Using Linear Programming Techniques for Balancing Automatic Case Assignments in Field Surveys

• Activity Number: 253
• Type: Contributed
• Date/Time: Tuesday, July 31, 2007 : 8:30 AM to 10:20 AM
• Sponsor: Section on Survey Research Methods
• Author(s): Edward English*+ and Steven Pedlow
• Companies: NORC at the University of Chicago and NORC at the University of Chicago
• Keywords: Area Probability Sample ; Interviewer allocation

Abstract

National field surveys attempt to hire field interviewers who reside proximate to selected cases. Traditionally, case assignment was a very labor-intensive activity, with staff manually comparing national maps to determine where cases are in comparison with interviewers. NORC has been using an automatic distance-based case assignment procedure that generally assigns cases to the closest interviewer. However, many interviewers have been allocated unreasonably high work loads, while others had been assigned none. The current research uses linear programming to create constraints during allocation in order to lessen the degree of manual intervention and thus balance loads automatically. Our method realized an increase in operational efficiency without a significant loss in accuracy, and thus shows promise as a method for integrating GIS, programming, and survey research.