An Apple Orchard Survey of Ontario County 213 



Value of the general average of the group. — In comparing the yields and 

 incomes of the different classes in the same table, it is not enough to con- 

 sider merely the figures of the table. In many cases, the absolute value 

 of one class may be above another, yet if both are below the general average 

 of the population neither should be considered of especial merit, for a 

 desirable practice is one that produces a yield and income above the general 

 average. 



The general average is of value, also, in determining whether or not 

 there is any relation between the condition or practice and the yield and 

 income. When we find that in classes large enough to be considered 

 statistically the yields and incomes of the different classes not only are 

 very close to each other but approximate the general average, we may 

 conclude that the condition under discussion has no influence, but that 

 the law of probability is in operation. If, on the other hand, the classes 

 fall into distinct groups, one lying above the general average and one below, 

 it is quite evident that something other than the law of chance is at work. 

 We may conclude safely, then, that the one practice is beneficial and the 

 other injurious. 



Shall the orchard or the acre he used as unit? — It has been shown that 

 there are various ways of determining an average. In orchard surveys a 

 further difficulty may be encountered, namely, the question of the units 

 whose values are to be averaged. If we consider the orchard as a unit 

 representing a certain set of conditions, the effect of various types of or- 

 chard management is better shown by a yield per acre per orchard. The 

 conditions in each orchard, moreover, are determined largely by the man 

 who is working the orchard ; and the figures for all the acres of the orchards 

 are the estimate of this one individual. By taking the orchard as unit, 

 therefore, we may not only determine an average representative of the 

 condition of the orchards, but may deal with the farmer who is responsible 

 for the condition. The orchard represents the man for whom these results 

 are being determined; the acre, in itself, represents so much land that 

 may be used well or ill. It does not express the conditions in the orchard 

 but is used as a measure of the size of the orchard. When the acre is 

 used as unit, the mean is weighted, as far as the orchard is concerned, 

 and the size overwhelms the results. This weighting is purely relative, 

 for the mean obtained by using the acre as unit is a correct mean for the 

 district which those acres represent. But such a mean does not give, 

 necessarily, a correct idea of the results of the conditions of the orchards 

 in that district. On the other hand, the mean obtained by using the or- 

 chards as unit would not give a correct idea of the mean per acre. If 

 a knowledge of the total amounts raised or the value of the crops is para- 

 mount, the acre is the only unit to be used. 



