|
Abstract:
|
In multiday diary surveys such as the Food Acquisition and Purchase Survey (FoodAPS-1), households make participation and reporting decisions every day. While some households remain committed throughout the reporting period, others may attrite or only respond on a subset of the days. This results in varying nonresponse patterns among households at the day level. Previous research (Hu et al. 2019) suggests that the current household level nonresponse adjustment may not adequately correct bias in estimates arising from households that initially agree to participate in the FoodAPS-1, but then fail to report on specific days. Furthermore, individual members within a sampled household may also make their own participation and reporting decisions. The current household-level nonresponse adjustment methods ignore these within household nonresponse patterns. This research project aims to address both these deficiencies, namely the lack of adjustments for nonresponse at the individual household member (person) and day level, by developing a set of person-day level nonresponse adjusted weights for the FoodAPS-1.
|
