270 likes | 399 Views
Multiple Linear Regression. (MLR). Testing the additional contribution made by adding an independent variable. Predicting SALARY using YRSRANK. Predicting SALARY using YRSRANK. SST = SSY = variation in SALARY. Predicting SALARY using YRSRANK. SST = SSY = variation in SALARY.
E N D
Multiple Linear Regression (MLR) Testing the additional contribution made by adding an independent variable.
Predicting SALARY using YRSRANK SST = SSY = variation in SALARY
Predicting SALARY using YRSRANK SST = SSY = variation in SALARY
Predicting SALARY using YRSRANK SST = SSY = variation in SALARY variation in SALARY= 1,009,361,697
Predicting SALARY using YRSRANK SST = SSY = variation in SALARY variation in SALARY= 1,009,361,697 SSR = variation explained by regression
Predicting SALARY using YRSRANK SST = SSY = variation in SALARY variation in SALARY= 1,009,361,697 SSR = variation explained by regression SSR = 117,824,722
Predicting SALARY using YRSRANK SST = SSY = variation in SALARY variation in SALARY= 1,009,361,697 SSR = variation explained by regression SSR = 117,824,722 R Square = SSR/SST ≈ .1167 or 11.67%
Predicting SALARY using YRSRANK SST = SSY = variation in SALARY variation in SALARY= 1,009,361,697 SSR = variation explained by regression SSR = 117,824,722 R Square = SSR/SST ≈ .1167 or 11.67%
Predicting SALARY using YRSRANK SST = SSY = variation in SALARY variation in SALARY= 1,009,361,697 Adding RANK as a second independent variable will explain more of the variation in SALARY, but will it be a significant amount? SSR = variation explained by regression SSR = 117,824,722 R Square = SSR/SST ≈ .1167 or 11.67%
Predicting SALARY using RANK and YRSRANK SST = SSY = variation in SALARY
Predicting SALARY using RANK and YRSRANK SST = SSY = variation in SALARY
Predicting SALARY using RANK and YRSRANK SST = SSY = variation in SALARY variation in SALARY= 1,009,361,697
Predicting SALARY using RANK and YRSRANK SST = SSY = variation in SALARY variation in SALARY= 1,009,361,697 SSR = variation explained by regression
Predicting SALARY using RANK and YRSRANK SST = SSY = variation in SALARY variation in SALARY= 1,009,361,697 SSR = variation explained by regression SSR(RANK and YRSRANK) = 683,715,472.1
Predicting SALARY using RANK and YRSRANK SST = SSY = variation in SALARY variation in SALARY= 1,009,361,697 SSR = variation explained by regression SSR(RANK and YRSRANK) = 683,715,472.1 R Square = SSR/SST ≈ .6774 or 67.74%
Predicting SALARY using RANK and YRSRANK SST = SSY = variation in SALARY variation in SALARY= 1,009,361,697 SSR = variation explained by regression SSR(RANK and YRSRANK) = 683,715,472.1 R Square = SSR/SST ≈ .6774 or 67.74%
Predicting SALARY using YRSRANK SST = SSY = variation in SALARY variation in SALARY= 1,009,361,697 Adding RANK as a second independent variable will explain more of the variation in SALARY, but will it be a significant amount? SSR = variation explained by regression SSR (YRSRANK) = 117,824,722 R Square = SSR/SST ≈ .1167 or 11.67%
Predicting SALARY using RANK and YRSRANK SST = SSY = variation in SALARY variation in SALARY= 1,009,361,697 SSR = variation explained by regression SSR(RANK and YRSRANK) = 683,715,472.1 R Square = SSR/SST ≈ .6774 or 67.74% SSR (YRSRANK) = 117,824,722
Predicting SALARY using RANK and YRSRANK SST = SSY = variation in SALARY variation in SALARY= 1,009,361,697 We may determine if this additional contribution is significant by performing a partial F-test. SSR = variation explained by regression SSR(RANK and YRSRANK) = 683,715,472.1 R Square = SSR/SST ≈ .6774 or 67.74% Additionalcontribution made by adding RANK = SSR(RANK | YRSRANK) = 683,715,472.1 - 117,824,722 = 565,890,750.1 SSR (YRSRANK) = 117,824,722
Partial F-test (α = .05) Additionalcontribution made by adding RANK = SSR(RANK | YRSRANK) = 683,715,472.1 - 117,824,722 = 565,890,750.1, the numerator.
Partial F-test (α = .05) In Simple Linear Regression, what was the relationship between the F-test and the t-test? The square root of the F ≈ 9.5969, the t value for RANK.
Predicting SALARY using RANK and YRSRANK The partial F-test and the t-test are equivalent, provided that one is examining the additional contribution of a single independent variable.