Using Computer Audio-Recorded Interviewing to Assess Interviewer Coding Error
Computer-assisted audio recordings provide a new approach for detecting and correcting interviewer coding error. For questions with categorical or “other specify” responses, it is possible for the interviewer to misinterpret, abbreviate, or improperly key the respondent’s answer. In this paper, we discuss the utility and effectiveness of computer audio-recorded interviewing (CARI) for detecting how accurately field interviewers capture responses to open-ended questions with pre-coded response lists. CARI is used in field studies as a means of monitoring interviewer performance and detecting interviewer falsification. For this study, we evaluated coding accuracy by comparing the keyed data with the audio recording of what the respondent said. We examined questions that differ in terms of the length of the response lists, the complexity of the concepts being captured, and the availability of an “other specify” category for capturing verbatim responses. We present the type and rate of errors detected, and discuss implications for questionnaire design, interviewer training, and data quality. Findings are based on the Study of Community Family Life, an in-person survey of 13,000 households in urban areas, sponsored by the Administration for Children and Families, Department of Health and Human Services.