【摘要】 本文研究了线性回归模型中试验点（x’_i,y_i）和（x’_j,,y_j）对参数β的最小二乘估计β的联合影响。在用welsch统计量Q₁²度量指标属于I的试验点的联合影响中,确定点（x’_i,y_i）和(x_ji,y_j)的交互影响作用,h_ij扮演一重要角色。而且,当h_ij=0时,不管γ_i和γ_j的值,有Q₁²=Q_i²+Q_j²;当h_ij≠0时,h_ij,γ_i和γ_j的符号和大小确定Q₁²比Q_i²+Q_j²大多少或小多少。更多还原

【Abstract】 This article studies the joint influence of cases （x_i′,y_i） and（x_j′,y_i） for the LSE β of β in Linear Regression Model.The h_ii plays an important role in deterning the interaction effect of cases（x_i^′, y_i）and（x_j^′, y_j） in the joint influence of I as measured by Welsch statistic Q₁².Moreover then Q_I²=Q_i²+Q_i², when h_ij=0, regardless of the values of γ_i and γⁱ and when h^ji=0, the signs and magnitudes of h_ij, γ_i and γ_j determine how much Q_I² is greater than or less than Q_i²+Q².更多还原