论文部分内容阅读
1、回归效果的显著性检验 1.1相关系数的显著性检验 在第五讲中已经阐明,相关系数r是表征X与Y两个变量间线性关系密切程度的一个重要特征参数。定性地说|r|愈大,Y与X间的线性关系愈密切。这意味着,只有当|r|值达到一定程度时,用回归直线表示Y与X之间的关系,即配回归直线,才有意义。那么,定量地说|r|究竟多大时,Y与X之间才存在着线性关系,配回归直线才有意义呢?这就是相关系数显著性检验所要研
1. Significance test of regression effect 1.1 Significance test of correlation coefficient It has been clarified in the fifth lecture that the correlation coefficient r is an important characteristic parameter that characterizes the close linear relationship between X and Y variables. Qualitatively speaking | r | the greater the linear relationship between Y and X more closely. This means that only when the value of | r | reaches a certain level, it is meaningful to use the regression line to represent the relationship between Y and X, that is, return to the straight line. Then, quantitatively say | r | exactly what time, there is only a linear relationship between Y and X, with return to the straight line makes sense? This is the correlation coefficient significance test to be studied