Markides K S, Martin H W
J Gerontol. 1979 Jan;34(1):86-93. doi: 10.1093/geronj/34.1.86.
A path analysis model of life satisfaction is proposed and applied to data from interviews with 141 persons aged 60 years and over. Predictor variables examined are self-reported health, income, education and an activity index. The analysis was conducted separately for the two sexes. Health and activityemerge as strong predictors of life satisfaction while income influences life satisfaction indirectly via activity. The importance of investigating direct and indirect effects of variables via path analysis is discussed.
提出了一个生活满意度的路径分析模型，并将其应用于对141名60岁及以上人群的访谈数据。所考察的预测变量包括自我报告的健康状况、收入、教育程度和活动指数。分析分别针对男性和女性进行。健康和活动是生活满意度的有力预测因素，而收入则通过活动间接影响生活满意度。讨论了通过路径分析研究变量直接和间接影响的重要性。
