Study on the optimum test for the drug trials in crop diseases

• Main Authors: Suh-pyng Liu, 劉素萍
• Other Authors: Ming-lai Shen
• Format: Others
• Language: zh-TW
• Published: 1999
• Online Access: http://ndltd.ncl.edu.tw/handle/09571466121984989177

碩士 === 國立臺灣大學 === 農藝學研究所 === 87 === We can know the effects of the drugs in crop diseases by using optimum test. The study proposes biological assay that is one of statistical experiments for determining the relative potency of several drugs. And the conclusion can be a basis of the choice of drugs in crop diseases. Generally, we adopt indirect assay when using biological assay. So the study focus on the discussion of indirect assay. As we know, indirect assay has two important parts. One situation is indirect assay based upon quantitative responses, there are two different models for this situation: parallel-line assay and slope-ratio assay. When the regression lines are parallel, we adopt parallel-line assay, otherwise, we adopt slope ratio assay. The other situation is indirect assay based upon quantal responses. Probit transformation and logistic transformation are used in the iterative estimation of parameters ａ and ｂ in linear regression models here. We will discuss both of them also. In consequence of analysis of real data, we expect that the scales of the levels of doses are similar. And a suitable transformation is helpful for the precision of analysis. The potencies of drugs are the same when the relative potencies Ｒare close to “1” and the initial points of the regression lines are identical. Moreover, in the analysis of quantal responses, we got the same results by using two kinds of transformations ( probit transformation and logistic transformation ).