Shared understanding and data quality in the SCF

Article type: Research Article

Authors: Bricker, Jesse | Kennickell, Arthur B.

Affiliations: Board of Governors of the Federal Reserve System (Ret.), Washington, DC 20551, USA. E-mail: Arthur.Kennickell@gmail.com

Abstract: In a survey, data quality is most vulnerable is at the point of data collection. Because the Survey of Consumer Finances (SCF) queries respondents on topics that are particularly technical and private, the role of the interviewer as facilitator is crucial; to be an effective facilitator, the interviewer must be well informed and motivated. To this end, SCF interviewers participate in trainings with NORC staff, and historically Federal Reserve Board (FRB) staff has reviewed completed cases for data quality issues. Although the review is partly driven by software that is used to identify many classes of potential systematic problems, the focus on reviewing individual cases has made larger data quality patterns less salient. Moreover, because the review usually proceeds with a considerable lag, it has often been difficult to get to the roots of problems, to educate interviewers where necessary and to change other procedures where appropriate. Understanding and well-conceived change rest on effective communication, but despite the best efforts of all involved, communication between the FRB staff and the field staff has been inefficient. This paper focuses on advances in the communication channel in the 2013 SCF. The work builds on the recognition that goals are easier to reach when everyone involved in the data collection process has a common understanding of them. To move closer to this ideal, some detailed elements of data quality review have been diffused through NORC staff. By engaging with the actual data, all players in the data creation process have developed a stronger common understanding of emergent issues and communicate more clearly about elements of data quality. This progress will serve as a basis for subsequent improvement.

Keywords: Data quality, communication, survey methods

DOI: 10.3233/SJI-160300

Journal: Statistical Journal of the IAOS, vol. 33, no. 1, pp. 165-174, 2017