Using survey paradata for the fieldwork corrections and optimization in adaptive design

Abstract

Paradata are data about the data collection process (interviewers’ observations, number of phone calls or visits to a household, time spent completing the questionnaire, etc.). The article considers the opportunities to use paradata in adaptive design where separate elements may vary depending on various categories of respondents during the fieldwork. Researchers should take into consideration both the fieldwork costs and nonresponse bias. The paradata analysis shows that interviewers’ assessments and observations (description of the district and house, assessment of the likelihood of the household’s participation in the survey) can be accurate enough to be used in adaptive design. Date, time and number of contacts help to predict the optimal day and time for the interviewers’ contact; researcher can prioritize certain categories of respondents to increase their response rate during the fieldwork. The author also considers shortcomings and difficulties in using the paradata in adaptive design.