The Effects of Questionnaire and Content Changes on Race Data: Results of a Replication of 1990 Race and Origin Questions
Elizabeth Martin, Manuel de la Puente, Claudette Bennett
KEY WORDS: split-panel experiment, race reporting, Census 2000, questionnaire design, mail survey, Hispanic origin, question order effects
ABSTRACT
In 1997, the Office of Management and Budget revised methods for collecting and reporting data on race and ethnicity in government surveys and censuses. The most publicized change was to allow respondents to report more than one race. Census 2000 introduced additional changes in question sequence, format, wording, and categories. To evaluate the effects of the changes on race and Hispanic origin data, 1990 questions on race and Hispanic origin were replicated in an experiment conducted during Census 2000. An experimental panel of 10,000 households was mailed a 1990-style short form. A control panel of 5,000 households received the Census 2000 mail questionnaire. This paper compares data from 1990-style and Census 2000 mail questionnaires to evaluate data quality differences. Results show that Census 2000 questionnaire changes substantially improved the completeness of race and Hispanic origin reporting. Race reporting was also affected by the questionnaire changes, with significantly more reports of two or more races, fewer reports of Native Hawaiian and Other Pacific Islander race, and fewer reports of Some Other Race in the Census 2000 questionnaire. Race reporting by Hispanics was particularly affected by the questionnaire changes.
CITATION: "The Effects of Questionnaire and Content Changes to Race and Hispanic Origin Items: Results of a Replication of the 1990 Census Short Form in Census 2000," Proceedings of the Survey Research Methods Section (American Statistical Association). 2001.
