U.S. flag

An official website of the United States government

Skip Header


Testing Alternative Relationship and Marital Status Questions on the 2013 American Community Survey-Questionnaire Design Test

Written by:

Executive Summary

An important shift in American family life is greater recognition of same - sex couples. Measurement of relationship and marital status must keep pace with this change to ensure the relevance, accuracy, and quality of our data on same- and opposite-sex married and unmarried couples. In this report, I review quantitative testing of revised relationship and marital status items in the 2013 American Community Survey - Questionnaire Design Test (ACS - QDT). Specifically, I compare results from a control questionnaire containing the relationship and marital status items currently used in production to those from an experimental questionnaire containing revised items. I find that unit response rates do not vary according to the version of the relationship and marital status items used. Further, I find no evidence that respondents receiving the revised relationship and marital status items are less likely to respond to these items in particular. Indeed, nonresponse on the marital status item is slightly lower in the test panel compared to the control panel. Also important, distributions on the relationship and marital status items do not differ by panel. Finally, consistency of couple's reports of relationship and marital status does not differ between the control and test panels. A weakness is that, because same-sex married and unmarried couples comprise a very small percentage of households, the ACS-QDT contains a small number of these couples. Although I find data quality to be comparable when using the current versus revised relationship and marital status items, I am uncertain whether data quality is truly comparable or the limited sample size inhibited the detection of differences in quality. For this reason, additional testing of the revised relationship and marital status questions using still larger samples is needed.

Page Last Revised - October 8, 2021
Is this page helpful?
Thumbs Up Image Yes Thumbs Down Image No
NO THANKS
255 characters maximum 255 characters maximum reached
Thank you for your feedback.
Comments or suggestions?

Top

Back to Header