19 



Productive work units per $1 of outside labor: Man. Horse. 



Outside labor, man and team 0.2 0.4 



Outside labor, man only 0. 4 



Outside labor, marketing (man and horse) 0. 2$ 0. 2f 



PRINCIPLES OF TABULATION. 



The interpretation of farm management survey data consists in 

 the main of the discovery and measurement of causal relations be- 

 tween the various classes of facts contained in the data. In re- 

 search work in general such causal relations are studied by pro- 

 ducing changes in one class of phenomena and observing the 

 resulting changes in others. In farm management survey data the 

 variations to be studied are already at hand, the problem being to 

 arrange the data in such manner as to display the variations of one 

 factor and then note the variation in others supposed to depend 

 in some way on it. 



A farm management survey record consists of the various items 

 constituting an analysis of the farm business. It is the interde- 

 pendence of these various items that is to be studied. The general 

 problem is to find what relation exists between the magnitude of the 

 various items and the efficiency of the farming. Farm management 

 studies have shown that success in farming depends largely on the 

 following factors: (1) Type of land tenure; (2) type of farming; 

 (3) magnitude of business; and (4) the various factors of efficiency, 

 such as yield per acre, income per cow, adequacy and economy of 

 equipment, diversity of business, productive work units per man 

 and per horse, etc. 



In any particular case the problem is to discover the relation of 

 the above factors to success in farming under the conditions which 

 prevail in the locality to which the data relate. 



Consideration of a particular case will best serve to show the 

 principles involved in a study of causal relations. The relation 

 between s' ze of farm and labor income may be used for this purpose. 



Two methods of procedure present themselves. First, we may 

 arrange the farms in groups according to increasing size of farm 

 and find the corresponding variations in the average labor income 



