You are on page 1of 10

Tittle: The relationship between the weigh of a car and the fuel consumption

Research Question: Is there a correlation between the weight of a car and


its fuel consumption?
Word Count: 1915

Table of Contents
Introduction. 3
Method. 3
Data collection. 4
Calculations. 4
Conclusion. 10
Interpretation. 11
Validity. 11
Areas of improvement. 12
Appendix. 13
Bibliography. 13
Introduction
I have always been interested in cars and had an easy time learning anything about them and
have a great substantial amount of knowledge about them. I also thought since I will be
looking into purchasing a car why not look at what would be the best for me financially. I did
this through looking at if there is a correlation between the weight of a car and the fuel
consumption that car has. My hypothesis for this study is that there will be a strong
correlation between the weigh and millage and that the weight will be a leading factor to
having a lower or a higher fuel consumption. My reasoning for this is that the heavier a car
might be the more energy it will require to move, thus, consuming more fuel and air. My null
hypothesis being that there is no real correlation and that the weight of a car is not a leading
factor to having a high fuel consumption or a low fuel consumption.

Method
I did this through
1. Collecting the data from a site that has specifications of multiple cars of their
weights, fuel consumption (per 100km) and much more.
2. Arranged the data according to their weight brand and car model to give a
broader and accurate data to the type of cars that are on the road.
3. Reconstructed the data vertically so that it could be used in a graph as the
previous way didn’t provide an accurate scatter plot graph.
4. Used a scatter plot graph to be able to determine if there was a correlation
between the weight of a car and the fuel consumption it has.
5. Found the averages and used the averages in a different scatter plot in hopes of
producing a better scatter plot graph.
6. Used the “Pearson’s Correlation Coefficient formula”[1] to calculate the exact
strength of the coloration.
7. Checked the validity of the answer in the “Pearson’s Correlation Coefficient
formula”[2]through calculating it in the calculator.

Data collection

The data collected as shown above are all from a car website known as
“https://www.autoevolution.com/cars/ford-focus-st-estate-2012.html#aeng_ford-focus-st-
estate-2012-20-gtdi-ecoboost-250-hp”[3]. I focused on having the car year models between
2012 and 2017 due to the new environmental laws, cars have to change their fuel
consumption to meet the new standard thus having these years will make the data more
accurate and up to date with most cars on the road. I also decided to find the averages of the
values in hopes of producing a better graph as you can see below.

Calculations
After I have found the data I thought that getting the averages of the values would be better
way of representing it on a graph.
Although you can’t really see if there is an actual correlation so I decided to rearrange my
data and add it all into the following graph and it gave a clear distinct answer.

As you can see there is no correlation between the weight of a car and the mileage of that car.
This meaning that through the use of the scatter plot diagram it shows that the weight of a car
is not a determining factor of the fuel consumption. Due to there being such a huge amount of
data it is difficult to see how strong the correlation is exactly. One way to determine the exact
strength of the correlation is through the “Pearson Correlation Coefficient formula”[4].
First, we rearrange the data into the following table and further fill out the equations. As seen
below
The table above shows that
= 1979.52
= 8.98
Now that we have the data we need we need. We can fill in the missing number in “Pearson’s
Coefficient Formula”[5].

r 0.315
Now to analyse what this value means.
for there to be a strong correlation to be in existence then the r value would be as close to one
as possible. The further away the value is from one the greater the non-correlation gap
becomes.
Although the calculations done here could be incorrect and one way to double check this
methods validity is through the calculator as shown below.
First, we add all the data correctly into the lists available in our calculator. To get there we
press STAT and then EDIT.

After that something like this should show.


This is where I added the lists of data I collected. Each category of data in a different list. I.e.,
L1, L2. As shown below.

After the lists are in it’s time to calculate the r value. I did this by pressing STAT again, but
now pressing the right arrow and moving to CALC. after that I move down to LinReg and

pressed ENTER to get the values.

Here comes the final step to calculating the r value. Now you have to assign the
correct lists to the x and y list so that the calculator knows what to calculate.
After that press enter by calculate.
you should get the values you need to find the strength of the correlation. As
show below.

As you can see the r value on the calculator (r=0.315) exactly equals the value that I found in
the formula, proving that this is indeed the correct value of the strength of the correlation
between car weights and the mileage of that car.

