Summary Statistics

df = read.csv("~/Desktop/carValue.csv")
     attach(df)
     summary(df)
##      Car                Size            Family.Sedan    Upscale.Sedan   
     ##  Length:54          Length:54          Min.   :0.0000   Min.   :0.0000  
     ##  Class :character   Class :character   1st Qu.:0.0000   1st Qu.:0.0000  
     ##  Mode  :character   Mode  :character   Median :0.0000   Median :0.0000  
     ##                                        Mean   :0.3704   Mean   :0.3889  
     ##                                        3rd Qu.:1.0000   3rd Qu.:1.0000  
     ##                                        Max.   :1.0000   Max.   :1.0000  
     ##      Price         Cost_Mile      RoadTestScore   PredictedReliability
     ##  Min.   :16419   Min.   :0.4400   Min.   :52.00   Min.   :1.000       
     ##  1st Qu.:21922   1st Qu.:0.5700   1st Qu.:73.00   1st Qu.:3.000       
     ##  Median :28918   Median :0.6700   Median :78.00   Median :3.000       
     ##  Mean   :28340   Mean   :0.6567   Mean   :78.07   Mean   :3.407       
     ##  3rd Qu.:34102   3rd Qu.:0.7475   3rd Qu.:84.00   3rd Qu.:4.000       
     ##  Max.   :39850   Max.   :0.8300   Max.   :95.00   Max.   :5.000       
     ##    ValueScore   
     ##  Min.   :0.820  
     ##  1st Qu.:1.173  
     ##  Median :1.335  
     ##  Mean   :1.354  
     ##  3rd Qu.:1.502  
     ##  Max.   :1.990
str(df)
## 'data.frame':    54 obs. of  9 variables:
     ##  $ Car                 : chr  "Toyota Corolla (base, manual)" "Mazda3 i Touring (manual)" "Toyota Corolla LE" "Mazda3 i Touring" ...
     ##  $ Size                : chr  "Small Sedan" "Small Sedan" "Small Sedan" "Small Sedan" ...
     ##  $ Family.Sedan        : int  0 0 0 0 0 0 0 0 0 0 ...
     ##  $ Upscale.Sedan       : int  0 0 0 0 0 0 0 0 0 0 ...
     ##  $ Price               : int  16419 18895 18404 19745 18445 20150 19040 20280 16595 20300 ...
     ##  $ Cost_Mile           : num  0.44 0.5 0.47 0.52 0.53 0.57 0.57 0.52 0.47 0.54 ...
     ##  $ RoadTestScore       : int  70 74 71 70 80 74 71 68 61 60 ...
     ##  $ PredictedReliability: int  4 5 4 5 3 4 3 2 2 3 ...
     ##  $ ValueScore          : num  1.99 1.94 1.89 1.82 1.64 1.51 1.32 1.3 1.25 1.24 ...

1) Estimated regression; DV: Cost/mile – IV: family sedan & upscale sedan

lm1 = lm(Cost_Mile~Family.Sedan+Upscale.Sedan)
     summary(lm1)
## 
     ## Call:
     ## lm(formula = Cost_Mile ~ Family.Sedan + Upscale.Sedan)
     ## 
     ## Residuals:
     ##       Min        1Q    Median        3Q       Max 
     ## -0.103333 -0.049833  0.006667  0.036667  0.098000 
     ## 
     ## Coefficients:
     ##               Estimate Std. Error t value Pr(>|t|)    
     ## (Intercept)    0.52308    0.01443  36.248  < 2e-16 ***
     ## Family.Sedan   0.11892    0.01854   6.416 4.56e-08 ***
     ## Upscale.Sedan  0.23026    0.01836  12.540  < 2e-16 ***
     ## ---
     ## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
     ## 
     ## Residual standard error: 0.05203 on 51 degrees of freedom
     ## Multiple R-squared:  0.758,  Adjusted R-squared:  0.7485 
     ## F-statistic: 79.89 on 2 and 51 DF,  p-value: < 2.2e-16

