Suzuki T, Kudo A
Environ Health Perspect. 1979 Oct;32:131-41. doi: 10.1289/ehp.7932131.
Hayashi's Quantification II is a method of multivariate discrimination analysis to manipulate attribute data as predictor variables. It is very useful in the medical research field for estimation, diagnosis, prognosis, evaluation of epidemiological factors, and other problems based on multiplicity of attribute data. In Japan, this method is so well known that most of the computer program packages include the Hayashi Quantification, but it seems to be yet unfamiliar with the method for researchers outside Japan. In view of this situation, we introduced 19 selected articles of recent applications of the Quantification II in Japanese medical research. In reviewing these papers, special mention is made to clarify how the researchers were satisfied with findings provided by the method. At the same time, some recommendations are made about terminology and program packages. Also a brief discussion of the background of the quantification methods is given with special reference to the Behaviormetric Society of Japan.
林氏量化法II是一种多元判别分析方法，用于将属性数据作为预测变量进行处理。在医学研究领域，基于属性数据的多样性，它在估计、诊断、预后、流行病学因素评估及其他问题上非常有用。在日本，这种方法广为人知，大多数计算机程序包都包含林氏量化法，但对于日本以外的研究人员来说，似乎还不熟悉该方法。鉴于这种情况，我们介绍了19篇近期林氏量化法II在日本医学研究中应用的精选文章。在回顾这些论文时，特别提及以阐明研究人员对该方法提供的结果的满意程度。同时，对术语和程序包提出了一些建议。还特别参考日本行为计量学会对量化方法的背景进行了简要讨论。
