We conducted this research to evaluate the cognitive and motivational properties of two mailing packages proposed for inclusion in the 1997 National Survey. One was designated the “accordion” and the other the “rollfold”. The cognitive properties included how well respondents navigated through the forms, operated the folds, and comprehended the questions. We also evaluated the motivational properties of the design features that differed between the mailing packages. This included respondents' preferences for one design feature over another and their reports of whether or not any of the design features would motivate them to complete the census. During this part of the interview, we occasionally included a comparison to the best performing mailing package from the 1996 National Content Survey, (the “green booklet”). We also asked respondents to comment on individual pictures and messages from the accordion form. Finally, when evaluating the cover pages, we included a third, stand-alone cover page.

These mailing packages performed better than those previously tested. We attribute this to two major changes. The first is that these mailing packages look more official than some of the packages tested last time. As a result, respondents didn't perceive these mailing packages as junk mail. The second change was with the navigational quality of the form. These forms were designed with a well-defined vertical flow, beginning with the 'Start Here' instruction and continuing until the very end. One of the greatest achievements with these forms was that all but two respondents began in the right place, and all but one successfully navigated through the form without missing or skipping questions. In addition, the critical concept of the census was conveyed. As a result, not one respondent mistakenly repeated information about himself or herself.