Biostatistics : Development of Bacic Theory and Ist Application

[毒理学数据分析]。发表了“毒理学数据的统计分析”（Yoshimura，Ohashi）。这是第一本用日语写的书。还出版了“毒理学研究中的生物统计学”（Sakuma，Yanagawa，Yoshimura编辑）。这是第一届国际毒理学生物统计学会议的会议记录。提出了一种确定无明显不良效应水平的新方法。在啮齿动物微核试验的情况下，还探索了致突变试验数据的统计分析。[流行病学方法]。“人类癌症的生物统计学”（Yanagawa，Ohashi，Tango，Yanagimoto编辑）这是第四届日美人类癌症研究生物统计学会议的会议记录，也出版了“医学科学统计学”（Tango）。这两篇出版物对医学领域的研究人员产生了巨大的影响。推广了Mantel-Haenszel估计 ...更多信息 对于2xJ表（J>2）。提出了一种疾病空间聚集的统计检验方法。从因果关系的角度对互动和规范的含义进行了深入的研究和澄清。当存在效应修正时，给出了综合风险比的估计量。[临床数据的统计方法]。一套新的统计方法的基础上累积卡方开发。这些方法有望取代经典的方差分析。统计检验程序，调整背景因素的不平衡，并检测等效性或超过等效性的两种药物的开发。探讨了Meta分析的方法，使研究者能够处理多中心临床试验。【基础理论】。本文提出了一种新的生长曲线分析方法，即在生长曲线中引入参数间的线性模型，推导了该方法的算法，并编制了可在微机上使用的计算机程序。正如所看到的，例如，在邀请特别讲座“良好的统计估计，从皮塔哥的观点”发表在日本数学学会，基本理论的生物统计学也非常先进。少

[Analysis of Toxicological Data]. "Statistical Analysis of Toxicological Data" is published (Yoshimura, Ohashi) . This is the first book of this kind written in Japanese. Also published is the "Biostatistics in the Study of Toxicology" (Sakuma, Yanagawa, Yoshimura eds.) which is the proceedings of the first International Biostatistics Conference in the Toxicology. A new method for the determination the of the no-observed-adverse-effect level is developed. Statistical analysis of data in mutageneity assays in the case of rodent micronucleus assay is also explored.[Methodologies in Epidemiology]. "Biostatistics in Human Cancer" (Yanagawa, Ohashi, Tango, Yanagimoto eds.) which is the proceedings of the fourth Japan-US Biostatistics Conference in the Study of Human Cancer is published : also published is the "Statistics for medical Science" (Tango). These two publications have given an inmense effect to researchers in the field of medical science. Mantel-Haenszel estimators are generalized … More for 2xJ table (J>2) . A statistical test for spaciai aggrigation of disease is developed. Meanings of interaction and standardization are syudied in depth and clarified from the point of view of causality. Estimators of summary hazard ratio are developed when there exist effect modification.[Statistical Methods for Clinical Data]. A set of new statistical methods are developed based on accumulated chi-squares. These methods are anticipated to replace the classical analysis of variance. Statistical testing procedures which adjust for the inbalance of background factors and to detect the equivalence or more than equivalence of two drugs is developed. A method is explored on the Meta Analysis so that it enables researchers to treat multicenter clinical trials.[Basic thory]. A new analysis of growth curves is explored by introducing linear models among parameters in growth-curve, algorithm of the method is developed and computer program is explored which can be used on parsonal computer. As is seen, for example, in the invited special lecture "Goodness of statistical estimator, from Pitagorian point of view" delivered at the Mathematical Society of Japan, basic theory of biostatistics is also very much advanced. Less