Testing for Differential Responses in a Multiple Category Scale: a Case Study on Self-Rated Health among Foreign- and Native-Born Asian Americans

• Activity Number: 467
• Type: Contributed
• Date/Time: Wednesday, August 9, 2006 : 2:00 PM to 3:50 PM
• Sponsor: Social Statistics Section
• Author(s): Elena Erosheva*+ and Emily C. Walton and David T. Takeuchi
• Companies: University of Washington and University of Washington and University of Washington
• Keywords: propensity score matching ; symmetry models ; categorical data ; self-rated health

Abstract

Using data on Asian Americans from the National Latino and Asian American Study, we test whether immigrants are less likely to report the extreme ends of the 5-category self-rated health scale than their native-born counterparts. We use propensity score matching to derive groups who share similar demographic and health characteristics. Each native-born person is matched to a foreign-born of the same ethnicity by nearest available Mahalanobis metric within a caliper defined by the propensity score. Propensity score framework allows us to make descriptive comparisons of self-rated health responses by nativity status, controlling for background characteristics. We find that nativity is not associated with higher likelihood of reporting the extreme ends of the health status scale. In addition, we find no evidence of imbalances in endorsement of any particular category between the two groups.