許斯淵博士生學術專題演講108/05/27

  • 2019-05-27
  • 楊 文敏

國立政治大學統計學系
     
主講人:許斯淵博士候選人 (政治大學統計學系)
    目:Suppression and Enhancement in Multiple Linear Regression
時   間:民國108527 (星期一) 下午130 
      點:國立政治大學逸仙樓050101教室
      要:
             For a linear regression model with two predictor variables, the effects of the correlation between the two predictors on estimated standardized regression coefficients and R2 have been well studied. However, the role the correlation plays may sometimes be overstated, such that confusions and misconceptions may arise. In this talk, I will revisit the issue from the perspective of a semipartial correlation coefficient. In addition, I will also take a geometrical approach to illustrate how estimated standardized regression coefficients and R2 behave as the correlation varies.