Ea1uatmg Data Collection Qualit3 After Piotocol Changes

It is important to evaluate data quality in any survey, but especially after changes in protocols and measurement systems. When revisions have been made in a longitudinal survey, it can be particularly important to make estimates of error rates for the entire data collection process and possibly for specific domains associated with the process. One examines this problem from the context of the National Resources Inventory, a longitudinal survey of the nation natural resources. The goal was to evaluate the quality of the data collection process after protocol and organizational changes were implemented for the 2005 survey. Because estimates of error rates were of interest, one used a probability sample to select segments for quality review.