U.S. Census Bureau

Cognitive Interview Results and Recommendations for the 1996 National Content Test

Nancy Bates, Karen Bogen, Theresa DeMaio, Eleanor Gerber, Jennifer Hess, Cleo Jenkins, Elizabeth Martin, Jeffrey Moore, Jennifer Rothgeb, and Dawn Von Thurn

KEY WORDS: cognitive testing; think-aloud technique; census long form; skip instructions; comprehension problems

ABSTRACT

Cognitive interviews (n = 10) were conducted to determine participants’ comprehension and navigation during completion of particular questions from the Census long form for the 1996 National Content Test. Results indicated the following global issues: 1) respondents had a very difficult time following skip instructions, especially when embedded in lots of other text; 2) respondents were sometimes confused about which person for whom to answer; 3) respondents did not skip unnumbered items; and 4) respondents exhibited frustration when answering the same question multiple times. There were few comprehension problems with the roster and coverage questions, but some comprehension and definitional problems existed for the education questions. Several respondents switched the focus from one household member to another during the disability questions, and there were some issues with comprehension of these questions. Very few issues existed for the labor force questions. The non-cash income questions produced multiple comprehension problems. Recommendations for question wording and questionnaire formatting are included.

CITATION:

Source: U.S. Census Bureau, Statistical Research Division

Created: May 2, 2007
Last revised: May 2, 2007