From Respondent Debriefings to Pilot Test and Beyond: A Comprehensive Redesign of a Questionnaire Measuring Foreign Direct Investment

In 2004, survey methodologists at the U.S. Census Bureau began a project with the U.S. Bureau of Economic Analysis (BEA) to aid in the evaluation and redesign of two of their self-administered survey questionnaires. These surveys, one annual and the other quarterly, collect foreign direct investment (FDI) data from U.S. companies that are foreign-owned. Although BEA goes to great lengths to assure that its surveys are not unduly burdensome, the forms may impose significant burden for some respondents, and data verification may be labor-intensive for BEA analysts. Some respondents expressed difficulties with comprehending the FDI concept and other particulars of the surveys, which may differ subtly but significantly from the ways they think about their companies. Our redesign strategies were aimed at facilitating respondents' comprehension of the FDI reporting requirements, as well as reducing the overall burden of completing the forms. To maximize the effectiveness of our redesign, we undertook a comprehensive process using multiple research methods to gain a thorough understanding of the conceptual underpinnings of the surveys and the problems respondents encountered when completing them. The project was concluded when the redesigned questionnaire for the quarterly BE-605 was fielded in March 2007. This paper summarizes the various methods used throughout the project, and presents key findings from various qualitative and quantitative analyses. The focus will be on the BE-605 but will include relevant findings from the annual form (BE-15), because the two forms share the same conceptual bases and because the redesign strategies were developed to address similar reporting issues for each.