In: Statistics and Probability
Numbers (in thousands) of forest fires over the year and the number (in hundred thousands) of aces burned for 7 recent years are shown.
Number of fires (x): 72 69 58 47 84 62 57
Number of acres burned (y): 73 70 59 48 85 63 58
The correlation coefficient for the data is r = 1 and standard deviation = 0.05.
Should the regression analysis be done?
Find the equation of the regression line. Round to three decimal places.
y' = ax+b
a =
b =
Here Dependent variable is Number of acres burned (y) and independent variable is Number of fires (x)
-------------------
As correlation between the two variables is 1.
But here point is that whether this correlation is meaningfull or nuinsance.So,this is a meaning ful correlation since logically we can estimate the Number of acres burned by using the number of fires in the forest.
--------------------
Question:-
Should the regression analysis be done?
Yes,Regression can be performed for the above data.
--------------------
Question:-
Find the equation of the regression line. Round to three decimal places.
y' = ax+b
a = Slope of the line
b = Intercept made by the line with the y-axis
Formulas to calculate a and b
Where xbar and ybar are the arithmetic means of x and y respectively.
Calculations:-(Model fitting)
(Excel is being used here)
cov(x,y) = 144.4672 ---- =covariance.s(array1,array2)
var(x) = 144.4672 ---- = var.s(number1,..number m)
Hence a= 144.4672 / 144.4672
a = 1
And to calculate b;
xbar = 64.14286 ----=average(number1,...,number m )
ybar = 65.14286
b = 65.14286 - 1* 64.14286
b = 1
Therefore,
The regression line for the above data is
i.e
which is obvious since correlation(y,x) = 1