In 2006, the BHPS introduced dependent interviewing (DI). In this paper, we examine DI as a questionnaire design strategy to determine change in respondents’ circumstances affects the interaction dynamic between interviewer and respondent. Analysing transcripts of 64 survey interviews, we find that the interview dynamic between interviewers and respondents depends on whether change has occurred in the respondent’s situation.
Focusing on an application of DI that gathers data on current employment characteristics, find that the introduction of DI in the BHPS rendered the series of questions repetitive which lead to significant shortcutting, in the form of interruption, on the part of respondents who did not experience any change in circumstances. We also find that when a respondent’s current
employment situation is no longer the same as the information presented to them – that is, under circumstances of true change – standardized interviewing breaks down. Respondents engage in stray talk which provides information relevant to answering subsequent questions, or often renders the answers to subsequent questions obvious.
We reason that this is due to the conversational tension inherent in survey interviewing which is enhanced through the use of DI. Under conditions of true change, respondents seem to infer the answers required. They provide additional information which renders the interaction at subsequent questions strained by a tension between the directive for standardised procedures and conversational norms. Though a failure to follow rules of standardisation may not impact data quality negatively, per se, we find that the sorts of errors resulting from this breakdown in the DI context are of a type others have found to negatively impact data quality.
Presented by:
Noah Uhrig, ISER and Emanuela Sala, ISER
Date & time:
February 4, 2009 1:00 pm - February 4, 2009 2:00 pm
Internal seminars home