Summary: | Obtaining reliable and valid data on sensitive questions represents a longstanding challenge for public health, particularly HIV research. To overcome the challenge, we assessed a construal level theory (CLT)-based novel method. The method was previously established and pilot-tested using the Brief Sexual Openness Scale (BSOS). This scale consists of five items assessing attitudes toward premarital sex, multiple sexual partners, homosexuality, extramarital sex, and commercial sex, all rated on a standard 5-point Likert scale. In addition to self-assessment, the participants were asked to assess rural residents, urban residents, and foreigners. The self-assessment plus the assessment of the three other groups were all used as subconstructs of one latent construct: sexual openness. The method was validated with data from 1,132 rural-to-urban migrants (mean age = 32.5, SD = 7.9; 49.6% female) recruited in China. Consistent with CLT, the Cronbach alpha of the BSOS as a conventional tool increased with social distance, from .81 for self-assessment to .97 for assessing foreigners. In addition to a satisfactory fit of the data to a one-factor model (CFI = .94, TLI = .93, RMSEA = .08), a common factor was separated from the four perspective factors (i.e., migrants' self-perspective and their perspectives of rural residents, urban residents and foreigners) through a trifactor modeling analysis (CFI = .95, TLI = .94, RMSEA = .08). Relative to its conventional form, CTL-based BSOS was more reliable (alpha: .96 vs .81) and valid in predicting sexual desire, frequency of dating, age of first sex, multiple sexual partners and STD history. This novel technique can be used to assess sexual openness, and possibly other sensitive questions among Chinese domestic migrants.
|