1990 年 6 巻 1 号 p. 1-13
There are many problems to be solved in social surveys, which aim to reveal information about social climates, family relations, life styles and individual social attitudes. The present study endeavors to assess these data vis-a-vis the prevention of senile dementia. While the prospective follow-up study would be best, it is not practical approach in this instance. In the present study, we planned the social survey specifically for the aged, a situation which brought up numerous problems in design and data collection. In this paper, those problems are described, along with some comments regarding improvements for such a survey. Finally, a data-analytic model, designed to elicit useful information from incomplete data, is presented.