Learning Resources

Home »  » Courses » Mathematics » Mathematics 3204 (delisted) » Unit 03 » Set 02 ILO 02 » Solution
Investigation 6: Pennies from Heaven

Test yourself

Year

1975

1980

1985

1990

1995

2000

Population

1600

1475

1355

1245

1145

1055

Let 1975 be year 0. The other years then become year 5, 10, 15 , 20, etc. Enter this in List L1. Enter the corresponding populations in List L2. This is shown below:

Now draw a scatter plot of this data that we have entered in Lists L1 and L2. This is shown below:

It is difficult to tell from the graph if it is linear or exponential. We will perform both a linear and an exponential regression and see which one "fits the best".

First the linear regression is shown below:

The r is referred to as a correlation coefficient and r2 as a coefficient of determination. They basically tell how good the regression equation fits the data. The closer they are to 1 the better the fit. Compare the values of these coefficients for the linear regression at the left and the exponential regression below. (Note: to get r and r2 displayed, you have to go to [CATALOGUE] and press [ENTER] at the DiagnosticOn command).

Note the values of r and r2 for the exponential regression on the left. They are much closer to 1 than they were for the linear regression. The data in the table is thus more exponential than it is linear. The desired function that best models the data is thus:

y = 1601 (0.9834)x

Next the exponential regression is completed:

To find what the population will be in 2005 if this trend continues, we substitute x = 30 (2005 is 30 years after the start of the data) into the regression equation. This gives:

y = 1601 (0.9834)30 = 968.9

Since the population must be a whole number, we round the answer to 969.