Analysis of survery data

调查数据分析

• 批准号: 8146-2009
• 负责人: Thompson, Mary
• 金额: $ 2.19万
• 依托单位: University of Waterloo
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2015
• 资助国家: 加拿大
• 起止时间: 2015-01-01 至 2016-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=568561
• 关键词: Analysis; survery; data

The research proposal concerns the statistical modelling of association and dependence among observations in time or space, with particular emphasis on survey data and survey methodology. I am proposing work on three aspects of the analysis of complex survey data. The first topic is that of density estimation for the joint distribution of event times and for spatial analysis; the former is relevant when we are interested in assessing whether one type of event might trigger another, and the latter will be applicable to the visualization of spatial distributions of disease incidence. The second topic is estimation in multilevel models, wherein explanatory variables at more than one level of aggregation are used to model responses; since this kind of modeling is increasingly widely used, resampling methods which allow for the complex sampling design in estimating components of variance, and which can be made available to analysts, will be explored. Finally, I will study applications of hierarchical models to account for population heterogeneity in the analysis of survey data, with particular application to cross-cultural survey methods. This topic is of great interest to social scientists and health scientists in the context of international surveys.

研究提案涉及在时间或空间上对观察之间的联系和相关性进行统计建模，特别强调调查数据和调查方法。我建议在分析复杂调查数据的三个方面开展工作。第一个主题是事件时间联合分布的密度估计和空间分析；当我们有兴趣评估一种类型的事件是否可能引发另一种事件时，前者是相关的，而后者将适用于疾病发病率的空间分布的可视化。第二个专题是多水平模型中的估计，其中在一个以上的聚合水平上使用解释变量来对反应进行建模；由于这种建模越来越广泛地使用，因此将探索允许在估计方差分量时进行复杂抽样设计的重抽样方法，并且可供分析人员使用。最后，我将研究层次模型在调查数据分析中解释人口异质性的应用，特别是跨文化调查方法的应用。在国际调查的背景下，这一主题引起了社会科学家和卫生科学家的极大兴趣。