16 



This figure eliminates every other causal factor than diversity, 

 and hence may be safely used in tabulations designed to bring out 

 the effect of diversity on the efficiency of the farm business. 



58. Adjusted labor income. In the older settled portions of the 

 country magnitude of business is perhaps the most important factor 

 in the success of the farmer. This means that, in the majority of 

 cases, it has more influence on labor income than any other factor 

 This being the case, it is difficult to bring out the effect of other 

 factors unless the number of farm records is sufficiently large to 

 insure the elimination of the effect of magnitude of business by the 

 law of averages, or to permit grouping the farms first on magnitude 

 of business and then subdividing each group on the basis of the 

 factor to be studied. In any case, the law of averages will not elimi- 

 nate the effect of magnitude of business if the factor to be studied 

 itself depends on magnitude of business. 1 This difficulty may 

 be obviated as follows: First group the farms on the basis of magni- 

 tude of business, making the limits of the groups as small as possible 

 and still have enough farms (50 to 100 if possible) to give a reliable 

 average for the labor income in each group; then express the labor 

 income of each farm in percentage of the average labor income of 

 its size group. We may call this percentage the "adjusted labor 

 income." It is a measure of success in farming as affected by fac- 

 tors other than magnitude of business. Thus if the labor income 

 of a 50-acre farm is 80 per cent of the average for its group, and that 

 of a 125-acre farm is 80 per cent of the average for its group, these 

 two farms are considered to be equally efficient, though the larger 

 farm may have much the larger labor income. 



This adjusted labor income should be calculated for each farm 

 and recorded on the survey record sheet, or the special factor card 

 if there is one. It is especially useful in dealing with factors which 

 are influenced by size of farm. Thus, in studying the relation 

 between type of tenure and labor income, we are confronted with 

 the difficulty that in some localities, especially in the Northern 



1 The law of averages, which plays so important a part in the study of farm man- 

 agement survey data, may be briefly stated as follows: In cases where variations 

 from a normal are just as liable to be plus as minus, the plus and minus variations 

 will cancel each other if tke number of cases averaged is large. 



