Continental Breakfast begins at 8:00 am
Co-Hosted by:
- The Harvard Program on Survey Research
- The New England Chapter, American Association for Public Opinion Research
-
The American Statistical Association
Event Overview: This half day mini-conference will explore new ideas in preventing, detecting, and remediating fabrication in survey research. Presentations, discussion, and Q&A will focus on both practical and theoretical considerations in addressing fabrication in today's challenging survey environment, with a particular focus on fieldwork in international settings.
Presentations:
The Cheater Problem Revisited: Lessons from Six Decades of State Department Polling
Regina Faranda
Acting Director, Office of Opinion Research, US Department of State
Systems and Processes for Assuring Data Quality
Rita Thissen
Manager, Center for Technology Solutions, Research Computing Division RTI International
Preventing Data Falsification in Survey Research: Lessons from the Arab Barometer
Michael Robbins
Project Director, The Arab Barometer
Discussants:
Fritz Scheuren
NORC at the University of Chicago
Senior Fellow and Vice President
Alan Zaslavsky
Harvard Medical School
Professor of Health Care Policy (Statistics)
The event is free and open to the public.
This is part of a 2014-15 series of events and articles exploring new frontiers in preventing, detecting, and remediating fabrication in survey research. The first event, held in December, 2014, focused on interviewer-level fabrication and also touched on evidence of apparent machine assisted fabrication, where a computer is used to generate data, as opposed to more traditional "curbstoning". This event is currently being written up by the panelists for submission to the Statistical Journal of the International Association of Official Statistics (IAOS). The IAOS recently published an article by Peter Winker exploring a new method of detecting fabrication using synthetic data.
Another event is planned for June with the Washington Statistical Society, and is in the planning stages. And a session on fabrication was accepted for the Total Survey Error conference in Baltimore, MD, to be held in September 2015.
Partial support for this event is provided to the Institute for Quantitative Social Sciences at Harvard University through the Eric C. Mindich fund for conferences in experimental social sciences.