论文部分内容阅读
四、回归(一)前面论述了在静脉内输入肝素后测量相应增加的游离甲状腺索(FT_4)的假设。当发现一种药物是有效的,就要知道其他因素对药效的影响。这里将估量FT_4的初始水平对药物使其增高的影响。假定表1为10个研究对象的FT_4升高值。首先用点图把FT_4的变化与初始值相联系。再如图1所示通过点子画条直线,直线和数据的适配程度可由各点与直线的纵向距离平方和来估量。配合最好的是平方和最小的直线,叫最小二乘方直线,可用y=a+bx表示。这是线性回归方程,其中a=截距,b=斜率,即X每增加一个单位时Y值的变化。图1的a=-0.377,b=0.4937,其特定方程是y=-0.377+0.4937X。估计各点在回归直线周围的密集程度的合适指标
IV. Regression (1) The hypothesis of the measurement of the corresponding increase in free thyroxine (FT_4) after intravenous heparin was previously discussed. When a drug is found to be effective, it is necessary to know the effect of other factors on the efficacy of the drug. Here, the initial level of FT_4 will be estimated to have an increased effect of the drug. Assume that Table 1 is the FT_4 elevation of 10 subjects. First, use the point diagram to relate the change in FT_4 to the initial value. Then, as shown in Fig. 1, the straight line and the data’s fitting degree can be estimated by the sum of squares of the vertical distances between the points and the straight line. The best match is the smallest straight line of squares, called the least square straight line, which can be expressed as y=a+bx. This is a linear regression equation, where a = intercept and b = slope, which is the change in Y for each increment of X. A=-0.377 in FIG. 1, b=0.4937, and its specific equation is y=-0.377+0.4937X. A suitable indicator for estimating the density of points around the regression line