Toseland R, Rasch J
Int J Aging Hum Dev. 1979;10(2):203-11. doi: 10.2190/n4ha-lqy4-rr3u-j800.
The Automatic Interaction Detector (AID3) was used to develop a model based on the interaction of predictors of life satisfaction. The sample consisted of 871 people over fifty-five years of age. Thirty-one potential predictors were used representing demographics, environmental variables, and social psychological variables. The findings indicate that nine predictors explained 22.1% of the variance in life satisfaction scores. The most important predictors of life satisfaction were family life satisfaction, personal health satisfaction, and satisfaction with dwelling. The interactions between the predictors indicated that a simple linear-monotonic relationship between the predictors was too restrictive.
自动交互检测器（AID3）被用于基于生活满意度预测因素的交互作用来开发一个模型。样本由871名55岁以上的人组成。使用了31个潜在的预测因素，代表人口统计学、环境变量和社会心理变量。研究结果表明，9个预测因素解释了生活满意度得分中22.1%的方差。生活满意度最重要的预测因素是家庭生活满意度、个人健康满意度和对居住环境的满意度。预测因素之间的交互作用表明，预测因素之间简单的线性单调关系过于受限。
