In this study, an attempt was made to delineate the structure of environmental factors thatinfluence national health and ascertain the difference of these influences, from the view point of a quantative approach, between two countries that have different cultures and ethnic compositions. As a consequence of past studies, i. e., those reporting how environmental factores influenced national health in Japan. Using the same paradigm, U. S. data was analyzed, i. e., the structure of environmental factors and how they influenced national health. The distribution of the data provided some reasons for the differences between Japan and the U. S..
In this article the author discusses some of the major methodological problems in quantitative political research, centering around W. Weaver's distinction between simplicity, disorganized complexity and organized complexity. Then the author examines two kinds of empirical studies that he has undertaken, political economy and international relations, in light of the three kinds of methodological difficulties confronting scientifically oriented political scientists.
The prognoses of adult patients with nephrotic syndrome were analyzed by two types of Markov model. One is the 3 state Markov model which consists of complete remission, incomplete remission and death states and the other one is the 4 state one in which the complete remission state in the 3 state model is devided into 2 states, never-relapse state and one having a possibility of relapse. The results were as follows; 1)the 3 state model estimated extremely poor or somewhat different prognosis comparing with the observed one. 2)The estimated prognosis by the 4 state model showed a good agreement with the actual one and the clinical empirical knowledge. Furthermore, the transition matrix of this model had acceptable stability. These suggested the validity of the prognostic estimation of adult nephrotic syndrome by the 4 state Markov model.
A method, called “isometric line mapping(ILMAP)”, is described for plotting contours of h which is a function of χ and y. An h value at each grid point on the χ-y plane is estimated according to the weights based on the distances to all measured points. Furthermore, the h values are smoothed by means of the moving average method. Two applications to psychological data(Ichikawa, 1981; Carroll, 1972)are presented to show that ILMAP is useful for visually grasping the effects of χ and y upon h.
To assess the contribution of the partial knowledge of examinees to the scores of multiple-choice tests, we have proposed two mathematical models, the Binary-model and the Terminal model. These models both were based on the probability with which an examinee recognizes a choice correctly, and on the arrangement of choices. As an application of these models, the distributions of scores were given for the one-best-response type and the two-responses type of questions of the national board examinations for the medical and dental practices. The distributions of scores by these two models were compared.