Online Program

Return to main conference page

All Times ET

Program is Subject to Change

Tuesday, June 15
Tue, Jun 15, 2:30 PM - 4:30 PM
Balancing Response Burden and Data Quality

Impact of Instant Feedback on Data Quality Follow Up (308148)

Amy Couzens, RTI International 
*Scott Ginder, RTI International 
Bryan Rhodes, RTI International 

Keywords: quality, data collection, costs

This presentation illustrates how RTI compares current and past survey responses in real-time to improve data quality (DQ) to reduce data collection costs. Survey DQ is critical to data users, but methods for increasing DQ in household surveys differ from establishment survey methods. For longitudinal establishment surveys, trend checks are useful to ensure data are consistent with prior submissions. If trend checks and other reviews only occur after data submission, options for reconciling discrepancies can be limited and relatively costly. In contrast, resolving potential inconsistencies in the data at the time of response adds a promising new avenue to pursue clarification of because (a) corrections can be made while the information is still “fresh” and (b) the need for DQ follow-up (DQFU) contacts after the time of submission is reduced. The Bureau of Justice Statistics (BJS) Mortality in Correctional Institutions (MCI) collection supports BJS' mission to “collect, analyze, publish, and disseminate information on crime, criminal offenders, victims of crime, and the operation of justice systems at all levels of government…(providing) data…critical to federal, state, and local policymakers…” and is the only national statistical collection obtaining comprehensive information about deaths in the nation’s approximately 3,000 jails. The 2018 MCI web collection contained the typical within-form validations, plus a newly implemented instant DQFU (iDQFU) module. The iDQFU module compared the jail’s data to prior year responses and encouraged respondents to make corrections (to current or prior year values) or provide brief explanations for differences. Respondent participation in iDQFU was voluntary – survey responses could be finalized without taking action in the iDQFU module. Our presentation will provide an overview of the MCI iDQFU module, present results regarding participation by respondents, and summarize the impact the module had on DQ and other DQFU efforts.