The purpose of the 2010 Census Alternative Questionnaire Experiment: Census 2000 Form Replication Panel  was to determine the combined data effects of all questionnaire changes made to the 2010 Census mail questionnaire.

Since 1970, the U.S. Census Bureau has implemented an experimental program to evaluate a variety of alternative methodologies and questionnaire design strategies. Traditionally, an experiment has been conducted during the decennial census in order to evaluate the cumulative effects of all content changes to the form from the previous census. As was done in Census 2000, the 2010 Census Alternative Questionnaire Experiment compared questionnaire content from two censuses. Specifically, we compared the results from a control panel using the 2010 Census questionnaire to those obtained from a questionnaire that replicated the Census 2000 questionnaire wording, categories, order, and other essential design features such as the use of icons in the column headers. Numerous changes were made to the 2010 Census questionnaire, compared to the short form used in Census 2000. In addition to changes in the overall questionnaire format and appearance, almost every census data item underwent at least some change in terms of response categories, instructions, and/or question wording. Although most changes were critically tested during the mid-decade to evaluate their impact, some minor changes were implemented based on subject matter expertise. By comparing the Census 2000-style questionnaire with the 2010 Census questionnaire in the same time frame, we can eliminate the impact of real changes to the population and can more clearly assess the combined effects of the questionnaire design changes.

The specific research questions for this experiment are:

1. How do the changes made to the 2010 Census questionnaire affect overall return rates, form completeness, item nonresponse, and specific item distributions?
2. Will the addition of the undercount and overcount questions on the 2010 Census questionnaire and the changes to the residency rules description result in a significant difference in household population count when comparing the 2000 Content panel to the 2010 Census form?
3. What effect will removing the “Foster child” response option from the relationship question have on the distribution of the other response categories, specifically “Other relative” and “Other nonrelative”?