I want to interview non-native speakers of English, but I am not a native speaker, so to avoid certain implications of this (unnatural setting, accommodation, etc.), I am thinking of sending them the questions and ask them to record themselves. Has anyone done this before? Are any methodological drawbacks I should consider?
Thanks!
Self-produced audio records can be made. We analyze them with NVIVO-Plus.
The only methodological problem is that the interview protocol should be very detailed. Similar systems, with self-registration, are used in distance language classes (web).
I would look at the quality of the interaction between the language barrier. Is your sample involved in some way other than the questionnaire?
Hi, I am looking at phonetic variation and attitude towards accents, so the questionnaire and reading lists.
Unless you're following Labov's (1972) vernacular principle as a guiding principle, the only methodological issue that's worth considering the population that your results can be generalized to when not using an interviewer.
As far as the content of their answers, I don't think you'd have any problems, but in analyzing their pronunciation, you'll basically only be able to say, "This is how these speakers pronounce these sounds in the context of responding to questions from an imaginary interviewer in a room by themselves." I'm not sure that's actually useful data, but perhaps it would still serve your purposes, depending on what your purposes are.
I can think of two alternatives, though:
1) Have an assistant who is a native speaker or who is part of the interviewee's community in some way perhaps do the interview. As with all sociolinguistic interviews, you'll still have to limit your conclusions to "this is how these people speak when being interviewed by this sort of person", but it's at least a more natural context than answering written questions out loud.
2) You can use Perrot's (1995) strategy and have two interviewees sit down to discuss their answers to each question with each other. The downside here would be that they could influence each other's answers, but the upside is that you've created a fairly natural context. In Perrot's case, she was studying speakers of a mixed language, so she needed them to speak to each other to ensure that they'd actually mix their languages seeing as they might not do so if they perceive the interviewer as being a standard speaker of one language or another and not someone who can easily parse a heavy mixture of the two.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics
- Survey Design
- Questionnaire