Conferences

With the recent emphasis on analysis of collection process data (“paradata”), paradata research has been focusing on a better understanding of the data collection processes, phases in the collection process, and a link between the collection strategies and collection outcomes. Various tools have been developed for survey managers to better assess the data collection progress and put them in a position to adapt their collection strategies on the fly (active management). The need for a more flexible approach was acknowledged and Responsive Design Strategy was proposed (Groves and Heeringa (2006), Laflamme and Mohl (2007)).

In essence, Responsive Design Strategy aims at adjusting data collection strategies throughout the collection period in response to how well the collection progresses. Responsive Design needs to be simple, cost effective and relatively low-risk to avoid any negative effect on the quality of data and resulting estimates. Responsive Design has to take operational constraints into consideration and has to be endorsed and supported by the data collection managers for this Strategy to work. Finally, Responsive Design should ideally help better balance collection costs and data quality by refocusing the collection effort accordingly.

The first part of the paper introduces the key components of the Responsive Design Strategy proposed within Statistics Canada. The second part of the paper presents practical issues related to implementation of the Responsive Design Strategy and how the theoretical approach has to be transformed to adapt to the realities of data collection mechanisms on a specific example of a survey conducted by Statistics Canada.