Journal of Official Statistics, Vol.19, No.4, 2003. pp. 403419

Current Issue
Personal Reference Library (PRL)
Personal Page

Improving Navigational Performance in U.S. Census 2000 by Altering the Visually Administered Languages of Branching Instructions

This article reports results from an experiment in which the branching instructions in the long form of the 2000 U.S. Decennial Census were varied in a national experiment administered to 25,000 households. The experimental manipulations included testing a simple change in the verbal language of the current design from ``Skip to' to ``Go to.' Three additional combined manipulations were also tested in which the graphic (e.g., larger bolder font), symbolic (e.g., arrows), and verbal (e.g., additional statements) were altered. Whereas the simple change in the verbal language only (from skip to go) had no effect on branching error rates, the combined manipulations of the visually administered languages tended to reduce the branching errors. Results of these experiments provide evidence that errors in following branching instructions can be reduced for general populations through the careful manipulation of the visually administered languages.

Questionnaire design; errors of commission; errors of omission.

Copyright Statistics Sweden 1996-2018.  Open Access
ISSN 0282-423X
Created and Maintained by OKS Group