The factorial survey as a method for measuring attitudes in population surveys

析因调查是一种将调查研究的优势与实验设计的优势相结合的方法。受访者评价的是假想的对象或情景描述，而不是单一的项目。在过去的25年里，我们观察到这种方法在社会科学中的使用不断增加。然而，对于因子调查质量的实际实施和系统评估，仍然缺乏经验支持的标准。该项目的主要目标是测试各种方法效果的相关性，并评估用于分析数据的不同统计模型，同时考虑到其特定的多层次结构和抽样程序。通过在项目第一阶段收集四个不同的数据集，我们系统地进行了一些调查实验。该项目现在的目标是全面分析收集到的数据，开展一项关于呈现方式的补充实验，并开发一个将普遍适用于调查研究的软件模块，用于进行因子调查。调查析因调查设计在一般人口调查中的适用性，并推断调查研究的一般建议，具有极其重要的意义。

The factorial survey is a method that links the advantages of survey research with those of experimental design. Respondents evaluate hypothetical object or situation descriptions rather than single items. Over the past 25 years we have observed a continually increasing use of this method in the social sciences. Nevertheless, empirically supported criteria for the practical implementation and systematic evaluation of the quality of factorial surveys are still lacking. The main objectives of the project are to test the relevance of various method effects and to evaluate different statistical models for analyzing data taking their specific multilevel structure und sampling procedure into account. By gathering four different data sets in the first phase of the project we systematically conducted a number of survey experiments. The project now aims to comprehensively analyze the data gathered, to carry out one supplementary experiment on presentation modes, and to develop a software module that will be generally applicable in survey research for conducting factorial surveys. It is of vast importance to investigate the suitability of factorial survey design in general population surveys and infer general recommendations for survey research.