The U.S. Census Bureau will conduct the 2013 American Community Survey (ACS) Questionnaire Design Test (QDT) to test four alternate questionnaire designs for the ACS mail questionnaire. In an interdivisional project with the American Community Survey Office and the Decennial Statistical Studies Division, the Center for Survey Measurement conducted cognitive pretesting of navigation issues for three of the five QDT questionnaires. Pretesting targeted older respondents with low education levels and low internet experience, as these characteristics were found to be prevalent among respondents who choose the ACS mail questionnaire option over the ACS internet questionnaire option.

General findings from a group of 21 respondents include: (1) most respondents had neutral reactions when completing the forms and thought the questionnaires were easy to complete; (2) some respondents did not follow skip instructions properly as a result of not paying careful attention; (3) respondents had no problems with marking check boxes or entering responses in write-in fields; and (4) some respondents did not roster themselves on the questionnaire because they mistakenly thought the information requested on the cover page of the form was the same as the information needed on the house hold roster section that followed the cover page. In terms of form-specific findings, the longest of the three forms had the most comments about appearing too long, and the form designed for Optical Character Recognition had relatively fewer skip instruction mistakes made by respondents in comparison to the other two forms. These results will be used in conjunction with quantitative results from the QDT field test to inform future ACS designs.