Interpritition: with a unit change in cost/mile

2) Regression 2: DV: Value Score – IV: cost/mile, road-test score, predicted reliability, family and upscale sedan

lm2 = lm(ValueScore ~ Cost_Mile + RoadTestScore+PredictedReliability+
                Family.Sedan+Upscale.Sedan)
     summary(lm2)
## 
     ## Call:
     ## lm(formula = ValueScore ~ Cost_Mile + RoadTestScore + PredictedReliability + 
     ##     Family.Sedan + Upscale.Sedan)
     ## 
     ## Residuals:
     ##      Min       1Q   Median       3Q      Max 
     ## -0.14750 -0.04675 -0.00125  0.04603  0.17186 
     ## 
     ## Coefficients:
     ##                       Estimate Std. Error t value Pr(>|t|)    
     ## (Intercept)           1.370959   0.139662   9.816 4.63e-13 ***
     ## Cost_Mile            -2.265881   0.193841 -11.689 1.20e-15 ***
     ## RoadTestScore         0.011133   0.001313   8.477 4.22e-11 ***
     ## PredictedReliability  0.166210   0.010433  15.931  < 2e-16 ***
     ## Family.Sedan          0.022778   0.037989   0.600    0.552    
     ## Upscale.Sedan         0.068111   0.053706   1.268    0.211    
     ## ---
     ## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
     ## 
     ## Residual standard error: 0.07199 on 48 degrees of freedom
     ## Multiple R-squared:  0.9353, Adjusted R-squared:  0.9286 
     ## F-statistic: 138.8 on 5 and 48 DF,  p-value: < 2.2e-16

3) Stepwise variable elemination

lm3 = lm(ValueScore ~ Cost_Mile + RoadTestScore + PredictedReliability)
     summary(lm3)
## 
     ## Call:
     ## lm(formula = ValueScore ~ Cost_Mile + RoadTestScore + PredictedReliability)
     ## 
     ## Residuals:
     ##       Min        1Q    Median        3Q       Max 
     ## -0.146647 -0.050088  0.006191  0.043384  0.187797 
     ## 
     ## Coefficients:
     ##                      Estimate Std. Error t value Pr(>|t|)    
     ## (Intercept)           1.24443    0.09273  13.420  < 2e-16 ***
     ## Cost_Mile            -2.04325    0.10471 -19.514  < 2e-16 ***
     ## RoadTestScore         0.01138    0.00123   9.252 2.06e-12 ***
     ## PredictedReliability  0.16510    0.01016  16.257  < 2e-16 ***
     ## ---
     ## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
     ## 
     ## Residual standard error: 0.07213 on 50 degrees of freedom
     ## Multiple R-squared:  0.9324, Adjusted R-squared:  0.9283 
     ## F-statistic: 229.7 on 3 and 50 DF,  p-value: < 2.2e-16

4) Do smaller cars better values than larger cars?

The analysis does not prove this claim as ther is no significant diffrence. Estimates are 0.022 and 0.068, pith both p-values larger than 0.05.

5) Predicting Value Score given Road test score

lm4 = lm(ValueScore~RoadTestScore)
     summary(lm4)
## 
     ## Call:
     ## lm(formula = ValueScore ~ RoadTestScore)
     ## 
     ## Residuals:
     ##      Min       1Q   Median       3Q      Max 
     ## -0.50996 -0.21083 -0.02092  0.15999  0.68317 
     ## 
     ## Coefficients:
     ##               Estimate Std. Error t value Pr(>|t|)   
     ## (Intercept)   0.902038   0.318681   2.831  0.00659 **
     ## RoadTestScore 0.005783   0.004055   1.426  0.15984   
     ## ---
     ## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
     ## 
     ## Residual standard error: 0.2668 on 52 degrees of freedom
     ## Multiple R-squared:  0.03763,    Adjusted R-squared:  0.01913 
     ## F-statistic: 2.033 on 1 and 52 DF,  p-value: 0.1598

6) Predicting value score given predictive reliability

lm5 = lm(ValueScore~PredictedReliability)
     summary(lm5)
## 
     ## Call:
     ## lm(formula = ValueScore ~ PredictedReliability)
     ## 
     ## Residuals:
     ##      Min       1Q   Median       3Q      Max 
     ## -0.42955 -0.13373 -0.02624  0.13209  0.53377 
     ## 
     ## Coefficients:
     ##                      Estimate Std. Error t value Pr(>|t|)    
     ## (Intercept)           0.76293    0.10141   7.524 7.25e-10 ***
     ## PredictedReliability  0.17332    0.02858   6.065 1.52e-07 ***
     ## ---
     ## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
     ## 
     ## Residual standard error: 0.2081 on 52 degrees of freedom
     ## Multiple R-squared:  0.4143, Adjusted R-squared:  0.4031 
     ## F-statistic: 36.79 on 1 and 52 DF,  p-value: 1.518e-07

7) Conclusion

There is a significant diffrence between value score and predicted reliabilty. But there is no significance between value score and Road test score.