions. Examining the context provided greater understanding and verification of their perceptions and highlighted associated ideas that could be used as a basis for further analysis. Foucault hypothesised the use of a “table” (Foucault 2002:74) to discover the matrix of ideas connected to a phenomenon: a similar idea to Yin's cross-case synthesis technique of approaching data through the creation of word tables (Yin 2009:156). Yin suggested that it is the pattern found in such word tables, rather than numerical tallies, that allows for interpretations of this level of data analysis (Yin 2009:160). The transcripts were particularly useful because it was from these that the context could be found: any ideas associated with “religion” were placed into a word table for each case study, and were later used as a point of comparison with the other case studies (see Appendix 10 for an example word table).
To ensure that a robust assessment of the findings was achieved, alternative terms also needed to be examined as points of comparison, to see if “religion” was actually a core idea in the pupils’ perceptions. Thus I went back to the findings from the starter activities and investigated what other terms were used, paying particular attention to any words or phrases that were comparable to those alternative motivations discovered during my literature review (and which formed the basis of Q13 in the survey). To ensure that a concise summary of the frequency data was presented, I combined certain words due to their proximity – for example, bomb and bombs were put together because one is simply the plural version of the same word; other words, such as “harm” and “hurt”, were combined due to their similarity. However, some words were ignored because they were either too broad (e.g. “people”) or only used once.
The same analytical process used for “religion” was then repeated for those words of interest for each case study (frequency counts and word tables) before the analytical data between the cases was compared: from this analysis, other ideas emerged, requiring additional frequency counts and word tables to ensure a thorough exploration of the perceived concepts associated with terrorism. Exploring these “rival” concepts (Yin 2009:160) was essential in ensuring that the research questions were answered and valid conclusions were reached.
After the analysis of these specific words, a secondary layer of analysis was conducted, to provide more insights into the symbols and representations of terrorism. I thus focussed on the details of the examples discussed, because they shed light on how pupils’ perceived specific events or groups they associated with terrorism. As Foucault
stated, the examples associated with a word or phenomenon are subject to the randomness of human recall (Foucault 2002:309) and can highlight some of the social power-dynamics at play.
For the analysis of terrorism examples, I began with the same approach discussed above: I initially read through all the data for each case study to discover any commonality in pupil responses, then highlighted a number of noticeable examples (such as 9/11 or bombers) and began with a frequency count from the starter and survey data. Afterwards, I explored the transcript data and created word-tables based on the details provided. These tables were then compared to the previous word-tables within that case study, to see if, and how, the examples were associated with wider comprehension of ideas associated with terrorism. Again, once each case study was completed, this information was compared across all case studies, to highlight any wider social power-knowledge dynamics that may have influenced the pupils' perceptions of terrorism.