This study, 2010 Census Language Program: Pretesting of Census 2010 Questionnaire in Five Languages (Task Order 10), was designed to pretest the 2010 mailout/mailback self-administered census form that has been translated into Chinese, Korean, Vietnamese, and Russian. Cognitive interviews in English were conducted to determine whether issues found in any of the four target language s present similar problems in English; this testing verifies the presence of the problems. Because the Census Bureau has completed pretesting the content of the Spanish language census form in the context of a bilingual “swimlane” questionnaire, it was not included in this project.

This research project aims to identify and correct translation issues at the lexical (wording), syntactic (grammar and naturalness in target language), and pragmatic (sociocultural context and appropriateness) levels. According to Pan (Pan, 2007; Pan et al., 2006), cognitive testing of a translated instrument is effective in detecting translation issues at the lexical, syntactic, and pragmatic levels. Specifically, cognitive testing goes beyond the lexical and syntactic levels and can resolve more subtle issues that only surface at the pragmatic level. For example, cognitive testing can identify issues related to respondent comprehension and reaction because of cultural expectations and assumptions. Results from cognitive testing can then be used to achieve the intended effect and to improve translation quality so that it meets the Census Bureau Guideline for Translation (U.S. Census Bureau, 2004).