56 BULLETIN 1400, U. S. DEPARTMENT OF AGRICULTURE 
A little further discussion is required to explain why these results 
come out as they do, since for several of the items the results appear 
inconsistent. Thus the regressions for cows, area, and labor index 
are all negative, because each measures the effect of changing one 
of these factors without changing any of the others; that is, adding 
one cow to the dairy without selling any more milk would reduce 
returns bv $22 on the average; employing an additional man for a 
month, without selling any more products, would decrease returns by 
the extent of his wages, or about $44. 
After eliminating the effects of these factors the results show 
that an additional SI 00 income from crops, dairy enterprise, poultry, 
or hogs added just about the same to the operator's earnings. This 
does not mean that this represented the profit on each $100 sale of 
these items, but that on the average of the area $100 added to the 
receipts of one of them, while holding all other factors constant, 
would have added the amount indicated to the operator's earnings. 
Under these conditions the extra return would have had to be due 
almost entirely to his own extra efforts, since the extra product 
would have to be produced from the same area of land or nerd of 
cows, without the use of more labor; therefore it is but natural that 
a large part of such extra return would go to the operator himself. 
To some extent this does measure the effect upon their farm earnings 
of the extent to which different men exert themselves; the returns 
shown being the measure of this greater output from the same 
equipment when the extra energy is employed in different enter- 
prises. It is also interesting to note that on the average each of 
these four enterprises had about the same balance between receipts 
and operator's earnings — an additional $100 income from any one 
of the four adding between $60 and $70 to earnings 
It is significant that neither the beef nor sheep enterprises were 
anything like so productive, giving statistical proof of the conclu- 
sion reached earlier that these two enterprises are not economically 
suited to Chester County conditions. 
One other correlation was made with the data from the 357 farms 
to see how far differences in operators' earnings could be accounted 
for on the basis of differences in receipts from the different enter- 
prises without taking into account size or efficiency of the farm. 
The six factors representing receipts — G, H, L K, L, and M — corre- 
lated with operators' earnings, gave a multiple correlation of R = 
0.729. The coefficients of determination of the six factors in this 
solution were: 
Relative impor- 
Factor: tance, per cent 
G. Crop receipts 7. 3 
H . Dairy receipts 33. 
I. Beef receipts . 9 
k . Hog receipts 1. 4 
L. Sheep receipts . 
M. Poultry receipts 10. 
Combined importance 53. 2 
The relative importance of the different factors in this seven- 
variable solution differed but little from the earlier solution where 
13 variables were considered, except that in this one poultry receipts 
ranked ahead of crop receipts in importance. As compared with the 
