Stones M J, Kozma A
Exp Aging Res. 1986 Spring;12(1):23-9. doi: 10.1080/03610738608259431.
The intention of the study was to test between two causal models of relationships between happiness and five of its correlates (i.e., housing satisfaction, financial satisfaction, perceived health, locus of control, and activity level). In Model 1, happiness was identified as a propensity variable that is a source of influence on the lower order correlates. Model 2 is a traditional formulation whereby levels of happiness are determined by its correlates. The models were tested by causal modeling procedures, using 18-month longitudinal data from 408 elderly subjects. The findings support Model 1 over Model 2 for 17/18 predictions that differentiated the models. The status of happiness as a trait was discussed with reference to: the type of items included in the happiness scale; and relationships with variables not included in the research.
该研究的目的是检验幸福与其五个相关因素（即住房满意度、财务满意度、感知健康、控制点和活动水平）之间关系的两种因果模型。在模型1中，幸福被确定为一个倾向变量，是对较低层次相关因素产生影响的源头。模型2是一种传统的表述方式，即幸福水平由其相关因素决定。使用来自408名老年受试者的18个月纵向数据，通过因果建模程序对这些模型进行了检验。对于区分这两个模型的17/18个预测，研究结果支持模型1而非模型2。参照幸福量表中所包含项目的类型以及与研究中未纳入变量的关系，对幸福作为一种特质的地位进行了讨论。
