Allegany College
Pre-Calculus 119 Final Project
By: Will Spencer
Instructor: Dr. Mark Shore
Data Source: National Center
for Health Services
Data set found at http://www.cdc.gov/nchs/fastats/pdf/47_28t12.pdf
Objective: To determine if there is a continual increase of life expectancy, and to see if there is an overlap of life expectancy of Black males versus white males.
Independent
Variable: Years (T=0=1900 in decades.
T=10=1910)
Dependant Measure
1: Life
expectancy in White Males
Dependant Measure
2: Life
expectancy in Black Males
Data Points for
Life Expectancy for White Males:
|
Independent
Variable (Years) |
Dependant
Variable 1 (Life Expectancy
of White Males) |
|
|
|
|
1910 |
48.6 |
|
1920 |
54.4 |
|
1930 |
59.7 |
|
1940 |
62.1 |
|
1950 |
66.5 |
|
1960 |
67.4 |
|
1970 |
68.0 |
|
1980 |
70.7 |
|
1990 |
72.7 |
|
|
|
Interpolation: Finding the curve of best fit within the range of the data.
The following chart shows
the model with the highest correlation coefficient based on the data from 1910
to 1990.
Equation
Type: Coefficient: Model:
Linear R2= 0.9283 y= 0.2787x + 49.411
Quadratic R2= 0.985 y= -0.003x2 + 0.5828x + 43.836
Exponential R2= 0.8993 y= 50.042e0.0046x
Logarithmic R2= 0.989 y= 10.952Ln(x) + 22.548
Power R2= 0.9926 y= 31.843x0.1826
Cubic R2=
0.9942 y= 6E-05x3
- 0.0116x2 + 0.9429x + 40.075
The Equation with the highest correlation coefficient is the Cubic equation, which will be used now.
This chart reflects the actual data, compared to the predicted data using the Cubic equation.
|
Year |
Age in Years
using the Actual Data |
Age in Years
using the Predicted Data with the Cubic Equation |
|
|
|
|
|
1910 |
48.6 |
48.40 |
|
1920 |
54.4 |
54.75 |
|
1930 |
59.7 |
59.47 |
|
1940 |
62.1 |
62.89 |
|
1950 |
66.5 |
65.37 |
|
1960 |
67.4 |
67.23 |
|
1970 |
68.0 |
68.83 |
|
1980 |
70.7 |
70.51 |
|
1990 |
72.7 |
72.60 |
|
|
|
|
In the long run, what does your linear equation predict for your data? This equation is not good for extrapolation because it predicts a steady rise in life expectancy into infinity. It does not level out anywhere, and therefore is invalid.
In the long run, what does your quadratic equation predict for your data? This equation is not good for extrapolation because after the year 1995, there is a steady increase in the life expectancy, and since we are alive today, it must be invalid.
In the long run, what does your cubic equation predict for your data? In the long run, the cubic equation is not good. It predicts that the life expectancy will continue to rise until infinity. This is not good for extrapolation. However, this equation is perfect for interpolation.
In the long run, what does your forth power equation predict for your data? This equation is not good for extrapolation because it starts with the life expectancy at infinity, and coming down at the year 1862 then increasing again, only to finally rise into infinity.
In the long run, what does your exponential equation predict for your data? This equation is good for interpolation, but like some of the others, the life expectancy continues to raise to infinity.
In the long run, what does your logarithmic equation predict for your data? This equation is okay for interpolation, and is a little better for extrapolation. The life expectancy slowly rises into infinity, but after an infinite amount of years.
In the long run, what does your power equation predict for your data? This equation is also decent for interpolation, but in the long run, slowly increases into infinity. However, it is the best for extrapolation because of the slow rate of the increase into infinity. This makes sense because more than likely, technology will increase at about the same rate, making longer life expectancies possible.
In your opinion, which equation is best for extrapolation? Why? Probably the best equation for extrapolation would have to be the power equation. It still raises the life expectancy into infinity, but it does it at a much slower rate than any of the other equations.
The power equation will be used to predict some extreme values for the future.
|
Year |
Predicted Value
using the Power equation |
|
|
|
|
2050 (150) |
79.51 |
|
3000 (200) |
83.80 |
|
3500 (250) |
87.28 |
|
4000 (300) |
90.24 |
|
5000 (400) |
95.11 |

Data Points for
Life Expectancy for Black Males:
|
Independent
Variable: (Years) |
Dependant Variable 2(Life
Expectancy, Black Males) |
|
|
|
|
1910 |
33.8 |
|
1920 |
45.5 |
|
1930 |
47.3 |
|
1940 |
51.5 |
|
1950 |
59.1 |
|
1960 |
61.1 |
|
1970 |
60.0 |
|
1980 |
63.8 |
|
1990 |
64.5 |
|
|
|
Interpolation: Finding the curve that best fits within the range of the data.
