r - 计算 R 中多元回归中每个变量的解释方差

Question

我正在使用多个预测变量进行多元回归，以测试人们是否可能会说他们将签署另一份合同（李克特量表）。我需要计算我创建的问题的每个平均聚类的额外额外方差，看看除了具有很强的 beta 系数之外，这个聚类是否实际上有助于解释人们选择签署额外合同的选择。可重现的例子：

indepndent:
avg_direct_supervisor = c(4.66,4,2,2.33,2.66,3.5)
avg_friends = c(4,3.5,4,1,2.5,5)
avg_moving = c(3.4,5,2,3.5,4,3)

dependent: 
sign_contract = c(3,4,5,3,4,2)

现在我进行了多元回归

avg_direct_supervisor = c(4.66,4,2,2.33,2.66,3.5)
avg_friends = c(4,3.5,4,1,2.5,5)
avg_moving = c(3.4,5,2,3.5,4,3)
sign_contract = c(3,4,5,3,4,2)

trial <- data.frame(avg_direct_supervisor,avg_friends,avg_moving,sign_contract)

trial_model <- lm(data = trial,formula = sign_contract~.)

summary(trial_model)

Residuals:
       1        2        3        4        5        6 
 0.50480  0.50450  0.99131 -0.61958  0.09018 -1.47121 

Coefficients:
                      Estimate Std. Error t value Pr(>|t|)
(Intercept)             3.8425     3.3096   1.161    0.365
avg_direct_supervisor  -0.7697     0.9777  -0.787    0.514
avg_friends             0.2357     0.6660   0.354    0.757
avg_moving              0.3813     0.9426   0.405    0.725
Residual standard error: 1.423 on 2 degrees of freedom

Multiple R-squared:  0.2639,    Adjusted R-squared:  -0.8402 

F-statistic: 0.239 on 3 and 2 DF,  p-value: 0.8644

标签： rregressionlinear-regressionvariance