Conclusion
In conclusion of my study, I have found that there is an extremely weak correlation between
the weight of the car and the fuel consumption that car archives. Thus, my hypothesis is
incorrect and I have to accept my null hypothesis which is that there is no real correlation
between the weight of a car and the fuel consumption. One explanation for this could be that
it doesn’t depend on the weight of the car but depends on the size of the engine and more
specifically the size of the piston and the number of pistons. This will all measure into what
performance it gives off. To see if they are actual factors to the mileage the car can achieve
more study’s will have to be conducted. Although from previous experience the performance
of the car will have a great effect on the fuel consumption. The reason for that would be that
the actual driving force that makes the car moves is the explosion of fuel and air. Now if the
car needs to go quicker than the amount of fuel and air will be increased to create a bigger
explosion. Or the more pistons the car has the more explosions it needs to function thus
consuming more fuel and air. A Bigger explosion equals more power, thus, cars that have
more horsepower use more fuel. Another factor could be the aerodynamics of the car. A
study was done that showed the more “aerodynamic your car is the less drag it has causing
less fuel consumption”[6]. The reason for this being that there are less forces holding your car
back and essentially stopping your car.

Interpretation
To find if weight was indeed a factor to fuel consumption I had to find out if there was a
correlation between fuel consumption and the weight of the car. I first collected all the data I
needed to come to my conclusion and then inserted all the data into graphs to see If there was
a correlation. I thought that, using the averages of the weights of the types of cars I found
would show a better scatter plot but I was wrong so I then pursued in using all the data. After
reconstructing the data vertically. I used that data to make a new scatter plot and found a
clearer answer. There was no clear correlation between the weight of the car and the fuel
efficiency of it. Although it did no show exactly how strong the correlation it was. This is
where I decided to find the exact strength of the correlation and found it through the
“Pearson’s Coefficient formula”[7]and with a validity check on the calculator that the
correlation was indeed very weak. Thus, I had to accept my null hypothesis which is that, the
weight of the car is not a leading factor to its fuel efficiency.

Validity
I went through great lengths to make sure that, the data I collected was correct, the data I
arranged was done correctly and the calculations I did were done correctly. For collecting the
data. I decided to have different brands and different types of cars. Thus, I had to be sure that
for instance a hatchback Volkswagen was a hatchback Volkswagen on the website I collected
the data from. So, I decided to use google images typed in the car brand and type I was
looking for. Did research on it and found that it was truly the type I was looking for. Then
and only then when I was satisfied went to the website typed the car brand and the models
name in found the specifications I needed and added it to my data. There was one more
aspect I kept consistent to make sure the data was correct and that was I took the fuel mileage
from city driving and highway driving and combined them, this the website already did for
me. I also got the weight of the car with a full tank of fuel because who drives around on a
completely empty tank of fuel.
The second thing I did after collecting the data was arrange the weighs of the car from
smallest to largest on excel. Now you can’t do that and then arrange the fuel mileage from
smallest to largest because then you won’t have the correct data, as it will be assigned
incorrectly. So, I made sure that the weight of the car had its correct fuel mileage.
The final step I did was the calculations and the procedures I did to make sure that the
calculations were correct are as follows. After I saw in the graph that there was no real
correlation I decided I wanted to know exactly how strong/weak the correlation was. Thus, I
decided to do that through the “Pearson’s Coefficient formula”[8]. I went through my data
allocations thoroughly and checked and rechecked that my data I put in the formula was
correct, but that wasn’t enough for me so I decided to do the calculation in the calculator as
well. I re-listed all the data in the L1 and L2 lists then did the calculation. After I did that and
found the correlation values were exactly the same I was satisfied.

Areas of
improvement
Some areas of improvement I think I could have done were, the data I collected should have
been from the same car year. For instance, not having it from 2012 to 2017 but only 2017
models. That could have given a different result. Another area of improvement could have
been to double check if the car models were what they said they were. You can never be to
certain that the data you collected were correct or the calculations you did were correct.
One last improvement would be to maybe find another way of finding the correlation or
maybe testing the validity to make sure the answer I got was entirely correct.

Appendix
Bibliography
Car specification Website: https://www.autoevolution.com/cars/ford-
focus-st-estate-2012.html#aeng_ford-focus-st-estate-2012-20-gtdi-
ecoboost-250-hp
Aerodynamics and fuel consumption study:
http://illumin.usc.edu/252/drag-reduction-the-pursuit-of-better-fuel-
economy/

[1] IB Mathematics studies SL book, Pg. 321


[2] Ibid, Pg. 321
[3] Car statistics website
[4] Ibid, Pg. 321
[5] Ibid, Pg. 321
[6] Aerodynamics and fuel study.
[7] Ibid, Pg. 321
[8] Ibid, Pg. 321

You might also like