Simple Regression--Least Squares


 With hand-held calculators, simple regression can be
figured easily. However, it is worthwhile to understand
the procedure in case such a calculator is not available.


Table VI-2.


Data and calculations for
regression.


 Xl X2 Y

 0.0 0.0 5.50
 0.0 0.0 5.20
 1.5 2.25 7.00
 1.5 2.25 6.35
 3.0 9.0 6.95
 3.0 9.0 7.00
Total 9.0 22.50 38.00

X or Y 1.5 3.75 6.333

Note: X1 = N and X2 = N2

In order to compare the statistical method with the
visiographic method, the same data set will be used. To
facilitate the calculation by a calculator, the data will
be tabulated in a somewhat different form (Table VI-2).
To avoid the use of very large numbers, the data have been
coded. An appropriate coding procedure is to consider N
in units of 100 kg and yield of rice in metric tons. In
this way, all the values fall between 0 and 10. The
required calculations follow. Take careful note of
capital and lower-case letters.

 Ex12 = ZX12 [(EZX)2 / n]

 where EX12 = (1.5)2 + (1.5)2 + (3.0)2 + (3.0)2

 and (EX1)2 / n = 92 / 6
 = 13.5

 then Zx12 = 22.5 13.5
 =9.0


quadratic