|
Abstract:
|
The Agricultural Labor Survey (Ag Labor) is a bi-annual survey administered for the purpose of collecting data on the number of hired workers, average hours worked, and wage rates. The Ag Labor survey has a panel design, which means that for each data collection session, respondents are asked to provide the same information for two different reference periods. This request may place unnecessary burden on respondents and can also lead to false rates of change due to variation in question interpretation, recall errors, and variation in open-ended response coding. The current study uses mixed methods to explore the impact of interviewer behavior on data quality by assessing respondents' response process when presented with currently reported data (CRD). In addition to the effects of CRD on data quality, the results of panel surveys are also prone to seam effects. This effect can produce fabricated changes that could lead to misinterpretation of the study results. The study also explores the impact of seam effects on data quality by assessing the magnitude of difference for various variables.
|
