The estimation problem of the population quantity of the red sea bream(Pagrus major)in the Seto inland sea is dealt, and one method of assesing the population based on the tagging experiments is proposed. The cohort analysis, which has been the popular method of assesing the population, has the many difficulties to execute and also take a long sampling time. The proposed method is thought to be better than the cohort analysis in these points.
Policy-making and -execution by representatives make a great impact on social system. If inappropriate representatives are elected and unsuitable policies are executecd, unreasonable social impact and enormous social cost will be produced. Scientific knowledge on the way of electing representatives are now desired. Most mayors seems to be re-elected if they want. Some mayors, however, fail and new mayors win in some cities. This paper aims to analyse relation between change of mayors and voters' preference structure. A conceptual model of mayor's election is built so as to analyse the effect of voters' preference structure on the state of social system. Computer simulation shows the results:a mayor elected is changed more often and the state of social system becomes more unstable as the preference structure of voters changes from Goal-Realization-Oriented to Policy-Attitude-Oriented. Hotelling's Theorem(Median Voter's Theorem)is shown as a special case by the simulation.
In this paper, first, similarity between categories is formulated using AIC. Then the relation between dual scaling solution and category pooling process conducted by AIC is discussed . Second, a method analyzing cross tabulated data is proposed, which is useful for handling multicategorized variables. In this method, categories are pooled into some groups so as to minimize AIC of cross table composed by the pooled category groups. Then, some tables that not significantly different from the MAICE table are selected, and several inductive results are inferred and confirmed by comparing dual scaling solutions on the selected tables and by analyzing the category pooling process. This confirmation procedure also provides deeper understanding of dual scaling solutio Applying the method to two cross tables of vehicle possession in Tukuba Science City, it is shown that proposed method is practically efficient for drawing conclusions, even in the cases several categories have rather small samples and the table apparently has no significance.
Sinece1978, the candidates for the Kitasato University, Medical School have been requested to answer the translated edition of MMPI. For these5years, two kinds of Japanese editions were adopted for test alternately, and total of3894male and786female candidates took this test. Using this material, the authors have studied the influence of translation into Japanese on the response for each MMPI item and profile pattern. The results were as follows: (1)Excluding D and Si scales, the mean scores of the remaining 12scales of either Doshisha or Nihon-edition were located within a difference of one standard deviation from the mean values of the norm group. (2)By means of item by item comparison, considerably large difference between the two editions was observed in the percentage of cases giving?gtrue?hresponse. This fact suggested that difference between the editions had to be taken into consideration when individual items were used separately or scales were newly constructed by using items with high risk of different response pattern.
A method of simultaneous equating for scales with the Rasch measurement model was proposed. In this method, two difficulty estimates from each of all common items were obtained and the difference between the two was composed of linking constant and sampling error. According to the weighted-least-squares principle, linking constants are estimated so as to minimize the sum of squares of sampling errors. To illustrate the proposed method, equating was carried out for a set of simulated tests and real verbal tests. As a result of using the new method, better statistical estimates were found when compared to the method used by Engelhard and Osberg(1983).
In Part I and II, comparative studies for the analysis of association in cross-classifications having ordered categories were developed. In this manuscript, GSK approach is described to fit linear models for ordinal data arising from a variety of situations. Aspects of applications are illustrated through several examples for which results of other methods are also provided