经济社会调查中的空间平衡抽样设计

Spatial Balanced Sampling Design in Economic and Social Surveys

在经济社会调查中,总体单元之间的空间相关性普遍存在,对传统抽样设计提出了挑战。针对这一问题,提出了使用经纬度坐标作为空间辅助信息,借助空间平衡抽样算法获取样本的设计思路。该种算法利用总体单元之间的空间距离设计抽样算法更新包含概率,使空间上距离较近的单元倾向于不同时进入样本,从而使样本单元在空间上均匀覆盖。实证研究结果表明,随着样本量连续增加,空间平衡抽样设计的估计量标准差在合理的抽样比范围内总是优于传统抽样设计,能够显著提高估计效率。

Spatial dependency in economic and social surveys brings new challenge for the traditional sampling methods.Spatial balanced sampling design based on the spatial auxiliary information,for instance,the longitude and latitude information,is raised to solve this problem.It is shown by the case study that,as the sample size continuously changed,spatial balanced sampling design could always gain more efficiency than traditional sampling designs.