r = -0.90
\(y=a+bx\)
\(My\ grumpingness = 125.96 - 8.94 * My\ sleep\)
\(\hat{Y}_i=b_0+b_1X_i+\epsilon_i\)
\(\epsilon_i = Y_i - \hat{Y}\)
評估迴歸模型的自變項(\(X_i\))解釋或預測應變項(\(Y\))的指標。
Assumption Checks
資料變項能構成線性關係\(y=a+bx\)
Collinearity statistics
(VIF)每項資料點的殘差(\(Y_i-\hat{Y_i}\))取樣來源彼此獨立
Autocorreltion test
(Durbin-Watson test)所有資料點的殘差(\(Y_i-\hat{Y_i}\))之取樣分佈符合常態分佈。
Normality Test
(Shapiro-Wilk test)Q-Q plot for statistics
所有資料點的殘差(\(Y_i-\hat{Y_i}\))之取樣分佈符合平均值是0,標準差是\(\sigma\)的常態分佈。
Normality Test
(Shapiro-Wilk test)Residual plots