Professional Documents
Culture Documents
The Monthly Flocculation of Gasoline Prices
The Monthly Flocculation of Gasoline Prices
of Gasoline Prices
Does time of year affect the price of gas?
12/19/2008
IB Math Studies
Sarah Beck
A. Statement of Task
The American Automobile Association (AAA) gives a daily American national
average of the fuel prices that are derived from credit card transactions from more
than 85,000 stations across the nation. Daily averages are able to inform the
consumer superficially in decisions of getting gas today or tomorrow. Daily averages
do not give the consumer a yearly outlook, which is the purpose of this
investigation. Monthly averages could help families and businesses save money
when they are planning a trip, that’s price is affected by the price of gasoline.
I plan to take national gasoline daily averages from a 10 year period, 1998-
2007, from the Energy Information Administration which has the official energy
statistics from the US government. From these daily averages, I plan to make
monthly averages. With the monthly averages, I will create a scatter plot,
containing a best fit line. Then, I will use the correlation coefficient to measure the
strength and the direction of a linear relationship between tw (U.S. Retail Gasoline
Prices, Regular Grade)o variables. I will also use the coefficient of determination
because it gives the proportion the variance (fluctuation) of one variable that is
predictable from the other variable. Using the slope of the best fit lines from the 12
months, I will compare them to the slope of the best fit line of all of the average gas
prices from 1998-2007 by their percentage differences. The comparison will show if
there is any correlation between gasoline and the month of year.
B. Data Collection
Y Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Avera
ear ges
1 1.13 1.08 1.04 1.05 1.09 1.09 1.07 1.05 1.03 1.04 1.02 0.98 1.0593
998 1 2 1 2 2 4 9 2 3 2 8 6 33
1 0.97 0.95 0.99 1.17 1.17 1.14 1.18 1.25 1.28 1.27 1.26 1.29 1.1650
999 2 5 1 7 8 8 9 5 4 4 8 83
2 1.30 1.36 1.54 1.50 1.49 1.61 1.59 1.51 1.58 1.55 1.55 1.48 1.51
000 1 9 1 6 8 7 3 2 9 5 9
2 1.47 1.48 1.44 1.56 1.72 1.64 1.48 1.42 1.53 1.36 1.26 1.13 1.461
001 2 4 7 4 9 2 7 1 2 3 1
2 1.13 1.13 1.24 1.40 1.42 1.40 1.41 1.42 1.42 1.44 1.44 1.39 1.3575
002 9 1 7 1 4 2 3 2 9 8 4
2 1.47 1.64 1.74 1.65 1.54 1.51 1.52 1.62 1.72 1.60 1.53 1.49 1.5907
003 3 1 8 9 2 4 4 8 8 3 5 4 5
2 1.59 1.67 1.76 1.83 2.00 2.04 1.93 1.89 1.89 2.02 2.01 1.88 1.8801
004 2 2 6 3 9 1 9 8 1 9 2 67
2 1.82 1.91 2.06 2.28 2.21 2.17 2.31 2.50 2.92 2.78 2.34 2.18 2.2953
005 3 8 5 3 6 6 6 6 7 5 3 6 33
2 2.31 2.31 2.40 2.75 2.94 2.91 2.99 2.98 2.58 2.27 2.24 2.33 2.5889
006 5 1 7 7 7 9 5 9 2 1 4 17
2 2.27 2.28 2.59 2.86 3.13 3.05 2.96 2.78 2.78 2.79 3.06 3.02 2.8005
007 4 5 2 2 1 2 9 3 9 83
T 1.54 1.58 1.68 1.80 1.87 1.86 1.84 1.84 1.87 1.81 1.77 1.72 1.7708
otal 92 46 33 98 62 03 94 66 72 68 56 14 67
Av.
C. Analysis
This section contains a short review of the type of math used during the
analysis portion of this investigation. Also included are a few definitions in order to
demonstrate a clear and understandable experiment.
Slope:
y = mx + b
Formula for
linear
regression
(best fit line)
slope =
= n( xy) - ( x)(
m
y)
n( x2) - ( x)2
y - m(
intercept x)
=
=b
n
x y xy x2
Key:
1 1.131 1.131 1
2 0.972 1.944 4
3 1.301 3.903 9
4There 1.472 5.888 16
are 10
terms;
therefo
re,
n=10
5 1.139 5.695 25
6 1.473 8.838 36
7 1.592 11.14 49
4
8 1.823 14.58 64
4
9 2.315 20.83 81
5
10 2.274 22.74 100
55 15.49 96.70 385
2 2
Therefore:
10(96.702) -
(55)(15.492)
114.96
slope =
==
m
10(385) - (55)2
825
= .1393454545 ≈
0.139
=
intercept
= y - m(
=b x)
15.492 -
. 0.7828 ≈ 0.783
0.139345454(55)
=
10
I then used the percentage difference equation to figure each month’s average
to the 10 year average. 10 Year Average January Average
Month Percentage
Difference
January -13.4%
February -11.1%
March -5%
April +2.17%
May +5.78%
June +4.93%
July +4.34%
August +4.19%
September +5.83%
October +2.56%
November +2.67%
December +2.83%
D. Evaluation
The mathematical applications applied to this experiment illustrated that
there is some significant relation between the price of gasoline per gallon and the
time of year.
I have hypothesized about why gas prices seem to have this cycle of price
fluctuation, which can be seen below.
My data in graphical form shows that year after year, there was a steady
increase in fuel prices per gallon. I also included the coefficient of determination (r2)
in order to show how well the measure of the regression line represents the data.
Because all of my coefficients of determination follow this value: 0.75 ≤ r2 < 0.90;
there is a strong correlation between my data and the best fit line. Having a strong
correlation is good because it shows that there was only a small room for error in
the best fit line.
E. Sources
Affairs, Bureau of Consular. Hurricane Season - Know Before You Go. 3 December
2008
<http://travel.state.gov/travel/cis_pa_tw/hurricane_season/hurricane_season_2915.h
tml>.