The following chart shows
the model with the highest correlation coefficient based on the data from 1910
to 1990.
Equation
Type: Coefficient: Model:
Linear R2= 0.887 y= 0.3545x + 36.342
Quadratic R2= 0.9681 y= -0.0047x2 + 0.8276x + 27.669
Exponential R2= 0.8349 y= 37.257e0.0071x
Logarithmic R2= 0.9721 y= 14.131Ln(x)
+ 1.4296
Power R2= 0.9683 y= 18.008x0.2902
Cubic R2= 0.9711 y= 4E-05x3 - 0.0111x2 +
1.0946x + 24.88
The Equation with the highest correlation coefficient is the Logarithmic equation, which will be used now.
This chart reflects the actual data, compared to the predicted data using the Logarithmic equation.
|
Year |
Age in Years (Actual Data) |
Age in Years
with Predicted data from Logarithmic equation |
|
|
|
|
|
1910 |
33.8 |
33.96 |
|
1920 |
45.5 |
43.76 |
|
1930 |
47.3 |
49.49 |
|
1940 |
51.5 |
53.55 |
|
1950 |
59.1 |
56.70 |
|
1960 |
61.1 |
59.28 |
|
1970 |
60.0 |
61.46 |
|
1980 |
63.8 |
63.35 |
|
1990 |
64.5 |
65.01 |
|
|
|
|
In the long run, what does your linear equation predict for your data? This equation is not good for extrapolation because it increases at a constant rate, and will increase until infinity.
In the long run, what does your quadratic equation predict for your data? This equation may be good for interpolation, but not extrapolation. The life expectancy increases until the year 1987, then it begins to decrease to below zero, which is impossible. And since we are alive today, 1987 would not be a valid ending date.
In the long run, what does your forth power equation predict for your data? This equation is not good for either interpolation or extrapolation. It begins with the life expectancy from previous years to be infinity, then in decreases and goes up and down again. It finally ends by going into infinity. This is not good for extrapolation.
In the long run, what does your cubic equation predict for your data? This equation is excellent for interpolation, but not for extrapolation. After 100 years, the life expectancy begins to increase too much making unrealistic values.
In the long run, what does your exponential equation predict for your data? This equation is good for interpolation, but much like the cubic equation, after 100 years the data begins to be unrealistic.
In the long run, what does your logarithmic equation predict for your data? This equation is not only good for interpolation, but it is also good for extrapolation. The increase of life expectancy is slow enough to give realistic values. For example, in the year 5000, the life expectancy is only 86 years old. This is great for extrapolation.
In the long run, what does your power equation predict for your data? This equation reacts similar to the logarithmic equation, but in the long run the numbers are not as realistic as the logarithmic equation, making it not the best for extrapolation.
In your opinion, which equation is best for extrapolation? Why? The best equation for extrapolation would have to be the logarithmic equation. In the long run its numbers are very realistic. A growth in technology would result in a longer life expectancy as the years go on, but it would have to be over many years. The logarithmic equation shows this beautifully.
The logarithmic equation will be used to predict some extreme values for the future.
|
Year |
Predicted Value
using the Logarithmic equation |
|
|
|
|
2050 (150) |
72.23 |
|
3000 (200) |
76.29 |
|
3500 (250) |
79.45 |
|
4000 (300) |
82.02 |
|
5000 (400) |
86.09 |

Equation that is best for extrapolation in Dependant Measure 1 (Life Expectancy of White Males):
Power R2=
0.9926 y= 31.843x0.1826
Equation that is best for extrapolation in Dependant Measure 2 (Life Expectancy of Black Males):
Logarithmic R2= 0.9721 y= 14.131Ln(x)
+ 1.4296
There is no intersection point on this data, which means that the average life expectancy for White Males have and according to this data, always will be higher than that of Black Males.
Project
Conclusions:
This data shows that the average life expectancy of white males have always been higher than that of black males. There are probably many reasons for this: sickle cell disease, segregation, and sadly, racism are all causes for a black males life expectancy to be shorter. However, with a steady increase in technology both white and black males have seen a steady increase in life expectancy and will continue to do so until a certain point.
