Using Item Response Theory to Examine the Self-Construal Scale With Asian American College Students in the Contiguous United States

Singelis's (1994) measure of independent/interdependent self-construals is frequently administered to Asian Americans in the contiguous United States despite the lack of validity and reliability evidence available pertaining to this population. The purpose of this study was to investigate the psychometric properties of Singelis's (1994) measure with a sample of Asian Americans college students in the contiguous United States. Item response theory was used to provide item-level information (reliabilities) and to examine differential item functioning among diverse subgroups of Asian Americans. A total of 281 Asian Americans from an undergraduate psychology subject pool of a West Coast university completed the Self-Construal Scale. Two, separate polytomous item response theory models were run reflecting each subscale. Each model demonstrated adequate fit. Test information curves, standard errors of the estimate, and item parameters suggested that each measure effectively captured the range of their respective constructs. Five items on the Independent scale and two items on the Interdependent measure were flagged for differential item functioning based on gender (women vs. men), generational status (first generation vs. second generation and above), and ethnicity (Southeast vs. East Asian) subgroups. Our results suggested that, overall, the items of both subscales function well among Asian Americans college students in the contiguous United States. However, future researchers should be mindful of the differential functioning of certain items among specific subgroups. These items may warrant future revision to better accommodate for the diversity within the Asian American population.