5.6 Interview length

In general, the rule applies that the longer an interview lasts, the more data is gathered. There are, however, strongly differing opinions on the matter of ideal length. Milroy and Gordon (cf. 2001: 58) cite Labov's (cf. 1984: 32) notion that an interview should ideally be 1-2 hours long, although they mention that 20-30 minutes can be enough to record interesting phonological data. If fluctuation in speech is of interest, they also quote Douglas-Cowie (1978) who argues that after about an hour a person usually settles down and uses more relaxed speech, even when speaking to a stranger. They contrast this with the opinion of Schilling-Estes (1998) who believes that a conversation moves in and out of styles for various reasons (strategies to help tapping into the vernacular are presented here). So apparently it is impossible to define the perfect interview length, it really depends on the research-project and its intention. In general, the longer the interview lasts, the more opportunities for studying the vernacular will present themselves to the researcher. But as a general rule interviews should not last longer than three hours as the informants tend to get bored after this time (cf. Crowley 2007: 95). Even though this time frame is suggested to be appropriate especially for the more tedious work of eliciting data from an undescribed language through lists of words and questions on grammar, it obviously proves equally applicable to all forms of one-on-one or small group interviews (other forms of interviews are described here).

Created with the Personal Edition of HelpNDoc: Free Web Help generator