Abstract: Subjective happiness, health and life satisfaction are usually measured by selfreported ratings. However, studies have shown that, in this type of measurement of subjective indices, there is a problem of “selfrating heterogeneity”, that is, the same score given by different respondents may very well represent quite different states in reality. This has been extensively evidenced in health measures. Considering the fact of family being the basic unit in the traditional Chinese society, measuring family happiness is, therefore, of unique significance. Taking the measurement of family happiness as an example, this study used the survey data from a representative national sample to examine if there was a problem of “selfrating heterogeneity” and if so, how to solve that problem. The findings revealed clear existence of this problem, including significant differences in the family happiness standards used by the people in urban vs. rural areas, and between people of different ages. To address the problem of “selfrating heterogeneity,” this study employed measurement anchors and used the “equal ratio method” to standardize the family happiness scores. Then the factors that influenced the “selfrating heterogeneity” were analyzed and the standardized scores were compared with the raw scores to determine the difference. Finally, the hierarchical linear model was applied to the examination of the impact of community integration on residents’ family happiness and the standardized and raw scores were compared as well.

Key words: family happiness , subjective measurement , measurement anchor , standardization