竞争性行为的性别差异——基于分解方法的研究  

Gender Difference in Competitive Behavior:Based on Quantile Regression Decomposition Method

作　　者：王疏影[1] 梁捷[2] 黄有光[3] 

机构地区：[1]西安交通大学金禾经济研究中心,陕西西安710049 [2]莫纳什大学经济系 [3]南洋理工大学经济系,新加坡637332

出　　处：《学术界》2014年第8期52-65,306,共14页Academics

摘　　要：长期以来,女性在竞争性行为表现上与男性存在显著且稳定的差异,很多领域都有"阴盛阳衰"现象。但真正可以公平观察两性竞争行为的平台极少,教育领域则被认为是一个最好的观察领域。尝试以中学生学习表现的差异作为研究竞争性行为性别差异的切入点,分别从个人因素,家庭因素,学校因素以及社会网络这四个维度研究竞争行为的性别差异。首先运用分位数回归技术进行统计建模与分析,再采用Neumark分解方法和基于分位数回归的分解方法对竞争性行为性别差异进行分解。研究结果发现个人因素、家庭因素和学校因素的系数差异是学习表现性别差异的主要来源,而对于网络因素维度,则主要由其特征差异决定学习表现的性别差异。Gender differences in competitive behavior prevail in many fields,including the education sector. However,there is no consistent finding on the gender differences in the study of competitive behavior. T his paper attempts to use the academic performance of middle school students to study gender differences in competitive behavior of four groups of factors( personal factors,family environment,school environment,as well as their social networks measured by an internal social network in each class.). W e apply quantile regression,as well as Neumark decomposition and quantile regression decomposition technique to analysis gender difference in competitive behavior. T he results show that the differences in the coefficients of personal factor,family environment and school environment are the main source of gender differences in competitive behavior in school. However,for the social network aspect,gender differences in competitive behavior in school are affected by the differences in characteristics primarily.

关 键 词：性别差异 竞争性行为 社会网络 分位数回归 分位数分解 

分 类 号：B848[哲学宗教—心理学]