International Usage of Social Survey Data

跨国研究吸引了越来越多的社会科学家。过去二十年来，涉及日本的大型跨国调查急剧增加。同时，跨国比较统计在影响各种问题的舆论和政策决定方面发挥着越来越重要的作用。与越来越多地注意调查统计相比，很少注意这些统计是如何产生的。事实上，虽然统计数据通常被接受，就像它们被报道一样，但它们的有效性、可靠性或真实性的基础并不像人们通常认为的那样坚实。我们发现，日本社会调查中采用的抽样方法通常是稳健的，但在日本受访者中普遍观察到的反应集，以及在美国开发和测试的“标准化”措施的文化有效性，提出了一个巨大的挑战。设计一项跨国调查涉及一个复杂的权衡网络，必须在战略拟订阶段加以权衡。我们强烈认为有必要为跨国调查建立一个理论模型和实际指导方针。必须考虑到一组相互依存的决策、样本的设计、无反应问题、收集系统的组织结构、测量的有效性和可比性以及调查成本。需要这样一种模式来指导全面调查系统的设计，以取代目前导致跨国调查的关键要素不够优化的直观方法。

Cross-national research has attracted a growing number of social scientists. The past two decades have witnessed a dramatic increase in large scale cross-national surveys involving Japan. At the same time, cross-national comparative statistics have come to play an increasingly important role in influencing public opinion and policy decisions on various issues. In contrast to increasing attention drawn to survey statistics, very limited attention has been given to how those statistics were generated. Indeed, while the statistics are usually accepted just as they are reported, the grounds for their validity, reliability or authenticity are not as solid as generally assumed. We found the sampling methods employed in Japanese social surveys are generally robust, but response sets commonly observed among Japanese respondents and the cultural validity of 'standardized' measures which were developed and tested in the United States, present an enormous challenge.Designing a cross-national survey involves a complicated network of trade-offs which must be weighed during the strategy formulation stage. We strongly feel a need for a theoretical model and practical guidelines for cross-national surveys. It is imperative to take into account an interdependent set of decisions, the design of sample, the nonresponse issue, the organizational structure of the collecting system, measurement validity and comparability as well as survey costs. Such a model is needed to guide the design of a total survey system to substitute the present intuitive approach that results in suboptimization of crucial elements of cross-national surveys.