STATISTICAL TREATMENT 



245 



putting the variables concerned into relationship to each other. Where 

 there are but two variables we may at times take the ratio of one variable 

 to the other and give our solution in terms of this ratio, but in most 

 problems the solution will involve the formulation of some other equa- 

 tional relationship. Such problems will involve curve fitting by either 

 an analytical or graphical method, and the equational relationships may 

 be based on some form of rationalization or may be entirely empirical. 

 Since in science we aim at as simple a rationalization as possible, the 

 results of many problems are stated either in terms of a linear equation 

 or in terms of an exponential equation. These two equations allow the 

 treatment respectively of the case where one variable changes at a uniform 

 absolute rate with regard to the other, and the case where one variable 

 changes at a constant relative rate with regard to the other. Since these 

 two equations are so widely applied, the methods of treating them will be 

 developed in the two following illustrations: 



As a possible application of a linear equation, let us consider the 

 observations reported by Laurens (2) on the effect of carbon-arc radiation 

 on essential hypertension in man. We may limit our discussion to two 

 of his measurements, that of systolic blood pressure before and after 



Table 8. — Systolic Blood Pressure of Subject F before and after Treatment 



WITH Carbon-arc Radiation 



