I DIOGRAPHIC C ASE S TUDIES
5 C HAPTER F IVE : D ATA ANALYSIS
5.6 V ERIFICATION AND VALIDATION OF DATA
5.6.1 Pilot Study Pilot Study Pilot Study Pilot Study
The semi-structured interview was piloted in September, 2014. The purposes of
undertaking a pilot study include to refine data collection plans by considering research
design and field processes, and to pilot the design of the interview (Yin, 2013). Van
Teijlingen and Hundley further recommend it as a way of pre-testing a particular
research instrument (2002).
The participant for the pilot of the semi-structured interview, TT, was recruited
after she attended a conference at which the researcher gave a short presentation about
her proposed research. Although TT’s daughter ‘Ellie’ (a pseudonym) was at that time
in full-time education, there had in the past been a number of periods when Ellie had
been withdrawn from school to be home-educated for one or more days a week. Ellie at
the point of participation was eleven years old and had been diagnosed with autism at
age 2.
The researcher attended TT’s house on 15th September, 2014, and recorded an
hour's interview. TT was given information about the study prior to the interview (both
verbally at the conference and in the form of the Information for Participants sheet – see
Appendix 1) and had the opportunity to discuss the study with the researcher or her
supervisors before taking part. She signed the consent form (Appendix 2) prior to the
interview and kept a copy of both it and the information sheet for future reference.
During the interview the researcher used the semi-structured interview schedule
narrative to flow freely (Gray, 2013; Van Teijlingen, 2014). This pilot prompted a
number of considerations:
Establishing protocols
Prior to the commencement of recording of the interview the researcher went
through the interview process, elucidating TT's freedom to stop the process at any time
and articulating what was going to happen to the data collected (transcription, inclusion
in this thesis and possible publication and presentation at conferences). On reflection, it
was decided that it would have been an advantage to include issues such as gaining
informed consent, permission to record and issues of anonymity during the recording
phase so that a record of them existed on the audio recording and in the transcript.
Use of pseudonyms
The researcher assured TT that all names would be taken out of the transcription
of the interview. However, she did not discuss the issue of what identification would to
be used instead. When she came to transcribe the interview and to consider the use of
pseudonyms, this raised several questions as to what identification should be used and,
indeed, who should choose it. The researcher’s response to these issue is discussed in
3.4.1.7 of this study.
Semi-structured interview schedule
The interview schedule asks open-ended questions in a way designed to enable
the participant's narrative to emerge naturally (van Teijlingen, 2014). Prompts during
the answering of these during the pilot study showed some tendency towards closed
questioning; the researcher did not not always enable the participant to 'tell her story'
Etherington, 2013; van Teijlingen, 2014) to emerge, and can be used to motivate
interviewees to given additional information (van Teijlingen, 2014). For example:
TT: Year One? I had a few problems with the school. She went missing a couple of times. I did warn them that she will go walk-about … I mean, one time she actually walked home, which was a worrying thing.
CL: When she was four? TT: Yes. Yeah – very young... CL: How far was that?
TT: For her, that was probably a good five-, ten-minute walk. CL: Across roads?
TT: Yeah. Not main roads, but bad enough …
CL: So, what did you do with school then, at that point?
TT: Well, I rang them up and said, '...Can you tell me how she is?' … and there was a very long pause, and finally they said, 'Oh, she's away doing something at the moment'. And I say, 'well, I'd quite like to find out. I'm quite happy to wait on the phone...' and eventually they had to admit that they didn't know where she was.
Telling of stories of this kind can yield rich data as 'people by nature lead storied
lives and tell stories of those lives' (Connelly & Clandinin, 1990, p. 2). However, the
transcript made the researcher aware of the sometimes closed nature of her
interruptions/prompts, and she made fewer, and of a more open nature, during the
formal data-gathering phase of the study.
Off-topic material
TT had a tendency to follow her own agenda and spent much of the interview
time 'off topic' as far as this research interest was concerned. She gave a number of
stories relating to her other children who do not have autism and with whom she has not
shared education. This prompted the researcher to consider to what extent she should
‘head off’ such narratives. During the formal interview phase of the study she allowed
such narratives to occur when to stop them risked interrupting the participant’s ‘flow’,
considered the option not to transcribe any content was truly ‘off topic’ (although, in the
formal data-gathering phase, off-topic content was rare and confined to interruptions
and interactions with third parties).