4.4.3 Semi-Structured Interviews
Interviewing is a process in which an interviewer and interviewee(s) engage in a ‘conversation with a purpose’ (Burgess, 1984, p.102).It is a neutral device to cast and assemble many facts in a participant’s field-based events (Kvale, 1996; Silverman, 2007; Yin, 2014).In linguistics, for example, some researchers use interviews to explore ‘what’ (e.g.Wagner & Wodak, 2006; Anderson, 2008) or ‘how’ (e.g.Holstein & Gubrium, 1995; Litosseliti, 2002; Tilbury & Colic-Peisker, 2006) the responses or narratives are grounded in and underpin different research purposes.Despite its widespread use in research, some researchers (e.g., Potter & Hepburn, 2005a, 2005b; Wolfson, 1976) criticise that the method of interviewing produces unnatural or manufactured data and a manipulated interviewer-respondent relationship.They claim that segmental data extracted from the whole interview seem to be researcher-preference oriented and context-free, and also that an interviewer designs the topics for his or her research’s sake and consequently holds a powerful role in the process of interviewing.
These criticisms seem to mitigate the quality of interview data, though there is no one single perfect research method to completely cover all variants in a phenomenon.As Labov (1972) observes, ‘intersubjective agreement is best reached by convergence of several kinds of data with complementary sources of error’ (p.97).Interviews can be good cross-reference sources of data to prompt and probe into scattered social, historical and individual information on participants.Not all interactional events in the real world are logically or systematically connected.Interviewing participants on certain pieces of information will create and accumulate knowledge to understand the participants’ life experiences and viewpoints, and then sufficiently approach particular research questions.
The semi-structured interview is the most commonly used to reach the ‘compromise’ (Dörnyei, 2007, p.136) between structured interview and open interview, yet it is not a mixture of the other two types.The structured interview is the most orally controlled form of written survey, and its wording and order of questions are predetermined.In contrast, the open interview is more like an unstructured conversation conducted in a flexible and exploratory manner.Standing in the middle of these two extremes, the semi-structured interview is ‘more open-ended and less structured’ (Merriam, 2009, p.90).It is guided by a list of questions and flexibly generates specific data from the interviewees.This requires a semi-structured interviewer to design not only clear topics covering the research purpose(s), but also intended topics enabling the mature development of the interview.
The semi-structured interviews conducted in the present research were to obtain as much rich sociocultural, historical and individual information from the participants as possible.As shown in Table 4.3, interviewees included six Chinese students and three non-Chinese teachers from the School of IC, as well as five non-Chinese students from the School of Social Science.
Table 4.3 Summary of interviewees
Notes:All pseudonyms.
The three non-Chinese teachers were those who delivered the filmed seminars.The five non-Chinese students were randomly introduced by three of the six Chinese student interviewees.Interviewing other members in the community helps ‘enter into the other person’s perspective’ (Patton, 2002, p.340) and find the variables of feeling, thought and experience which cannot be directly observed.Furthermore, as all non-Chinese interviewees came from different ethnic and cultural backgrounds, it was interesting and necessary to see how they interpreted each other’s social behaviour in similar or different ways, and in turn their interpretations would reveal the representation of the Chinese Humanities students’ social and individual identities through their recognition and negotiation of group membership in the community.
Before carrying out the interviews, I emailed each of the interviewees the list of questions and determined the interview places as to their willingness, either in a café or a quiet room.At the beginning of the interviews, I restated the research and interview purposes to them.The interviews were audio recorded and took place in either Chinese or English.The process of interviews denoted that a fertile researcher-participants relationship (Spradley, 1979) facilitated the generation of profound interview information, which benefited from the familiarity and mutual trust established during the initial free talks and personal contacts.
The interviews were fulfilled over two years.In 2014, I interviewed five non-Chinese interviewees from the Schools of Social Science and Science.I also conducted the first round of interviews with the six Chinese student participants and the three non-Chinese teachers from four different years after the video recordings of seminars.The interviews were intended to identify sociocultural, historical and individual variables and the interpersonal relationships between the interviewees and other group members in the community.They also implemented the interviewees’ reflection and identification of certain learning or teaching practices in the seminars.In order to track the history of the Chinese student participants’ learning experiences with reference to the group membership change in the context, the second round of interviews were performed as follow up.In 2015, I interviewed the same non-Chinese students whom I had interviewed in 2014, except Sara.The contact with the previous year 3 student Sara could not be made after she graduated and left the university.The interviews with the teachers were all in English, whereas Chinese was preferred in the interviews with Chinese students to eliminate their anxiety and any misunderstanding or vague expressions caused by a foreign language.
Design of the semi-structured interview questions followed Patton’s (2002) six types of interview questions; namely, experience and behaviour, opinion and values, feeling, knowledge, sensory, and background.Patton suggests that these six types of questions are not necessarily sequenced.Drawing on them, a matrix of the two rounds of interview questions was synthesised in an intertextual meaning of what happened, what is happening and what will happen along a timeline of past, present and future.What happened in the past in fact presupposed the students’ present learning experiences and attitudes, and the present reflexively pointed forward to what will happen in the future.The analytical results would facilitate an temporal-spatial understanding of disciplinary identities in the specific context.
Open-ended questions were asked in the interviews, because they engendered the interviewees’ unique definition of situation (Silverman, 2007).Patton (2002) advises that open-ended questions preferably begin with noncontroversial ‘what’ questions about present activities, experiences or behaviours, followed by ‘how’ questions to probe into interviewees’ opinions and feelings that are based on their descriptions of the events.He disagrees on taking ‘why’ questions into account, which in his explanation will close the questioning beyond a qualitative researcher’s expectation.However, Merriam (2009) argues that ‘why’ questions sometimes are good for clearing up speculation or creating new possibilities of questioning.The current SFL-oriented research follows the tradition of viewing a repertoire of text, spoken or written, as both an object and instrument (Halliday & Matthiessen, 2014).I hence tackle with ‘why’ questions in the interviews.