2. Methodology and Research Design
2.8. Procedures
2.8.5. Interview procedures
Semi-structured interviews were conducted with twelve children, thirteen educational practitioners and sixteen parents. This section will outline the methodological principles and linguistic considerations pertinent to these interviews, along with practical concerns such as interview recording, location and length. First, the researcher sought to adhere to key principles for effective interviewing, such as: establishing a rapport between the interviewer and the interviewee (Fontana & Frey, 2005); starting the interview with a less serious question to put the participant at ease (Giogi, 1985); posing clear, non-threatening questions (Robson & McCartan, 2016); avoiding interrupting participants (Wescott & Littleton, 2005); and completing the interview in the allocated time (Creswell & Poth, 2018). A ‘funnelling’ interview technique (Smith & Osborn, 2015) enabled the researcher and interviewee to move from the general to the specific, and follow-up questions were often utilised depending on the flow of the conversation. Each interview was concluded by asking participants if they would like to ask any questions themselves (Magnusson & Marecek, 2015).
Interviews were as attentive as possible to potential linguistic barriers faced by participants who speak English as an additional language. Although the use of interpreters can impact the rapport between the interviewer and interviewee, and questions may arise about the accuracy of translated responses (Arnot et al., 2014; van Nes, Abma, Jonsson, & Deeg, 2010), parent informants were given a choice regarding interpretation. The rationale for this decision was to ensure that their responses were not limited by their English language proficiency, although it was acknowledged that the transfer of meaning between languages may result in a less reliable interpretation of participants’ accounts (van Nes et al., 2010). Most participants were willing to be interviewed in English, however two participants opted to have an interpreter (Baheela in Urdu and Lena in Polish) as they felt more comfortable expressing their views in their home language. Interpreters were made aware of the research aims and questions prior to
interviews for contextualisation purposes (Liamputtong, 2010). It was acknowledged, however, that the phenomenological quality of these two accounts may have been compromised, as an interpretation may not fully convey participants’ meaning (Squires, 2009).
With one exception, all interviews were audio-recorded, which provided an accurate record of interviewees’ perspectives and meant that the interviewer could fully participate, without having to take notes (Willig, 2013). One child (Marco) wanted to participate in the interview but did not wish for the interview to be recorded. In this case, notes were taken during the interview, and a more comprehensive transcript was produced immediately after the interview. Participants were interviewed individually, but children and parents were given the choice to be accompanied by a family member, in keeping with the studies undertaken by Yu (2013) and Poon et al. (2014). Four children were accompanied by an adult during the interview (parent, teacher or teaching assistant). The majority of parents were interviewed individually, however one father and one grandmother were also interviewed together with the respective child’s mother (see Table 1). For convenience, four practitioners (Suzanne, Gill, Lucy, Rachel) from the same school were interviewed together and were questioned about their experiences with two autistic pupils (Jack and Glyn) who were bilingual in Welsh and English (see Table 1). It is acknowledged that a ‘focus group’ approach was not anticipated in the study design due to concerns that the phenomenological aspect of drawing out individual lived experience may be compromised (Smith et al., 2009). However, it was deemed an appropriate method in this instance because practitioners worked closely with one another, had in-depth knowledge of the two children being discussed, and the focus group generated ‘diversity and difference’ in perspectives (Lunt & Livingstone, 1996, p.96), which, in turn, enriched the data.
The location of interviews can significantly impact the construction and deconstruction of meaning (Herzog, 2012). Parents were consulted as to the most suitable location for their child. Although it is generally considered unconventional to grant participants a choice of location for the interview (Herzog, 2012), this procedure is common within the autism literature (Brewin, Renwick, & Fudge Schormans, 2008; Dillon & Underwood, 2012; Fox et al., 2017) and may have resulted in participants feeling more at ease, leading to more honest responses (Creswell & Poth, 2018). Interviewing children at school was considered more advantageous by the researcher because it is a context in which unfamiliar adults are normative (Beresford, Tozer, Rabiee, & Soper, 2004), however five children were interviewed at home. With regard to interviews with parents, seven took place in participants’ homes, six at their
child’s school, and one in a public space, on the request of the participant. All practitioners were interviewed in school, not only for logistic reasons but also because, despite possible distractions of their daily surroundings (Gillham, 2000), respondents were likely to speak more readily in the context being explored.
As for interview length, previous studies in the autism and EAL literatures vary widely; interviews with children range from 8-19 minutes (Makin, Hill, & Pellicano, 2017) to 45- 75 minutes (Poon et al., 2014), while interviews with parents vary from 17–46 minutes (McNerney et al., 2015) to up to 2 hours (Brewin et al., 2008). The reasons for such a fluctuation include possible developmental challenges faced by children on the autism spectrum and potential linguistic barriers for both children and parents with EAL. Interview lengths with school staff in existing literature are more consistent, lasting between 30 and 40 minutes on average (Anderson et al., 2016; Lindsay et al., 2013; Liu & Evans, 2016; Makin et al., 2017). In the current research, interviews with children lasted between 7 minutes 55 seconds and 23 minutes 14 seconds (average = 16 minutes 20 seconds). Interviews with adult participants were longer as they tended to give more detailed responses than child participants. Interviews with practitioners lasted between 9 minutes 13 seconds and 39 minutes 40 seconds (average = 25 minutes 3 seconds), while interviews with parents lasted between 12 minutes 58 seconds and 46 minutes 4 seconds (average = 29 minutes 2 seconds).