Conferences

Although children are increasingly seen as competent informants we know very little about their cognitive and interactive skills in interview settings. To tackle problems of construing adequate research instruments for children (of different age groups) we need to take those skills into account. A specific problem in any research is item-nonresponse. Don’t know (DK) as well as substantive response based on weak information or mere guessing undermine the data quality in social research. By applying cognitive information-processing models I try to explain children’s DK answers in interviews. Only with a deeper understanding of children’s cognitive information processing we can construe research instruments that are adequate for children.

The analysis is based on results from a semi-structured approach to interview children aged 5 to 11. Each of the 56 children was interviewed twice, once on the telephone and once face-to-face. Altogether 112 interviews resulted. By this qualitative procedure the context of the given DK answers can be analysed and inferences about reasons for DK answers within the question answer process can be made. Research Questions are: What causes DK responses in interviews with children? Where in the cognitive question-answer process can reasons be allocated? Does age as a correlate to cognitive abilities have an effect on either the total number of DK responses or the “cognitive” reason? What do response latencies tell us about the cognitive state of the interviewee and the validity of his (DK) response? Are there specific interactional patterns that accompany DK responses?

The qualitative approach can be used as a vehicle also for a better understanding of question-answer processes in more standardised interviews.