In: Statistics and Probability
1. The U.S. Department of Agriculture publishes data annually on various selected farm products. Shown here are the unit production figures (in millions of bushels) for three farm products for 10 years during a 20-year period. (leave 3 decimal places)
Corn |
Soybeans |
Wheat |
4152 |
1127 |
1352 |
6639 |
1798 |
2381 |
4175 |
1636 |
2420 |
7672 |
1861 |
2595 |
8876 |
2099 |
2424 |
8226 |
1940 |
2091 |
7131 |
1938 |
2108 |
4929 |
1549 |
1812 |
7525 |
1924 |
2037 |
7933 |
1922 |
2739 |
(1). If we want to predict the corn production by the production of soybeans and wheat. What is the dependent variable and what are the independent variables? (Written)
(2). Write down the multiple linear regression model. How to interpret the coefficients in the regression? (Written)
1. Here we want to predict the corn production by the production of soybeans and wheat.
So independent variable is soybeans and wheat and dependent variable is corn
2.
Sum of X1 = 17794
Sum of X2 = 21959
Sum of Y = 67258
Mean X1 = 1779.4
Mean X2 = 2195.9
Mean Y = 6725.8
Sum of squares (SSX1) = 700612.4
Sum of squares (SSX2) = 1494276.9
Sum of products (SPX1Y) = 3814621.8
Sum of products (SPX2Y) = 3477858.8
Sum of products (SPX1X2) = 739466.4
Regression Equation = ŷ = b1X1 +
b2X2 + a
b1 =
((SPX1Y)*(SSX2)-(SPX1X2)*(SPX2Y))
/
((SSX1)*(SSX2)-(SPX1X2)*(SPX1X2))
= 3128341511432.1/500098368444.6 = 6.25545
b2 =
((SPX2Y)*(SSX1)-(SPX1X2)*(SPX1Y))
/
((SSX1)*(SSX2)-(SPX1X2)*(SPX1X2))
= -384153649078.4/500098368444.6 = -0.76816
a = MY - b1MX1 -
b2MX2 = 6725.8 - (6.26*1779.4) -
(-0.77*2195.9) = -2718.35776
ŷ = 6.25545X1 - 0.76816X2 -
2718.35776
So for every increase in x1(=soyabeans), y (=corn) will increase to 6.25545
For every increase in x2(=wheat), y(=corn) will decrease to 0.76816