Gillham states that probes should be "simple, clear, direct and potent" for them to effectively direct and focus the conversation (p46). The interviewer must appreciate that the interviewee is helping them to understand. To this end, clarification shouldn't be used as a device, though reflecting what has been said and asking for justification when a judgment is made can be very effective in gaining deeper understanding (p47). Extending the narrative, by saying "Give me an example" or "Tell me a bit more", can help to contextualise a word and query the internal consistency of what the person is saying (p44). Reflecting can indicate the emotional state behind what has been said and may result in the interviewee developing and linking elements of what they have been thinking or saying (p51). This is a significant difference from the questionnaire where there is an assumption of ready-formed, accessible opinions, which may in fact be vague and ill-defined at that point (p52).
By using Pamela for Skype to automatically save each telephone interview, and storing face-to-face ones on an mp3 recorder, I am not as pressured into transcribing as-I-go. However there is a requirement to back-up and safely store such data. By storing the mp3 files for future transcription, I am missing out on systematically and regularly informing my methodology etc based on what has aired during the interviews, as well as reflection and action which would have resulted during the transcription and categorisation process. Presently, there is no time for me to transcribe and content analyse even one of the pilot interviews. To pay for transcription at this stage would be a waste of funds. The interim solution is to listen-again to each interview and to reflect and implement improvements prior to the January interviews.
Gillham states that content analysis is "about organising the substantive content of the interview" (p59). I will identify the substantive points and then subjectively put them into categories. The derived categories will need to be exclusive and exhaustive; each statement fitting into one category. Gillham suggests transcribing around 350 words per page, using a different typeface for distinct parts.
In the medium-term, I will immerse myself in the detail of the substantive content of the initial series of interviews. I will then begin to see the significance of what people have said, although I need to be aware that while categorising, I am also interpreting (p73). As I publish findings, a purpose of some of the interview content will be to illustrate the narrative being woven through the three years of the study. I will be selective and may need to discard some very relevant quotes. Discrepant quotes, which I have already noticed, will be added to give a qualifying insight. Again subjectively, I will need to reflect general points being made, including contradictions. I am unlikely to quantify patterns of interview statements in tabular form.
When writing-up findings, I should use the introductory paragraph to reinforce a framework for holding the quotations together, then allow the interviewees to 'take over' (p77). If they have made a common point, this should be illustrated by several quotations (p78). This helps to construct an audit trail, should any of my conclusions be inspected or challenged. How I arrived at my conclusions will directly affect the peoples’ perceptions of my judgment and justification (p79).
I am attempting (and in many cases been fortuitous) to keep the study broad in terms of: gender, education sector, experience, location and age. I have not spent time on identifying a sample and there may be some justification needed later on my decision to work with 'learning professionals', which may be considered too broad in relation to the research questions.
One other phenomenon to be aware of, if it exists or comes to exist, is the elite group or elite interview (p81). Where an interviewee has authority in one or more areas of the 'social web' in learning, they have a privileged position; a wide grasp. Their perspective will be different. These participants, many of whom will have signed-up to the edonis Ning site as opposed to those without expertise or prior experience and who predominantly participate via email, possess greater knowledge about the areas, topics and settings (p82). They may have higher expectations about the interview questioning and will be used to being in control when talking about the 'social web'. I will be more likely to get distinctive views and perspectives, with insights into hidden agendas, sensibilities, key documents and records, and formal/informal relationships within the system. The interviewee will need to be carefully consulted beforehand. The evidence will have authority and the respondent may need to be listened to in future directing of the research. However, Gillham warns of respondents becoming too close an ally, which may put distance between myself and others, including those who choose to participate, for example via email, and do not currently possess significant knowledge or experience of the ‘social web’. Gillham values even-handedness (p83), which is a challenge to me, particularly when I find that since November I have been investing great effort in the edonis Ning site and methodology blog, which of course some or many participants will have no interest in utilising during the study.
