de Lillo A, Meraviglia C
Department of Sociology, University of Milan, Italy.
Subst Use Misuse. 1998 Feb;33(3):751-64. doi: 10.3109/10826089809115894.
The paper focuses on the role of the spouse's occupation as a resource for mobile individuals, from the perspective that social positions are held by families, rather than by individuals. Three groups are confronted in terms of the role of the key variables and other relevant factors: men whose spouse does not have a paid job (group 1), men and women whose spouse has a paid job (group 2 and 3). The data set is provided by the national survey on social mobility in Italy, carried out in 1985; social achievements of members of the three groups are considered, including social origins and destinations, social position corresponding to respondent's first job, cultural background (educational achievement of respondent's father and mother found), respondent's education and spouse's social position. The techniques used are discriminant analysis and back propagation Neural Networks. Both techniques traced a clear boundary between group 1 and groups 2 and 3, which were discriminated mainly on the basis of the spouse's occupation; Artificial Neural Networks reached better classification results and allowed a deeper insight into the nonlinear effects of the discriminating variables for the three groups.
本文从社会地位由家庭而非个人占据的角度出发，聚焦于配偶职业作为流动个体资源的作用。就关键变量和其他相关因素的作用而言，有三组人群被加以对比：配偶没有带薪工作的男性（第一组）、配偶有带薪工作的男性和女性（第二组和第三组）。数据集由1985年开展的意大利全国社会流动调查提供；对三组人群成员的社会成就进行了考量，包括社会出身和去向、与受访者第一份工作相对应的社会地位、文化背景（受访者父母的教育成就）、受访者的教育程度以及配偶的社会地位。所使用的技术是判别分析和反向传播神经网络。两种技术都在第一组与第二组和第三组之间划出了一条清晰的界限，这两组主要是根据配偶的职业来区分的；人工神经网络得出了更好的分类结果，并能更深入地洞察区分变量对三组人群的非线性影响。
