Summary

Most of us believe automatic transmission vehicles consumes more fuel than manual transmission ones. There is a very good chance that this is a misconception. Further to add, the global crude oil price keeps going north, it is high time to do an indepth analysis on various parameters which are inherent part of automobiles. This report’s main objective is to answer the following questions,

To answer above questions we have conducted study on various parameters on 32 unique models of cars for different auto makers. Cars considered for this study are across a wide gamut of price range, quality and comfort. Along with Miles per Gallon and Auto/Manual transmission information, following factors are considered for this study

Regression Models

Let us discount all variables except the Weight, Power, Cylinder Count and Displacement along with Transmission type.

##                Estimate Std. Error   t value     Pr(>|t|)
## (Intercept) 38.20279869 3.66909647 10.412045 9.084987e-11
## wt          -3.30262301 1.13364263 -2.913284 7.256888e-03
## hp          -0.02796002 0.01392172 -2.008374 5.509659e-02
## cyl         -1.10637984 0.67635506 -1.635797 1.139322e-01
## disp         0.01225708 0.01170645  1.047036 3.047194e-01
## factor(am)1  1.55649163 1.44053603  1.080495 2.898430e-01

MPG vs Transmission

## (Intercept) factor(am)1 
##   17.147368    7.244939

Above model shows, cars have an average MPG of 17.147 with automatic transmission and manual ones MPG increases by 7.245

MPG vs Weight

## (Intercept)         mpg 
##    6.047255   -0.140862

Coefficients shows Miles per Gallon decreases as the weight increases.

No of Cylinders vs Weight

##  (Intercept) factor(cyl)6 factor(cyl)8 
##    2.2857273    0.8314156    1.7134870

Weight and Cylinder counts are quite correlated. Weight increases in proportion with no of cylinders

## Analysis of Variance Table
## 
## Model 1: mpg ~ wt
## Model 2: mpg ~ wt + hp
## Model 3: mpg ~ wt + hp + cyl
## Model 4: mpg ~ wt + hp + cyl + factor(am)
##   Res.Df    RSS Df Sum of Sq       F   Pr(>F)   
## 1     30 278.32                                 
## 2     29 195.05  1    83.274 13.2261 0.001147 **
## 3     28 176.62  1    18.427  2.9267 0.098592 . 
## 4     27 170.00  1     6.623  1.0519 0.314180   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

P value for model 2 is quite low and we can conclude Miles per Gallon is determined by Weight and Power

Let us reconfirm with another set of model including the effect of transmission.

## Analysis of Variance Table
## 
## Model 1: mpg ~ factor(am)
## Model 2: mpg ~ factor(am) + wt
## Model 3: mpg ~ factor(am) + wt + hp
## Model 4: mpg ~ factor(am) + wt + hp + cyl
##   Res.Df    RSS Df Sum of Sq       F    Pr(>F)    
## 1     30 720.90                                   
## 2     29 278.32  1    442.58 70.2925  5.39e-09 ***
## 3     28 180.29  1     98.03 15.5695 0.0005107 ***
## 4     27 170.00  1     10.29  1.6348 0.2119166    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

P value for mode 2 & 3 are low, here it is again proved the impact factors are from wt and power.

Inference of Regression Analysis

Thus concluding the impact of transmission is less than Weight and Power in determining Miles per Gallon.

Annexe

Sample Car Models

##  [1] "Mazda RX4"           "Mazda RX4 Wag"       "Datsun 710"         
##  [4] "Hornet 4 Drive"      "Hornet Sportabout"   "Valiant"            
##  [7] "Duster 360"          "Merc 240D"           "Merc 230"           
## [10] "Merc 280"            "Merc 280C"           "Merc 450SE"         
## [13] "Merc 450SL"          "Merc 450SLC"         "Cadillac Fleetwood" 
## [16] "Lincoln Continental" "Chrysler Imperial"   "Fiat 128"           
## [19] "Honda Civic"         "Toyota Corolla"      "Toyota Corona"      
## [22] "Dodge Challenger"    "AMC Javelin"         "Camaro Z28"         
## [25] "Pontiac Firebird"    "Fiat X1-9"           "Porsche 914-2"      
## [28] "Lotus Europa"        "Ford Pantera L"      "Ferrari Dino"       
## [31] "Maserati Bora"       "Volvo 142E"

Exploratory Analysis

We believe weight, displacement and cylinder count have tremendous effect on determining the efficiency of the car. Also net power(in hp) is a resultant of all other factors including MPG except transmission type. Further, Do any of these parameters have direct impact on the tranmission type.

Transmission Type

We have taken 19 Automatic Transmission and 13 Manual Transmission vehicles respectively for analysis.

Cylinder Count

The cars taken for analysis has 4, 6 and 8 cylinders.

Efficiency

We have factored the efficiency regardless of rest of the parameters of the cars into 3 categories based on MPG data. Around 6 cars are highly efficient giving almost 30 miles per gallon. 5 cars in the bottom end giving less than 15 miles per gallon.

Miles per Gallon vs Power(hp)

This plot clearly shows 5/19 manual cars gives better mileage than rest of the cars. Also not a single car is on the low mileage category of below 15MPG.

Miles per Gallon vs Weight

As weight increases, mileage decreases. Weight could be the highest fact that determines the fall in fuel consumptions. It is a linear down south trend noticed through regression line.

Weight vs No of Cylinder

I am quite curious, Is number of cylinders causing the increase in weight. Plot shows a linear increase in weight as the number of cylinder increases. We might have to statistically prove, one of the parameters, either weight or cylinder count is redundant. Also note, most of the automatic cars has less number of cylinders.

Regression Models