问题如下图:
选项:
A.
B.
C.
解释:
为什么正序列相关会使标准误变小?统计量变大?
NO.PZ2018101001000040问题如下Whiof the following statements about positive sericorrelation is most likely correct?A.It will not affethe consistenanestimation of regression coefficients.B.It will lea higher stanrerrors of regression coefficients.C.It will lea smaller t-statistiof regression coefficients.A is correct. 考点: Sericorrelation.解析: 本题考的是正序列相关的特征。正序列相关只会影响残值ε的波动,但对回归方程本身的精确度无影响,也不会影响系数的估计。但正序列相关会使标准误变小,造成t统计量变大。所以A的描述是正确的,B和C的描述是错误的。选择A。是指比如Yt=a0+a1X1+a2Yt-1+u,a1的估计值还是准确的吗?positive sericorrelation的定义可以再讲一下吗?
NO.PZ2018101001000040 B错在哪里
老师好,请问为什么C不对呢? 比如这题, If Tylor tests the null hypothesis whiis no sericorrelation in the regression resials anthe results shows ththe rbin-Watson statistic is 1.0165. Angiven the criticvalues the 0.05 significanlevel for the rbin-Watson statistic are =1.35 an=1.49. Whiof the followings is most likely correct? 就是1.0165小于1.35,所以是positive啊
负的序列自相关会导致t值和stanrerror变大还是小呢? 那异方差和条件异方差会导致t值和stanrerror变大还是小呢?
老师,三个有关残差项的assumptuon都满足A么? 不会影响mol的拟合度?