Disparate Modes and their Effect on Instrument Design
Mark Pierzchala ; Mathematica Policy Research, USA
Multimode surveys are increasingly common for a variety of reasons including attempts to increase response rates,
lower costs, and accelerate data collection. As practitioners of multimode surveys, we concern ourselves with a
variety of issues, but few are as important as the rendering of the electronic (e.g., Blaise) instrument. One concern
is that in a multimode survey we may find ourselves with different question forms, for the same item, across the
modes.
This paper defines and elaborates on the concept of disparate modes and their impact on instrument
design strategies. Two modes of data collection are defined as disparate if they differ in any one of three intrinsic
aspects of data collection mode: presentation (aural vs visual), administration (self vs interviewer), and medium
(dynamic vs passive). MPR has found that the greater the disparity between the modes, the harder it is to maintain
consistency in instrument design between modes.
| Contact: markpierzchala@cs.com |